Crime, Disorder and

Community Safety

Major changes are taking place in the crime control industry and by association
in the discipline of criminology itself. These changes, which involve the adop-
tion of local partnership approaches, new forms of managerialism as well as the
blurring of distinctions between crime, disorder and community safety, have
been described as constituting a watershed for criminology. Bringing together
nine original articles from leading national and international authorities on
these issues, this book is both a product of and a reflection on the changing
state of play. The articles address, from a range of different perspectives, the
changing nature of regulation in contemporary society and the simultaneous
reconstruction of the communities which are being regulated. Crime, Disorder
and Community Safety also offers a timely commentary on government attempts
to address questions of inequality and social exclusion as well as issues relating
to ethnic minorities and young people.

Roger Matthews is Professor of Sociology at Middlesex University. John Pitts is

Vauxhall Professor of Socio-Legal Studies at the University of Luton.
Crime, Disorder and
Community Safety
A new agenda?

Edited by Roger Matthews

and John Pitts

London and New York

First published 2001
by Routledge
11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canada

by Routledge
29 West 35th Street, New York, NY 10001

Routledge is an imprint of the Taylor & Francis Group

This edition published in the Taylor & Francis e-Library, 2002.
© 2001 Roger Matthews and John Pitts for editorial matter and selection;
individual contributors their contribution
All rights reserved. No part of this book may be reprinted or reproduced
or utilised in any form or by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying and recording,
or in any information storage or retrieval system, without permission in
writing from the publishers.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data

Crime, disorder, and community safety: a new agenda?/edited by Roger
Matthews and John Pitts.
Includes bibliographical references and index.
1. Crime prevention. 2. Crime. I. Matthews, Roger. II. Pitts, John.

HV7431 .C7 2001

364–dc21 00-065337

ISBN 0–415–24230–4 (hbk)

ISBN 0–415–24231–2 (pbk)
ISBN 0-203-47099-0 Master e-book ISBN

ISBN 0-203-77923-1 (Adobe eReader Format)

Contents

List of illustrations vii

Notes on contributors viii
Acknowledgements x

Introduction: beyond criminology? 1

R O G E R M AT T H E W S A N D J O H N P I T T S

1 Identity, community and social exclusion 26

JOCK YOUNG

2 Joined-up but fragmented: contradiction, ambiguity and

ambivalence at the heart of New Labour’s ‘Third Way’ 54
A D A M C R AW F O R D

3 Evaluation and evidence-led crime reduction policy

and practice 81
NICK TILLEY

4 Crime, community safety and toleration 98

LY N N H A N C O C K A N D R O G E R M AT T H E W S

5 ‘Broken Windows’ and the culture wars: a response to

selected critiques 120
GEORGE L. KELLING

6 Ethnic minorities and community safety 145

MARIAN FITZGERALD

7 The new correctionalism: young people, Youth Justice and

New Labour 167
JOHN PITTS
vi Contents

8 Crime victimisation and inequality in risk society 193

TIM HOPE

9 Distributive justice and crime 219

PA U L W I L E S A N D K E N P E A S E

Index 241
Illustrations

Tables
1.1 The binaries of social exclusion 43
8.1 Household and area risk factors for household property crime 202
8.2 Indices of individual and area deprivation and affluence 206
8.3 Property crime victimisation, deprivation and affluence 207
8.4 Correlation between property crime victimisation,
(a) household deprivation (b) household affluence, within
levels of areas deprivation 208
8.5 Risk factors for household property crime victimisation
by level of area deprivation 209
9.1 Risks of domestic burglary by household type 232
9.2 OCU rate variability by offence types 236

Figures
6.1 Arrests and searches: London 1997–1999 150
6.2 Searches, population and suspects: London 1998–1999 152
6.3 Recorded racial incidents in England and Wales, 1991–1999 157
6.4 Trends in types of racial incidents recorded by the
MPS, 1997–2000 158
6.5 Victims and suspects of racist incidents by ethnic group, 159
MPS 1997–2000
7.1 The Crime and Disorder Act (1998) 170
8.1 Inequality and property crime 203
8.2 Property crime rates and poverty concentration 210
9.1 Property crime by area crime levels 229
9.2 Property crime by area victim prevalence 230
9.3 Property crime by area victim concentration 231
9.4 Inequality at force and OCU level 233
Contributors

Adam Crawford is Professor of Criminology and Criminal Justice and the

Deputy Director of the Centre for Criminal Justice Studies, University of
Leeds. He is the author of Local Governance of Crime (1997), Crime
Prevention and Community Safety (1998) and co-editor of Integrating a Victim
Perspective within Criminal Justice (with J. Goodey, 2000). He has worked for
both the New Zealand Ministry of Justice and the Northern Ireland Office in
relation to crime prevention policy. He is part of a Home Office-funded team
which evaluates youth justice reforms, introduced by the Youth Justice and
Criminal Evidence Act (1999).
Marian FitzGerald is currently a Visiting Research Fellow at the London
School of Economics and has researched many aspects of British race rela-
tions policies. Her earliest work was on relations between the main political
parties and ethnic minorities, and she is the author of a report, entitled
Ethnic Minorities and the Criminal Justice System, for the Royal Commission on
Criminal Justice. Her recent publications include a major study of police
searches in London and she is presently co-directing new research into police
community relations in the capital.
Lynn Hancock is Senior Lecturer in Criminology at Middlesex University.
Before moving to Middlesex in 1997, she was a Lecturer in Criminology at
Keele University (1993–1996). She held a Leverhume Special Research
Fellowship between 1996 and 1998. Her research ‘Neighbourhood Change,
Crime and Urban Policy’ will be published in her forthcoming book Crime,
Community and Disorder.
Tim Hope is Professor and Head of the Department of Criminology at Keele
University, England. He has held positions at the Universities of Manchester and
Missouri-St. Louis, CACI Ltd., and the Home Office, where he was Principal
Research Officer. His research interests and publications lie in the fields of crime
prevention, community safety, victimology and the evaluation of crime preven-
tion programmes, where he is currently leading a consortium to evaluate the
Home Office Reducing Burglary Initiative. He recently edited Perspectives on
Crime Reduction (2000) and co-edited Crime, Risk and Insecurity: Law and Order
in Everyday Life and Political Discourse (with Richard Sparks, 2000).
 Contributors ix

George L. Kelling is a Professor in the School of Criminal Justice at Rutgers

University, a Research Fellow in the Kennedy School of Government at
Harvard University, and a Senior Fellow at the Manhattan Institute. He is
the co-author (with James Q. Wilson) of ‘Broken Windows’, published in
Atlantic Monthly in 1982, and Fixing Broken Windows: Restoring Order and
Reducing Crime in our Communities (with Catherine M. Coles, 1996).
Roger Matthews is Professor of Sociology at Middlesex University. His current
research interests are centred around issues relating to toleration and public
attitudes towards crime and justice. He has published a number of books
including Informal Justice (1988), Privatizing Criminal Justice (1999) and
Rethinking Criminology: The Realist Debate (with J. Young, 1992). His most
recent book is Doing Time: An Introduction to the Sociology of Imprisonment
(1999).
Ken Pease is Professor of Criminology at Huddersfield University. He has previ-
ously held Chairs at the Universities of Manchester and Saskatchewan. He is
currently on secondment to the Home Office Crime Reduction Programme
and sits on the Crime Panel of the DTI Foresight Programme.
John Pitts is Vauxhall Professor of Socio-Legal Studies at the University of
Luton. He has worked as a ‘special needs’ teacher, a street and club-based
youth worker, a youth justice development worker and a group worker in a
Young Offender Institution. His recent publications include Planning Safer
Communities (1998), Working with Young Offenders (1999) and Positive
Residential Practice: Learning the Lessons of the 1990s (2000).
Nick Tilley is Professor of Sociology at the Nottingham Trent University and a
consultant to the Home Office Policing and Reducing Crime Unit. He is the
co-author of Realist Evaluation (with Ray Pawson, 1997) and co-editor of
Surveillance of Public Space: CCTV, Street Lighting and Crime Prevention (with
Kate Painter, 1999).
Paul Wiles is currently Director of Research, Development and Statistics at the
Home Office. Previously, he was Professor of Criminology at the University
of Sheffield. His recent research interests have been in the analysis of
geographical crime patterns and more broadly in the ecology of crime. In the
past he has worked in various areas of research on crime and socio-legal
studies.
Jock Young is Professor of Sociology and Head of the Centre for Criminology at
Middlesex University. His first book was The Drugtakers (1971) and his most
recent book is The Exclusive Society (1999). He is the co-author of The New
Criminology (with Ian Taylor and Paul Walton, 1973). His current research
interests are in urban and cultural studies, and he is working on a book to be
entitled Strange Days: The Transition to Late Modernity.
Acknowledgements

This book developed out of a series of seminars and workshops which took place
in London and Luton in 1998. The seminar series was funded by the ESRC
through their Research Seminar Series. We would therefore like to thank the
ESRC for their support. We would also like to thank all those who took part in
the seminar programme. The participants included Afshan Ahmet, Rob Allen,
David Barrett, Jon Bright, Charlie Burr, Adam Crawford, Helen Edwards, Robin
Fletcher, Roger Grimshaw, Lynn Hancock, Simon Harding, Tim Hope, Mike
Hough, Michael Hutchinson-Rees, George Kelling, Alan Marlow, Kate Painter,
Ken Pease, David Porteous, Nick Tilley, Norman Warner and Paul Wiles. We
would also like to thank Julie Grogan for her assistance in organising the
various meetings and the staff of Middlesex University who provided the neces-
sary administrative support.
Roger Matthews and John Pitts
October 2000
Introduction
Beyond criminology?
Roger Matthews and John Pitts

Introduction
It used to be fashionable to claim that criminology was in crisis (Bottoms and
Preston 1980; Young 1986). This diagnosis reflected a general mood of uncer-
tainty and called for a reconsideration of the appropriate focus of criminological
investigation. The mood, however, has changed over the past few years and a
growing number of books and articles have appeared which suggest that crimi-
nology is currently in the process of fragmentation and dissolution (Ericson and
Carriere 1994). These recent developments have been attributed to changes in
the nature of social relations and to associated changes in the organisation of
social control (Garland 1996; Lea 1998). No longer, it is argued, is the state
purely concerned with individual acts of deviance. Increasingly, the focus is on
security and the control of aggregate populations differentiated according to
assessments of risk and dangerousness. The emergence of the so-called ‘risk
society’ has brought in its wake new modes of inclusion and exclusion (Sibley
1995; Young 1999). Some critics propose that these changes have in turn
eroded the conventional boundaries between social science disciplines as the
objects of investigation begin to overlap and as narrow and parochial interests
become less relevant.
Paradoxically, while some academics are beginning to mourn the demise of
criminology as a discipline, the size of the crime control industry and the
number of students studying criminology in many countries around the world is
rapidly growing (Braithwaite 2000). In England and Wales the official annual
cost of crime control currently stands at £8.2 billion, and this is projected to
increase to £10.6 billion by 2003–4. These costs do not, of course, include the
millions of pounds spent every year by private individuals and commercial
organisations. Recent reports from the US indicate that in some states expendi-
ture on crime control has exceeded that on education, while the amount spent
by federal, state and local governments in 1996 was a staggering $120 billion for
civil and criminal justice. This represented a 70 per cent increase over the
previous year (Currie 1998; Bureau of Justice 2000).
The growing number of people studying criminology is on the one hand a
function of the massive increase in media time dedicated to crime and
2 Roger Matthews and John Pitts

punishment and, on the other hand, a result of the increased likelihood of

employment within the burgeoning criminal justice industry. Thus, it would
seem that while the debates in criminology are expressing a sense of exhaus-
tion in relation to the subject area, the number of students studying
criminology is increasing steadily, as is the annual expenditure on ‘law and
order’.
It must be said that there has for some time been growing doubts about the
coherence of criminology as a discipline. Many writers have commented upon
the elusive and narrow conception of ‘crime’ which is used among academics
(Matthews and Young 1992; Muncie 2000; Schwendinger and Schwendinger
1975; Shearing 1989). If it is the case that the concept of ‘crime’ does not
provide a solid enough base for the construction of the discipline, it might be
suggested that its coherence lies in an established set of theories, or in a
concerted body of knowledge, or alternatively in a particular tradition or mode
of enquiry (Garland and Sparks 2000; Sparks 1997). Some commentators have
asked whether criminology is better seen as a subset of a grander and more
robust discipline, such as sociology, politics or psychology, or whether it should
be seen as little more than an amalgam of knowledges pieced together from a
range of disciplines. The answer depends on the degree of coherence one affords
to the body of knowledges which have become identified as ‘criminology’.
Clearly the suggestion that criminology is in the process of fragmentation
assumes a prior state of unity, which may be more imagined than real.
The uncertainty over the status of criminology has intensified as a result of a
series of debates which have taken place over the past two decades questioning
the proper object of enquiry of criminology from advocates of victimology,
radical and critical criminology and from feminist criminologists. The victi-
mology movement changed the primary focus from the offender to the victim
and in the process drew attention to the limitations of purely offender-based
criminology (Maguire and Pointing 1988; Mawby and Walklate 1994). The
second level of critique came from radical criminologists who took issue with
what they saw as the unduly narrow focus of conventional criminology, which
paid relatively little attention to white-collar crime, corporate crime and crimes
of the state. Together, these new radical and anti-criminologies demanded a
fundamental review of the established agenda of criminological investigation,
particularly where it was seen to deny progressive and emancipatory interests.
The third level of critique was mounted by feminists who claimed that the
assumptions of gender neutrality implicit in major criminological theories were
in fact misguided (Leonard 1982). Traditional criminology was found guilty of
being androcentric and gender biased. Moreover, it was argued that many key
issues in criminology had been neglected by male researchers or defined in
masculinist terms and that these issues required investigation from a feminist
standpoint (Cain 1990; Smart 1990).
Most of these critiques, however, did not involve the negation of crimi-
nology as such, but instead called for the revision and a reorientation of the
subject. These so-called anti-criminologies at their most constructive intro-
 Introduction: beyond criminology 3

duced new concerns, a wider range of issues and perspectives and generally
enriched the subject area. At their worst they were irrelevant and indulgent.
Much of the debate centred around the ‘real’ meaning of crime and the
construction of a ‘proper’ criminology. While these critical, radical, feminist and
realist criminologies attempted to change the focus of criminology, they
succeeded in reshaping and to some extent re-ligitimating, the subject. More
recent contributions, however, have been concerned less with the reorientation
of the subject and more with its demise. This change of perspective has been
stimulated by widespread socio-economic developments and changing forms of
governance which have been associated with globalisation and the transition
from modernity to late modernity – or as some would have it, postmodernity
(Bauman 1995; Harvey 1989). These changes, in turn, involve a unique histor-
ical mutation in which the distinctions between crime and disorder are
becoming increasingly blurred and both are becoming incorporated within the
wider framework of community safety. In this process it is becoming increasingly
difficult to treat crime as a discreet object of study and to sustain criminology as
a distinct academic discipline.

From crime prevention to community safety

The period from the mid-1970s to the end of the 1990s witnessed a shift in
focus from situational to social crime prevention, to disorder and eventually to
community safety. This development can be regarded as one of a gradual
widening out of the analytic focus and at the same time a blurring of the
distinctions between these processes.
The interest in crime prevention was given a new lease of political and
academic life by Ron Clarke in the late 1970s. Clarke, who was Director of the
Home Office, played a key role in developing situational crime prevention poli-
cies by seeking to provide a response to crime which was both cost effective and
politically and socially acceptable (Clarke 1992). The enormous impact of these
situational measures is difficult to gauge retrospectively since many of them
have subsequently become an integral part of our everyday lives. By the mid-
1980s, however, problems of displacement and of ‘designing out’ a range of
serious crimes, including rape, domestic violence and racist attacks, indicated
the limits of the situational perspective and encouraged the promotion of social
prevention measures in the form of strengthening neighbourhoods by encour-
aging citizen involvement and fostering informal controls (Currie 1988;
Rosenbaum 1986).
However, by the mid-1990s the promise of developing more effective social
crime prevention measures was called into question by the observation that
crime control strategies were most difficult to mobilise in the neediest areas and
that dealing with crime in high-risk areas required a different type of approach.
It was increasingly felt within British policy circles that the problem of crime
required a more comprehensive and better co-ordinated response while
retaining a local focus. A series of inter-agency initiatives were established
4 Roger Matthews and John Pitts

during the 1980s to address the issue of crime, and a number of agencies have
become increasingly involved in the prevention and reduction of different
forms of crime. The logic of these developments was endorsed and extended by
the Morgan Report (1991), whose recommendation that the responsibility for
crime prevention and community safety should be devolved to local authorities
was implemented by an incoming Labour government. The subsequent passing
of the Crime and Disorder Act (1998) consciously linked crime control with
disorder and saw both as part of a more comprehensive community safety
strategy (Hughes 1998).
The shift from crime prevention to community safety has extended the terms
of reference and made everything from crime and incivilities to environmental
issues part of a more comprehensive agenda (Muncie 2000). The fact that
community safety is as elusive and loosely defined a term as crime only serves to
extend the open-endedness of this new agenda. Crime has increasingly become
fused with disorder and both have become enveloped within the more general
notion of community safety, despite the fact that the focus of the three terms,
the nature of the events involved, the established forms of intervention
employed and the characteristics of the victims is remarkably different. It is not
so much that crime has become normalised but rather it has become absorbed
within a more complex set of overarching relations. This widening of the frame
of reference has been reflected in the language and concepts which have been
used by policy-makers to address these issues. Such terms as ‘dangerousness’,
‘anti-social behaviour’, ‘security’ and ‘social exclusion’ are necessarily ill-defined
(Brown and Pratt 2000). The wide-ranging terminology represents a shift away
from the style of language that has been used in relation to the ‘war against
crime’, which talks of ‘targets’ and ‘campaigns’. From this ‘get tough’ crime
control perspective, the claim is that there are a relatively small number of
offenders, victims, groups, neighbourhoods (‘hot spots’) who are ‘responsible’ for
a disproportionate amount of crime and victimisation. However, as the issue of
crime becomes further integrated into the wider matrix of community safety the
focus necessarily shifts to aggregate problem populations – the underclass, the
marginalised and the socially excluded. Consequently, a hybrid discourse is
emerging which at one moment identifies the problem as a limited number of
deviant individuals, while at another regards whole sections of the population
as a problem. Whereas the war against crime is seen as something which can be
‘won’ and criminals reformed or deterred, the terms ‘underclass’ and the ‘socially
excluded’ come close to suggesting a permanent condition which has to be
managed rather than eliminated (Feeley and Simon 1994). Policy-makers and
practitioners espouse military metaphors and then in the same breath refer to
social support and individual needs. This new hybrid discourse is appearing in
various local authority-generated policy documents and in the newly established
sites such as local authority community safety forums. The multi-agency forums
are well situated to cultivate this ambiguous discourse since they typically incor-
porate the coercive powers of the police alongside the more welfare-oriented
approaches of social services and education departments. The formalisation of
 Introduction: beyond criminology 5

these multi-agency interventions has created new alliances and new lines of
responsibility. Importantly, the role of the key regulatory agencies is changing
significantly in this process. The police in particular have been ‘let off the hook’
and can now legitimately claim that neither crime control nor community
safety is their sole responsibility (Kinsey et al. 1986). At the same time, they are
required to participate in a wider range of interventions which shifts the
balance of their work away from crime control and towards public order and
social work functions.
It has in fact been evident for some time that crime is related to almost every
other negative indicator in society – poor health, limited educational facilities,
unreliable transport, bad housing and the more serious environmental problems.
Indeed, these negative indicators predictably overlap to the extent that high
crime areas also rank high on every other scale of social and economic misfor-
tune. Thus, multiple victimisation (the tendency for the same individual or
household to be a victim of a range of hazards) is becoming a more appropriate
point of reference than the more linear concept of repeat victimisation, which
has shown that individuals and victims are subject to serial forms of victimisa-
tion (Genn 1988; Farrell and Pease 1993; Sparks et al. 1977). Multiple
victimisation is the logical counterpart of multi-agency intervention, since one
of the original justifications for developing a multi-agency approach was the
recognition that a range of different agencies often intervene independently in
the lives of certain individuals in response to the various manifestations of the
same underlying problem.
The implications of this broader focus on the inter-related and the socially
and geographically concentrated nature of social problems is threefold. The first
is that crime may begin to lose its privileged position among other social prob-
lems. It may be the main problem in certain areas and for certain individuals,
but in the majority of places it will only constitute one component of an inte-
grated matrix of hazards. Second, it calls for a more comprehensive mode of
theorising and a consequent shift away from the preoccupation with piecemeal
and situational measures to forms of explanation which are more comprehen-
sive and can account for the concentration and compounding of different types
of victimisation. Social support theory provides one way of addressing these
issues. Its basic proposition is that the best predictor of the areas and households
with high levels of multiple victimisation are not those with low levels of social
control, as conservative criminologists have argued, but rather those with low
levels of social support (Cullen 1994). The difficulty, however, in a post-welfare
era is that the traditional networks of state and social support are themselves in
the process of dissolution and transformation. Consequently, introducing appro-
priate systems of support requires the development of new, innovative,
interventionist strategies. The third implication is that because multiple victim-
isation moves beyond a preoccupation with specific crimes, it then raises the
wider issue of social justice. The culmination of these processes may possibly
result in the radical reorientation of criminology, both as a framework of inter-
vention and of investigation. It will, however, almost certainly foster forms of
6 Roger Matthews and John Pitts

enquiry that transcend strict disciplinary boundaries and encourage researchers

to explore the relation between crime and other problems.
The shifting focus from crime prevention to community safety also raises the
question of what is meant by ‘community’. The assumption which lies behind
much of the community safety rhetoric is that society is comprised of a number
of fairly homogeneous communities expressing a high degree of consensus. But
as Jock Young argues in Chapter 1, the notion of ‘community’ itself is becoming
increasingly problematic, and any realistic community safety strategy needs to
come to terms with the changing nature of these communities and associated
forms of identity if they wish to address the issues of crime and community
safety.
The integration of crime control and community safety signals that everyone
is vulnerable to some degree, not only to the widely acknowledged forms of
crime and victimisation but also to a whole range of unforeseen hazards. From
the vantage point of community safety it follows that all citizens are more or
less directly concerned with the issue of security. The shift in responsibility to
local authorities potentially opens the door to new and broader forms of
accountability since there is a perceived need to consult with a wide body of
local opinion in order to identify priorities, elicit co-operation and establish
support. However, rather than create greater transparency, the move towards
local authority-driven community safety programmes has instead created new
forms of complexity and opacity. This has been achieved through a combina-
tion of the development of new forms of managerialism as well as through
decentralisation and privatisation.

The New Managerialism

The New Managerialism plays two important roles in this changing situation.
First, it provides an ostensibly neutral language in which policy formation and
decision-making can be located which serves to mediate the tensions generated
by the emerging hybrid discourses. The second major attribute of the New
Managerialism is its capacity to turn political and moral decisions into adminis-
trative and technical ones. In the name of cost-effectiveness, policy options are
counted and discounted. Within the performance indicator culture, which is
actively fostered by the New Managerialism, only that which is quantifiable is
relevant. The fact that many of these performance indicators are uncertain is
generally overlooked in the rush to produce figures and meet targets. The
reliance, for example, on official crime statistics as a measure of the effective-
ness of specific interventions or the efficiency of the relevant agencies is
extremely dubious. Although there has recently been some recognition in offi-
cial circles of the limitations of using official crime figures in this way, the
general unreliability of many performance indicators is not routinely addressed
(Home Office 2000). The performance indicator culture has encouraged new
forms of expertise in the manipulation of figures and in prioritising those activi-
ties which are more directly measurable, irrespective of their overall
 Introduction: beyond criminology 7

contribution to designated objectives, the interests of the ‘clients’ or the realisa-

tion of justice (Loveday 1999). The requirement of local authorities to produce
regular crime audits has also provided an example of how a body of unrelated
figures, drawn from disparate and unchecked sources, based in many cases on
dubious methodologies, can be cobbled together and used to provide a bench-
mark from which ‘progress’ is measured.
New Labour was quick to appropriate both the language and the principles of
the New Managerialism. Its emphasis on the rationality of ‘what works’, ‘joined-
up solutions’ and ‘empowerment’ fitted well with New Labour’s avowed
commitment to rational public policy and the devolution of power to ‘local
people’. As Anthony Bottoms (1995) has argued, the adoption of manageri-
alism within criminal justice incorporates a significant actuarial dimension that
presents policy choices as a function of competing risks which can be assessed
on an objective scale as well as on the basis of cost-effectiveness. Balancing off
these two apparently neutral and ‘scientific’ sets of criteria, however, opens up
considerable latitude for interpretation and discretion. But the central appeal of
the New Managerialism is that it offers a template for the new, more restricted
role of government – a ‘Third Way’, which New Labour believed was necessi-
tated by the unmanageability of the global economy and the privatisation of
personal life precipitated by the radical changes in the class structure of Britain
over the preceding thirty years (Giddens 1998). As Cooper and Lousada note:

The state of affairs described here arises largely where government has
vacated territory which it once organised, resourced, nurtured and presided
over with political authority – even if contested authority – but for which it
nevertheless will not or cannot abandon all responsibility. Where once it
willed the means, now it tends to prescribe the ends while others are left to
supply the means; and as with all instrumentally-oriented practices, the
logic governing the attainment of ends is the logic of rules. In short, where
government was, now audit regulation and management is. And so,
inevitably, this social ego is pervaded by obsessional anxieties and
behaviours about the loss of control.
(Cooper and Lousada 2000: xx)

Thus, the Crime and Disorder Act (1998), in handing new responsibilities over
to local authorities and local multi-agency partnerships, prescribes the goals
they should pursue, the targets they should achieve and the time scales in which
they should operate, but not the means for their achievement. However, as
Cooper and Lousada (2000) suggest, the anxieties about the political legitimacy
and electoral credibility engendered by this new form of ‘hands-off’ governance
leads to an unprecedented concern with the management of implementation,
which is specified in such detail that what at first sight appears to be an exercise
in devolution upon closer inspection looks to be a franchising operation.
‘Product support’ is provided via a steady flow of ‘evidence-based’ data about
‘what works’ and ‘pilot studies’ of new administrative arrangements and modes
8 Roger Matthews and John Pitts

of intervention, while ‘quality control’ is vouchsafed by statutory audits.

Meanwhile, ‘action teams’, comprising senior officials from the relevant
ministries, work on inter-ministerial synchronisation and the articulation of
central government activity with that of the local authorities and multi-agency
partnerships.
New Labour’s community safety strategy similarly devotes an inordinate
amount of time and energy to the detail of its day-to-day management and
implementation – the development of partnerships, monitoring systems and
the like – but little to any analysis of the origins and nature of the problems to
which these elaborate, joined-up solutions are to be applied. In the 115-page
Audit Commission report, Safety in Numbers (1999), for example, while only
half a page is devoted to the origins of crime, David Farrington and David
West’s (1993) by now familiar ‘risk factors’ are presented as the appropriate
mode of explanation, although it is noted in the report that: ‘Risk factors are
not causes of crime as such, but the likelihood of criminality increases as their
intensity or clustering increases’. The origins of these phenomena, or why they
are ‘clustered’ in the ways that they are, is apparently of little interest.
In the view of the government, the ‘joined-upness’ of solutions to youth
crime and community safety consists in the simultaneous management of those
‘risk factors’ associated with the onset of crime and delinquency, and the appli-
cation of ‘evidence-based’ technologies for the correction of familial, cognitive
or behavioural deficiencies which lead to persistence. Alongside this, systemic
anomalies are eradicated by rigorous audits and the imposition, by central
government, of financial penalties upon local government, the police and the
Crown Prosecution Service.
There is, however, an ongoing tension between the New Managerialism and
multi-agency partnerships since interventions are not so much mobilised by the
seriousness or urgency of the particular issue, but rather, in many cases, they are
motivated by the availability of funding the interests of specific agencies. The
actual decisions about priorities are often dependent upon personalised prefer-
ences or political priorities rather than a rigorous assessment of risk. Moreover,
lines of responsibility in multi-agency networks are frequently blurred with the
consequence that roles are unspecified and interventions may lack clear lines of
command and decision-making. Consequently, interventions tend to be uncer-
tain and poorly conceived, while active involvement may be short-lived,
particularly where agency and individual interests compete (Crawford 1998). In
many cases differences of objectives, styles of operation, and interests are played
down in order to sustain a loosely constructed and fragile consensus. The
increasingly diverse and amorphous nature of the object – community safety – is
reflected in the equivocal nature of the mechanisms which are responsible for
achieving it – multi-agency partnerships. As Adam Crawford argues in Chapter
2, the partnership approach, the New Managerialism and communitarianism
form the three central strands of New Labour’s ‘Third Way’ strategy, which aims
to move beyond old-style statism, on the one hand, and a reliance on the opera-
tion of a free market economy on the other. Each of these separate strands has a
 Introduction: beyond criminology 9

certain logic and coherence on its own, but when ‘joined-up’ they produce
unanticipated tensions.

Decentralisation and privatisation

A consequence of the political and academic onslaught which sought to limit
the powers of the sovereign state in the 1960s and 1970s was the development
of a decentralisation strategy through which state power would be dispersed and
in which a series of more approachable, accountable and more efficient agencies
would come to provide essential services. In the field of ‘law and order’ this
decentralising impulse took the form of delegalisation, decarceration and depro-
fessionalisation. This resulted in the proliferation of different forms of
privatisation, the expansion of the voluntary sector, the creation of greater
autonomy among regulatory agencies and the devolving of responsibility for
crime control to community organisations. In this process the aim was seen to
be to gradually shed state power from a central axis to a wide range of regulatory
bodies and for the state to become involved in ‘steering’ rather than ‘rowing’
(Cohen 1987).
This destructuring impulse provided the impetus for the development of new
forms of decentralised community control which was seen to promote forms of
inclusion by dealing with deviancy in more immediate and informal ways
within the community (Cohen 1985).The optimism which had been associated
in the early period with these destructuring tendencies, however, gradually
turned into pessimism. Critics argued that what drove these more informal
community-based interventions was not their greater effectiveness or benevo-
lence. Rather, they were adopted because they were cheaper and provided a way
of reducing state obligation to provide decent and consistent services, while
serving to depoliticise the process of crime control by turning public issues into
personal problems. At the same time, the proliferation of these apparently
benign agencies meant that the depth and intensity of intervention in many
cases was increased. As a result of ‘net widening’, it was argued, these interven-
tions were seen as being largely counterproductive since a growing number of
minor offenders were drawn ineluctably into the mainstream criminal justice
process with a predictable proportion ending up in prison at a later date
(McMahon 1990).Thus, rather than producing new forms of inclusion, the
decentralising strategy resulted in promoting new forms of social exclusion and
marginalisation. For some critics a disturbing consequence of the decentralising
impulse was that existing forms of state control remained in place, and rather
than curtail state power, this strategy produced an unwieldy complex of agencies
and organisations which traversed the conventional distinctions between the
public and the private realms and between state and civil society. Instead of a
straightforward process of decentralisation, what seemed to take place was a
complex blend of decentring and recentring processes, resulting in the prolifera-
tion of different sites of decision-making, increased bureaucratisation and a
10 Roger Matthews and John Pitts

more diffuse set of interventions arising from a growing array of increasingly

unaccountable agencies.
In retrospect, the model of the state power which the majority of criminolo-
gists held at this time was extremely limited. Consequently, most of the key
concepts which were used to explain these developments remained underdevel-
oped:

Take the very notion of decentralisation itself. This derives from a master
metaphor which sees political power in terms of centre and periphery. We
are asked to imagine such things as deviants being drawn away from the
centre, or the periphery being awarded the power that belonged to the
centre. But this metaphor rests on a largely unexamined view of state
power, which at times leans towards the most extreme forms of elitist
centralism, at other times towards the most amorphous forms of pluralism.
(Cohen 1987: 377)

In relation to implementation there was also a naïveté about the role and
interests of these newly formed or reconstructed agencies. It became apparent
that the agencies had their own specific interests, not least of which was their
own preservation and expansion. Less clear was the increasingly intricate
network of policy formation and implementation, the financial links and pecu-
niary constraints which were imposed, the hierarchical relations which
developed, and the growing pressure for these apparently autonomous and inde-
pendent bodies to pursue the same goals and adopt the same management
practices which had been previously developed in the existing state agencies.
Alongside such uncertainties was a lack of clarity about the ways in which
newly state-sponsored agencies were being established to monitor, assess, guide
and regulate the ‘independent’ agencies, such that the apparent cost savings
which were claimed were in some cases more than swallowed up by the costs of
monitoring and supervision.
The criticisms of the decentralisation strategy which surfaced in the mid-
1980s, however, did little to slow down this process, although they may have
influenced its direction. Privatisation in its various forms continued to be
promoted, particularly in relation to policing, crime prevention, the use of secu-
rity equipment and the contracting out of prison management (Johnston 2000;
Matthews 1989). The rapid development of multi-agency partnerships encour-
aged the involvement of the private sector alongside voluntary and statutory
agencies.
In the UK in the 1990s the decentralising impulse took a slightly different
turn following the implementation of the recommendation of the Morgan
Report (1991) to devolve responsibility for crime control and community safety
to local authorities. Emphasis was placed upon multi-agency partnerships and
the establishment of community safety strategy groups which were to be
responsible for assessing the nature of crime in each locality, developing a
community safety strategy and monitoring interventions. During the 1980s, the
 Introduction: beyond criminology 11

Thatcher-led Conservative government had tried to limit the powers of local

authorities, which were seen as unreliable, but a decade later this policy was
reversed by an incoming Labour government and local authorities in conjunc-
tion with the local police were given statutory responsibility for crime
prevention and community safety. Local authorities were seen as uniquely
placed to connect with the local population and thereby able to increase
citizen involvement in crime control while mobilising the wide range of agen-
cies which were deemed necessary to develop a comprehensive community
safety strategy – housing, planning, social services, education, leisure services,
youth services and others.
The potential problems associated with devolving responsibilities for crime
control and community safety in this way received limited attention. The urge
to shift responsibility to local authorities played down the existing unequal
distribution of needy people in different areas, as well as the geographical
concentration of victimisation. The placing of greater responsibility on local
authorities for this increased range of problems has further burdened certain
municipalities. Thus this development may be seen as a recipe for the consoli-
dation rather than the alleviation of injustice. At the same time it should be
noted that creating greater autonomy in local decision-making does not neces-
sarily involve greater empowerment or increased accountability. Nor does it
provide an alternative to the growth of a crime control bureaucracy, but instead
broadens its base. Although crime in itself may be locally extremely variable,
the level of crime in any particular area will also be a product of a wider process
of deflection and displacement as well as a function of social and economic
movements which affect the make-up and distribution of local populations
(Barr and Pease 1990). Similarly, community safety in the vast majority of
localities will be affected by environmental, health and transport issues which
reach far beyond the geographical boundaries of local boroughs and over which
particular local authorities may have little or no control (Young 1990).
Therefore, in relation to crime and community safety there are unresolved
issues about whether the local authority is the appropriate regulatory site and
whether it might be more appropriate to locate these responsibilities in regional
authorities if the aim is to achieve greater levels of distributive and social
justice.
This shift of responsibilities is likely, as has been suggested above, to have a
profound impact on the way in which crime control is perceived and carried
out. Although it is too early to properly assess the impact of this form of decen-
tralisation, the devolvement of powers and responsibilities to local authorities
could take one of two directions. The first, which has been the main develop-
ment to date, is that it could stimulate styles of management in which crime
control and community safety are tied to narrow and short-sighted administra-
tive concerns with community safety forums becoming victims of the New
Managerialism and of the associated performance indicator culture. This could
lead to a proliferation of bureaucratic processes and the overall expansion of an
ever more opaque and impenetrable criminal justice industry. It could continue
12 Roger Matthews and John Pitts

to be preoccupied with vertical forms of accountability, producing endless

documentation, glossy brochures and the construction of audits which are
designed to provide the basis of what is euphemistically called ‘evidence-led’
policy. As Nick Tilley suggests in Chapter 3, there are serious problems with
the contention that policy and practice can simply be ‘led’ by evidence. There
is a danger that this approach could lead towards a ‘virtual war’ against crime in
which the reality of crime is reduced to a process of ‘representations’ based on
forms of creative accounting, the massaging and recycling of figures, and the
creation of forums in which those responsible for crime control and community
safety are increasingly distanced from the messy business of victimisation and
suffering (Ignatieff 2000).
The alternative option is that this form of decentralisation could open the
door to wider public participation in both the decision-making process and in
the setting of the community safety agenda. That is, local authorities could
become a conduit for the development of new styles of participation and
develop a real autonomy and independence in decision-making which is open
and responsive to public interests. The potential of a more localised response in
creating greater accountability remains largely unrealised and although the
more democratically-minded local authorities may have a genuine desire to be
responsive to the needs and views of the local communities, the lines of
accountability appear at present to be mainly horizontal with the main pressure
coming from the government, the Home Office, the Audit Commission and the
police inspectorate.

Disorder
Since the 1980s there has been a growing concern in official circles, as well as
in local communities, with the issue of disorder. This concern, however, has not
entirely been stimulated by an increase in disorder in all its forms on our city
streets. Certainly there has been an increase in the visibility of certain types of
disorder, such as graffiti, vagrants and homeless people on the streets, aggressive
begging, and more recently ‘squeegee merchants’, while a number of urban
centres in Britain have become sites of disorder at the weekends involving
predominantly young people (while other centres, in contrast, have become
desolate and neglected). In many urban centres, however, the number of public
demonstrations and riots and the level of public drunkenness has remained
fairly stable, and in comparison to twenty or thirty years ago is more contained.
Thus in explaining the growing concern with disorder we cannot identify a
simple causal sequence which sees these responses, both by the general public
and by official bodies, as a direct result of a general increase in disorder itself.
The explanation instead must lie in the ways in which these incidents are
perceived and the changing levels of toleration which is expressed by different
groups in different localities. As Lynn Hancock and Roger Matthews argue in
Chapter 4, the term toleration is a critical but largely unexplored concept in
understanding changing public attitudes towards crime, disorder and commu-
 Introduction: beyond criminology 13

nity safety, since the level and nature of public tolerance will condition which
activities members of the community are willing to put up with and those
which they are not.
The changing conceptions and responses to street prostitution in the UK
since the 1950s provides an example of how public tolerance and social sensibili-
ties have changed. Although the visibility of street prostitution decreased in
most urban centres following the implementation of the recommendations of the
Wolfenden Report (1957), a number of residents groups across the country
campaigned vigorously for the removal of prostitution from their neighbour-
hoods. This resulted in street patrols by residents, and in some localities vigilante
squads were established. The groups involved a diverse mix of local residents and
included individuals drawn from different religious and ethnic backgrounds.
Their justification for establishing citizen patrols was that the police and local
authorities were not interested in dealing with the problem. This was predomi-
nantly a ‘bottom-up’ community-based movement involving a number of
like-minded but independent groups in a number of different urban locations.
The sociological question which arises is why these groups of residents, often
living in poor transient communities – which the literature suggests should have
low levels of informal control – mobilised so effectively around this issue in an
attempt, as they saw it, to win back ‘their’ streets and neighbourhoods (Matthews
1993). Previous generations had put up with the noise and disturbance associated
with street prostitution for years, but clearly a shift occurred in terms of public
tolerance and this appeared to be bound up with changing conceptions of defen-
sible space and changing expectations about privacy, security and risk. No small
factor in this equation was the role of women living in ‘red light’ districts who
felt they were living under a curfew and that they and their children should have
the right to walk freely in public spaces without fear of harassment or intimida-
tion. The mobilisation of local residents around this issue served to increase
confidence and was a source of empowerment for many local residents groups.
Dealing with disorder, as Zygmunt Bauman (1995) has observed, can have a
cleansing and cathartic effect: it represents a form of social purification.
Because of the continuous and quasi-legal nature of prostitution, the police
were ill-equipped to deal with the issue, and although a series of ‘vice squads’
were established in the 1970s, their methods relied mainly on the mobilisation
of the criminal law which appeared increasingly heavy-handed, inequitable and
inappropriate, particularly since many of the residents came to regard female
prostitutes as victims rather than offenders (Benson and Matthews 2000). In
many instances, residents were opposed to prostitution not prostitutes.
Therefore, they looked for different, non-legal and more appropriate ways of
addressing the problem which involved a combination of social and situational
measures. In some ways the growing concern with disorder reflected both the
increasing aspirations of citizens and the mobilisation of social movements
which were interested in defending, and where possible, improving urban life.
Increasingly, disorder in its various forms provides a bridge between crime on
the one hand, and community safety on the other. Just as it is the case that
14 Roger Matthews and John Pitts

certain forms of disorder involve illegalities (such as ‘soliciting for the purposes
of prostitution’), they also involve ‘quality of life’ issues which readily fuse with
growing concerns about environmentalism, health, transport and community
safety in general.
As George Kelling (1987) has pointed out, whereas crimes are mainly events
or acts which occur in a specific time and place, disorder constitutes a condition.
It represents not so much a direct confrontation between an individual victim
and offender, but an ongoing process which affects whole neighbourhoods. The
articulation of these concerns was presented in George Kelling and James Q.
Wilson’s article, ‘Broken Windows: The Police and Neighbourhood Safety’,
published in Atlantic Monthly in 1982. The article hit a nerve, or rather a series
of nerves, and its extensive international influence and the debate which it
generated is testimony to its significance both as an attempt to explain the link
between crime disorder and community safety and to provide a practical guide
for intervention. The immediate appeal of the article was that it offered a way
out which was both plausible and potentially cost-effective. The authors
claimed that there was a clear link between low level disorder, crime and neigh-
bourhood decline. They argued that the visible evidence of graffiti and public
drunkenness indicates that an area is a vulnerable target for criminal activities.
This prompts more mobile and respectable families to move away, with the
consequence that the existing system of informal controls breaks down, crime
and incivilities proliferate and eventually the neighbourhood slips into decline.
Thus Wilson and Kelling argue that:

at the community level disorder and crime are usually inextricably linked
in a kind of developmental sequence. Social psychologists and police offi-
cers tend to agree that if a window in a building is broken and if it is left
unrepaired, all the rest of the windows will soon be broken. This is as true
in nice neighbourhoods as in run-down ones.
(Wilson and Kelling 1982: 78)

The novelty and the attractiveness of this thesis lies in its emphasis on the
added value of intervening at the most immediate and accessible level. In oppo-
sition to those forms of crime control which concentrated on serious crime,
Wilson and Kelling offered a strategy which promised significant benefits by
dealing promptly with low level disorder, incivilities and minor offences. In this
way they promised to turn declining neighbourhoods around and to salvage
those who were in danger of ‘tipping’ into decline (Hope 1999).
The ‘Broken Windows’ thesis was for many years held up as the new wisdom
of criminology. However, it also became the object of criticism from a number
of directions. These critiques generally took one of three forms. Most commen-
tators took issue with the claimed developmental sequence between disorder,
crime and urban decline. In opposition, they argued that this account captured
at best only one side of the dialectic and at worst distorted the real develop-
mental sequence which was in fact the opposite of the one proposed by Wilson
 Introduction: beyond criminology 15

and Kelling. Critics argued that it was urban decline which was more likely to
affect the housing market, family stability and the composition of neighbour-
hoods and that as these elements changed informal controls would tend to
break down and the vulnerability of affected populations would increasingly
make them more attractive targets for crime and disorder (Matthews 1992).
Second, it was suggested that the thesis did not clearly differentiate the signifi-
cance and the social meaning of different kinds of disorder. Although it was
accepted that disorder was a loosely defined category, the important differences
between say prostitution and graffiti, or between homeless people sleeping in
doorways and rowdy youths became blurred (Sampson and Raudenbush 1999).
This conceptual slippage led to a third problem, the critics claimed, which was
that Wilson and Kelling failed to distinguish between policing problems and
welfare problems. Moving the homeless on or arresting them for loitering was
seen by many as inappropriate – the problems were merely put out of sight and
it failed to address the real causes, such as the shortage of available work or
affordable accommodation. In Chapter 5, George Kelling provides a response to
these critiques and restates the need to take issues of disorder seriously while
providing some suggestions of the ways in which the ‘Broken Windows’ thesis
might be developed and implemented.
Other critics have argued that disorder can have its uses and have pointed to
the dangers of attempting to suppress every form of disorder since this can result
in sanitised cities in which street life, in its various forms, becomes conspicu-
ously absent (Sennett 1970). Alternatively, the fight against disorder can
produce ‘fortress cities’, in which the use of public space is curtailed and law
enforcement is built into the very bricks and mortar of the urban landscape. In
Los Angeles, heralded by many as the city of the future, the fight against crime
and disorder has resulted in what Mike Davis (1998) has identified as the
creation of a series of ‘social control districts’ which embody new forms of
spatial discipline. These include: abatement districts, in which the traditional
police powers over nuisance are extended to the removal of graffiti and prostitu-
tion; enhancement districts, in which those committing certain offences within a
specified radius of designated public institutions can have their penalties
increased; containment districts which are designed to quarantine potentially
endemic social problems and, more usually, certain social types; and exclusion
districts, in which designated pariah groups are excluded from public spaces or
even city limits. Engaging in urban life in cities like Los Angeles becomes a
matter of moving between a series of ‘security bubbles’, which may take the
form of malls, offices or gated communities. It is a world in which order is
organised in relation to strict spatial divisions producing new geographies of
inclusion and exclusion.

The underclass and social exclusion

A corollary of the shift from crime control to community safety is a change in
focus from individual offenders to aggregate populations and a consequent
16 Roger Matthews and John Pitts

interest in the development of an ‘underclass’ and the socially excluded.

These two terms are used in different ways to refer to populations who are
depicted as being excluded from useful or meaningful participation in main-
stream society.
The concept of the underclass was developed and widely publicised through
the work of the conservative critic Charles Murray (1994). Murray attributes
the emergence of an underclass to the failings of a welfare system which creates
dependency and indolence. He argues that not only is the present welfare
system a disincentive for people to work, but it helps to foster a sub-economy of
crime and promotes illegitimacy. This in turn sustains a ‘culture of poverty’ and
intensifies the divisions between the haves and the have-nots. Whereas Murray
locates the development of the underclass in a counter-productive welfare
system and in the consequent lack of motivation of certain types of people to
seek gainful employment and support their families, Ralph Dahrendorf (1994)
and William Julius Wilson (1987, 1994) offer an alternative, ‘structuralist’
account of the emergence of an underclass by focusing upon the processes
leading to social and economic polarisation in late modern societies (Crowther
2000). Wilson identifies the underclass as comprising the long-term unem-
ployed, the persistent poor and the denizens of the ‘ghetto’ who have lost
regular and guaranteed access to markets, especially the labour market, and
effective political representation. In his later work Wilson uses the phrase the
‘ghetto poor’ in order to distance his critique from that of Murray.
Although contemporary politics and policy in the UK has increasingly been
influenced by the underclass thesis in general, and particularly in its more
voluntaristic variant as presented by Murray, in mainland Europe the notion of
‘social exclusion’ has gained far greater currency. The idea of social exclusion
has its origins in France in the early 1980s. Upon his election as President in
1981, François Mitterrand faced nation-wide riots in the multi-racial banlieue.
The analysis of these events proffered by Mitterrand’s advisers suggested that
French society was experiencing une grande mutation characterised by the dissi-
pation of industrial society and the working-class movement, the supplanting of
collectivism by individualism and subjectivism, mass unemployment and racial
conflict (Wieviorka 1994). It was evident that the multi-racial banlieue, the site
of the 1981 riots, was the place, par excellence, where the sense of French citi-
zenship was at its weakest because its denizens were wholly or partially
‘excluded’ from participation in the key social, educational, cultural and polit-
ical institutions of French life. As a result, Mitterrand’s subsequent ‘social
prevention initiative’ attempted not only to reduce crime and ameliorate mate-
rial poverty, but also to include les exclus in the key institutions of the state and
civil society in order to strengthen social cohesion and social order and re-
establish the political legitimacy of the Republic (Donzelot and Roman 1991).
As Graham Room observes, whereas:

The notion of poverty is primarily focused upon distributional issues: the

lack of resources at the disposal of an individual or household … social
 Introduction: beyond criminology 17

exclusion focus(es) primarily on relational issues, in other words, inade-

quate social participation, lack of social integration and lack of power.
(Room 1995: 5)

Defined in this way, ‘social exclusion’ is inextricably bound up with the notion
of citizenship. Further, it was the related issues of citizenship and ‘exclusion’
which Jacques Delors, who was a close political associate of Mitterrand, intro-
duced into the vocabulary of the European Union, most notably in the Delors
Report, Growth, Competitiveness and Employment (1993). Social exclusion and
citizenship both have a central role in contemporary political discourse in the
UK, but because the voluntaristic variant of the underclass thesis has such a
firm purchase within that discourse, there is always the sense that social exclu-
sion may be self-inflicted and that discussions of citizenship accentuate the
responsibilities of young people at the expense of their civil, political and
economic rights (Pitts 2001). However, the question of citizenship, and more
particularly whether a citizen’s race or culture has a bearing upon the service
he or she might receive from the police, was forcefully posed by the
Macpherson Report (1999) on the murder of Stephen Lawrence. Marian
FitzGerald, in Chapter 6, assesses the impact of Macpherson on the policing of
black and Asian citizens and the extent to which the policies which emerged
following the publication of the Macpherson Report have contributed to their
safety.
In order to respond to what is seen as the development of a more divided
society in which the gap between the rich and poor appears to be ever
widening, the Labour government established the Social Exclusion Unit. In its
major strategy document, it states that:

Over the last generation, this has become a more divided country. While
most areas have benefited from rising living standards, the poorest neigh-
bourhoods have tended to become rundown, more prone to crime, and
more cut off from the labour market. The national picture conceals pockets
of intense deprivation where the problems of unemployment and crime are
acute and hopelessly tangled up with poor health, housing and education.
They have become no go areas for some and no exit zones for others.
(Social Exclusion Unit 1998)

In a society in which the structural and social divisions are becoming

increasingly pronounced, the ability of the Social Exclusion Unit to do more
than identify and possibly ameliorate certain pockets of poverty and deprivation
appear to be limited. The proposed interventions such as job training, the
development of ‘more flexible’ benefits, improved housing management,
measures to ‘tackle’ anti-social neighbours, youth crime and drugs, youth work,
improved local services and initiatives to get central government departments
working more closely together appear to be both overly ambitious and at the
same time inadequate.
18 Roger Matthews and John Pitts

In Chapter 7, John Pitts provides an overview and assessment of the

Labour government’s strategy for dealing with problem and excluded youth
and argues that although this strategy may strive to tackle ‘alienation’
amongst the young it is unable to refurbish the structural links between
impoverished, destabilised, neighbourhoods and local and regional labour
markets; nor will it offer the financial incentives and tax breaks to the busi-
nesses and industries which would ‘kick-start’ moribund local economies
into life.
It has also been suggested that there are conceptual problems associated with
the use of the terms ‘underclass’ and ‘social exclusion’ in that, as with the
notion of decentralisation, they suggest a simple centre–periphery relation and
play down or ignore divisions and diversity in society. The implicit suggestion in
much of this literature is that if the marginalised could be properly incorporated
into the mainstream of society the problem would be resolved. This overlooks
the fact that the so-called ‘mainstream’ is itself hugely fragmented and divided
along class, gender, age and ethnic lines (Levitas 1998; Morris 1996). Similarly,
the notions of underclass and social exclusion suggest that deprivation and
inequality are peripheral in society rather than endemic. By playing down the
potentially positive aspects of difference and diversity it fosters a moral authori-
tarianism which urges conformity with the contrived consensus of the
mainstream. In reality, the stark dichotomy of inclusion and exclusion does not
stand up to empirical investigation. Even those who live in poverty or who
experience long-term unemployment may, as William Julius Wilson points out,
endorse mainstream social values:

Our research reveals that the beliefs of inner-city residents bear little
resemblance to the blanket media reports asserting that values have
plummeted in impoverished inner-city neighbourhoods or that people in
the inner-city have an entirely different value system. What is so striking
is that despite the overwhelming joblessness and poverty, black residents
in inner-city ghetto neighbourhoods actually verbally endorse, rather
than undermine, the basic American values concerning individual
initiative.
(Wilson 1996: 179)

The term social exclusion, however, fits well with the view that social depri-
vation is multifaceted and with the suggestion that the issues are wider than
poverty alone and involve moral considerations and notions of identity. It also
fits well with the commitment to multi-agency interventions. The implicit
suggestion, however, that the problem is overwhelmingly one of distribution, or
redistribution, has been challenged by those who see the focus on the under-
class and the socially excluded as a distraction from the development of a more
diverse and tolerant society, on the one hand, and by those who argue for a
more thoroughgoing economic and social reorganisation, on the other (Bowring
2000; Young 1990).
 Introduction: beyond criminology 19

The risk society

The advent of ‘late modernity’, it has been argued, marks the abandonment of
the modernist dream, that through judicious intervention in health, crime and
the environment, the state will progressively rid the world of risk. Instead, we
confront the reality that both the natural and the social worlds are far less
amenable to control than we had once supposed, and we can predict neither
outcomes nor consequences with any degree of certainty. As Anthony Giddens
(1992) comments:

To accept risk as risk … is to acknowledge that no aspects of our activities

follow a pre-destined course, and all are open to contingent happenings. In
this sense it is quite accurate to characterise modernity, as Ulrich Beck
does, as a ‘risk society’, a phrase which refers to more than just the fact that
modern life introduces new forms of danger which humanity has to face.
Living in the risk society means living with a calculate attitude to the open
possibilities of action, positive and negative, with which as individuals, and
globally, we are confronted in a continuous way in our contemporary social
existence.
(Giddens 1992: 28)

Paradoxically, as Tim Hope observes in Chapter 8, the modernist attempt to

understand the origins and nature of crime in order to devise social policy and
rehabilitative responses which address the supposed causes of crime, is giving
way to actuarial calculation, leading to categorisations of neighbourhoods and
populations of actual and potential victims and offenders. In this way risk
assessment is increasingly seen to displace the use of personal and moral criteria
as a basis of intervention (Feeley and Simon 1992). Tim Hope also raises the
issue of distributive justice and the distribution of crime and victimisation. This
theme is further developed by Paul Wiles and Ken Pease in the final chapter.
They argue that the issue of distributive justice is a fundamental but largely
unexplored concept within criminology, since it relates both to questions about
the relation between inequalities and punishment as well as to the possibility
and desirability of achieving a more equal distribution of crime and victimisa-
tion throughout society.
The actuarial language of risk gives the impression of being objective, calcu-
lable and scientific. The reality is that the modes of risk assessment which have
been developed to date in relation to crime are not particularly precise and
consequently their predictive capacity is uncertain. Thus as tool for manage-
ment or a guide for intervention, risk classification schemes are of limited
utility. The value choices which guide decisions can, however, be effectively
concealed, while most calculations of risk conveniently allow considerable
leeway in the interpretation of the findings and the possibility of overriding
outcomes when required. Of particular significance is the way in which the
New Managerialism combines with actuarialism to redefine the meaning of
certain categories. The way, for example, in which the meaning of ‘needs’ has
20 Roger Matthews and John Pitts

become redefined in this discourse, such that it becomes virtually synonymous

with ‘risk’ is instructive. The ‘new penology’ redefines needs of prisoners as
those factors or deficiencies which are thought to be linked to an individual’s
criminal career. As Kelly Hannah-Moffat has argued in relation to risk assess-
ment procedures in women’s prisons:

[However], an unintentional byproduct emerging in the correctional logic

of the new women’s prisons is that the concept of ‘need’ shifts from a vindi-
cation of a claim for resources (the feminist view) to a calculation of
criminal potential (the risk of recidivism). Thus, correctional strategies and
programmes now ‘govern at a distance’ by regulating women through their
needs. Unlike past feminist narratives on women’s needs that stress
women’s entitlement, the Correctional Service uses the language of needs
to facilitate responsibilization. The prisoner is expected to ‘cure’ herself and
manage her own risk by satisfying her crimogenic needs.
(Hannah-Moffat 1999: 84)

Needs assessment, therefore, becomes a variant of risk assessment and serves as

regulatory mechanism rather than a basis for therapeutic or welfare interven-
tion. The potential for extending the logic of this form of needs/risk analysis
outside of the correctional setting into community safety programmes is, of
course, considerable.
The growing preoccupation with actuarial policies has also been heightened
by the discovery, and dramatisation, of new forms of risk of which we may previ-
ously have been only dimly aware. Changing public sensibilities and awareness,
a steadily expanding and ever news-hungry media, proliferating pressure groups,
and a growing body of social and criminological research have been instru-
mental in revealing a much wider range of potential hazards. Moreover, hazards
which we might once have regarded as rare and isolated aberrations increasingly
appear to be endemic. Environmental pollution, contaminated foodstuffs,
growing health problems and disease, violence in its various forms, drug use
combined with other forms of crime and disorder appear as a major threat to the
realisation of ‘the good life’ (Crimmens and Pitts 1999).

Conclusion
The shifting emphasis from crime control to community safety which has been
institutionalised in local authority-based partnerships in the UK has been
described as a ‘sea change’ and a ‘watershed’ with some justification. The formal
coupling of crime and disorder in the Crime and Disorder Act (1998) and the
incorporation of both within the wider framework of community safety has had
a number of significant consequences, for criminology and for how crime is seen
in relation to other hazards. Although the crime control industry continues to
grow, the privileged position which crime has been accorded in recent years is
coming under question. New styles of governance are emerging and the
 Introduction: beyond criminology 21

discourses through which interventions are being formulated are changing

significantly. At the same time, policing strategies for crime are gradually being
transformed.
In the course of these developments it seems probable that we will see
existing divisions between administrative and managerialist approaches on the
one side, and the more critical and sociological perspectives on the other,
becoming wider and deeper. For the growing number of students and practi-
tioners in the field, the focus of criminology will become broader but less precise
while choices of orientation will become more difficult as debates become
increasingly polarised and distanced.
The situation in which those of us who have come to identify ourselves as
criminologists find ourselves is both disturbing and exciting. It is disturbing
inasmuch that a great deal of what we have taken for granted for many years in
criminology appears unreliable and uncertain. It is exciting in that new possibil-
ities and questions are opening up. We are beginning to re-evaluate some of the
old issues. There is, however, a problem in making sense of the changes and
responding appropriately, not just because of the scale and depth of changes
taking place, but also because of the pace of change. It is increasingly difficult to
keep up with developments. The certainties of the past have become today’s
problems, and they demand solutions which conventional approaches no longer
seem able to adequately grasp.
This book is a response to these changes. It represents an attempt to come to
terms with shifting practices and changing policies, and it reflects current uncer-
tainties and future possibilities. It is self-consciously diverse and uneven. In many
of the articles there is a sense of reflection and taking stock. At the same time,
however, there is evidence of a struggle to move on and to explore new ground.

References
Abel-Smith, B. and Townsend, P. (1964) The Poor and the Poorest, London: G. Bell and
Sons.
Audit Commission (1996) Misspent Youth, London: the Audit Commission.
—— (1999) Safety in Numbers, London: the Audit Commission.
Barr, R. and Pease, K. (1990) ‘Crime Placement, Displacement and Deflection’, in M.
Tonry and N. Morris (eds) Crime and Justice: An Annual Review of Research, Chicago:
University of Chicago Press.
Bauman, Z. (1995) Life in Fragments: Essays in Postmodern Morality, Oxford: Blackwell.
Benson, C. and Matthews, R. (2000) ‘Police and Prostitution: Vice Squads in Britain’, in
R. Weitzer (ed.) Sex for Sale, New York: Routledge.
Blagg, H. and Smith, D. (1989) Crime and Social Policy, London: Longman.
Bottoms, A. (1995) ‘The Philosophy and Politics of Punishment and Sentencing’, in C.
Clarkson and R. Morgan (eds) The Politics of Sentencing Reform, Oxford: Clarendon.
—— and McWilliams, W. (1979) ‘A Non-treatment Paradigm for Probation Practice’,
British Journal of Social Work, 9: 159–202.
—— and Preston, R. (1980) The Coming Penal Crisis, Edinburgh: Scottish Academic
Press.
22 Roger Matthews and John Pitts

Bowring, F. (2000) ‘Social Exclusion: Limitations of the Debate’, Critical Social Policy,
20(3): 307–30.
Braithwaite, J. (2000) ‘The New Regulatory State and the Transformation of Crimi-
nology’, British Journal of Criminology, 40: 223–38.
Brown, M. and Pratt, J. (2000) Dangerous Offenders: Punishment and Social Order,
London: Routledge.
Bureau of Justice (2000) Expenditure and Employment Statistics, US Department of Justice.
On-line. Available: www.ojp.usdoj.gov/bjs/eande.htm
Cain, M. (1990) ‘Realist Philosophy and Standpoint Epistemologies or Feminist Crimi-
nology as a Successor Science’, in L. Gelsthorpe and A. Morris (eds) Feminist
Perspectives in Criminology, Milton Keynes: Open University Press.
Clarke, R. (1992) Situational Crime Prevention: Successful Case Studies, New York: Harrow
and Heston.
Cloward, R. and Ohlin, L. (1960) Delinquency and Opportunity, London: Routledge and
Kegan Paul.
—— and Fox-Piven, F. (1972) Regulating the Poor: The Functions of Public Welfare,
London: Tavistock Publications.
Cohen, S. (1985) Visions of Social Control, Cambridge: Polity Press.
—— (1987) ‘Taking Decentralisation Seriously’, in J. Lowman, R. Menzies and T. Palys
(eds) Transcarceration: Essays in the Sociology of Social Control, Aldershot: Gower.
Cooper, A. and Lousada, J. (2000) The Meaning of Welfare, London: Venture Press.
Crawford, A. (1998) Crime Prevention and Community Safety: Politics, Policies and Prac-
tices, London: Longman.
Crimmens, D. and Pitts, J. (1999) Positive Residential Practice: The Lessons of the 1990s,
Lyme Regis: Russell House Publishing.
Crowther, C. (2000) ‘Thinking About the “Underclass”: Towards a Political Economy of
Policing’, Theoretical Criminology, 4(2): 149–67.
Cullen, F. (1994) ‘Social Support as an Organising Concept for Criminology’, Justice
Quarterly, 11(4): 527–59.
Currie, E. (1986) Confronting Crime: An American Challenge, New York: Pantheon.
—— (1988) ‘Two Visions of Community Crime Prevention’, in T. Hope and M. Shaw
(eds) Communities and Crime Reduction, London: HMSO.
—— (1998) Crime and Punishment in America, New York: Metropolitan Books/Henry
Holt and Company.
Dahrendorf, R. (1994) ‘The Changing Quality of Citizenship’, in B. van Steenbergen
The Condition of Citizenship, London: Sage Publications.
Davis, M. (1998) Ecology of Fear: Los Angeles and the Imagination of Disaster, New York:
Henry Holt and Company.
Donzelot, J. and Roman, J. (1991) Le déplacement de la question sociale, in J. Donzelot
(ed.) Face à l’exclusion, Paris: Editions Esprit.
Ericson, R. and Carriere, K. (1994) ‘The Fragmentation of Criminology’, in D. Nelken
(ed.) The Futures of Criminology, London: Sage.
Farrell, G. and Pease, K. (1993) Once Bitten, Twice Bitten: Repeat Victimisation and its
Implications for Crime Prevention, Crime Prevention Unit Paper, no. 46, London:
HMSO.
Farrington, D. and West, D. (1993) ‘Criminal, Penal and Life Histories of Chronic
Offenders: Risk and Protective Factors and Early Identification’, Criminal Behaviour
and Mental Health, 3: 492–523.
 Introduction: beyond criminology 23

Feeley, M. and Simon, J. (1992) ‘The New Penology: Notes on the Emerging Strategy of
Corrections and its Implementation’, Criminology, 30(4): 452–74.
—— (1994) ‘Actuarial Justice: The Emerging New Criminal Law’, in D. Nelken (ed.)
The Futures of Criminology, London: Sage.
Garland, D. (1996) ‘The Limits of the Sovereign State’, British Journal of Criminology,
36(4): 445–71.
—— and Sparks, R. (2000) ‘Criminology, Social Theory and the Challenge of Our
Times’, British Journal of Criminology, 40: 189–204.
Genn, H. (1988) ‘Multiple Victimisation’, in M. Maguire and J. Pointing (eds) Victims of
Crime: A New Deal?, Milton Keynes: Open University Press.
Giddens, A. (1992) The Transformation of Intimacy, Cambridge: Polity Press
—— (1998) The Third Way: The Renewal of Social Democracy, Cambridge: Polity Press.
Graham, J. and Bowling, B. (1995) Young People and Crime, London: Home Office.
Hannah-Moffat, K. (1999) ‘Moral Agent or Actuarial Subject: Risk and Canadian
Women’s Imprisonment’, Theoretical Criminology, 3(1): 71–94.
Harvey, D. (1989) The Condition of Postmodernity, Oxford: Blackwell.
Hirschi, T. (1969) The Causes of Delinquency, California: University of California Press.
Home Office (2000) Review of Crime Statistics: A Discussion Document, London: HMSO.
Hope, T. (1999) ‘Crime, Community and Inequality: The Rise of Disorder in British
Public Policy’, inaugural professorial lecture, University of Keele, October.
Hughes, G. (1998) Understanding Crime Prevention: Social Control, Risk and Late Moder-
nity, Buckingham: Open University Press.
Ignatieff, M. (2000) Virtual War: Kosovo and Beyond, London: Chatto and Windus.
Johnston, L. (2000) Policing Britain: Risk, Security and Governance, Harlow: Longman.
Kelling, G. (1987) ‘Acquiring a Taste for Order: Community and Police’, Crime and
Delinquency, 33(1): 90–103.
Kinsey, R., Lea, J. and Young, J. (1986) Losing the Fight Against Crime, Oxford: Blackwell.
Krisberg, B. and Austin, J. F. (1993) Reinventing Juvenile Justice, London: Sage Publishing.
Lea, J. (1998) ‘Criminology and Postmodernity’, in P. Walton and J. Young (eds) The
New Criminology Revisited, London: Macmillan.
Leonard, E. (1982) Women, Crime and Society: A Critique of Criminological Theory,
London: Longman.
Levitas, R. (1998) The Inclusive Society? Social Exclusion and New Labour, London:
Macmillan.
Lewis, O. (1966) La Vida: A Puerto Rican Family in the Culture of Poverty: San Juan and
New York, New York: Random House.
Loveday, B. (1999) ‘The Impact of Performance Culture on Criminal Justice Agencies in
England and Wales’, International Journal of the Sociology of Law, 27: 351–77.
McMahon, M. (1990) ‘Net Widening: Vagaries in the Use of a Concept’, British Journal
of Criminology, 30(2): 121–50.
Maguire, M. and Pointing, J. (1988) (eds) Victims of Crime: A New Deal?, Milton
Keynes: Open University Press.
Marris, P. and Rein, M. (1967) Dilemmas of Social Reform: Poverty and Community Action
in the United States, New York: Atherton Press.
Martinson, R. (1974) ‘What Works? Questions About Prison Reform’, The Public
Interest, Spring: 22–54.
Matthews, R. (1989) Privatising Criminal Justice, London: Sage.
—— (1992) ‘Replacing Broken Windows’, in R. Matthews and J. Young (eds) Issues in
Realist Criminology, London: Sage.
24 Roger Matthews and John Pitts

—— (1993) Kerb-crawling, Prostitution and Multi-agency Policing, Crime Prevention Unit

Paper, no. 43, London: HMSO.
—— and Young, J. (1992) ‘Reflections on Realism’, in J. Young and R. Matthews (eds)
Rethinking Criminology: The Realist Debate, London: Sage.
Mawby, R. and Walklate, S. (1994) Critical Victimology, London: Sage.
Morris, A., Giller, H., Swzed, E. and Geach, H. (1980) Justice for Children, London:
Macmillan.
Morris, L. (1996) ‘Dangerous Classes: Neglected Aspects of the Underclass Debate’, in E.
Mingione et al. (eds) Urban Poverty and the Underclass, Oxford: Blackwell.
Moynihan, D. P. (1969) Maximum Feasible Misunderstanding: Community Action in the
War on Poverty, New York: Free Press.
Muncie, J. (2000) ‘Decriminalising Criminology’, in G. Lewis, S. Gewitz and J. Clarke
(eds) Rethinking Social Policy, London: Sage.
Murray, C. (1994) Underclass: The Crisis Deepens, London: Institute of Economic Affairs.
Pitts, J. (2001) The New Politics of Juvenile Crime: Discipline or Solidarity, Basingstoke:
Macmillan.
Preston, R. H. (1980) ‘Social Theology and Penal Theory and Practice: The Collapse of
the Rehabilitative Ideal and the Search for an Alternative’, in A. E. Bottoms and R.
H. Preston (eds) The Coming Penal Crisis, Edinburgh: Scottish Academic Press.
Room, G. (1995) (ed.) Beyond the Threshold: The Measurement and Analysis of Social
Exclusion, Bristol: Polity Press.
Rosenbaum, D. (1986) Community Crime Prevention: Does it Work?, Beverley Hills: Sage.
Sampson, R. and Raudenbush, S. (1999) ‘Systematic Social Observation of Public
Spaces: A New Look at Disorder in Urban Neighborhoods’, American Journal of Soci-
ology, 105(3): 603–51.
Schwendinger, H. and Schwendinger, J. (1975) ‘Defenders of Order or Guardians of
Human Rights’, in I. Taylor, P. Walton and J. Young (eds) Critical Criminology,
London: Routledge.
Sennett, R. (1970) The Uses of Disorder: Personal Identity and City Life, New York:
Vintage.
Shearing, C. (1989) ‘Decriminalising Criminology’, Canadian Journal of Criminology,
31(2): 169–78.
Sibley, D. (1995) Geographies of Exclusion, London: Routledge.
Smart, C. (1990) ‘Feminist Approaches to Criminology or Postmodern Woman Meets
Atavistic Man’, in L. Gelsthorpe and A. Morris (eds) Feminist Perspectives in Crimi-
nology, Milton Keynes: Open University Press.
Social Exclusion Unit (1998) Bringing Britain Together: A National Strategy for Neighbour-
hood Renewal, Cm. 4045, London: HMSO.
Sparks, R. (1997) ‘Recent Social Theory and the Study of Crime and Punishment’, in M.
Maguire, R. Morgan and R. Reiner (eds) The Oxford Handbook of Criminology,
Oxford: Oxford University Press.
Sparks, R. F., Genn, H. and Dodd, D. (1977) Surveying Victims, Chichester: John Wiley.
Thorpe, D., Smith, D., Green, C. and Paley, J. (1980) Out of Care, London: Allen and
Unwin.
Utting, W., Bright, J. and Hendrickson, B. (1993) Crime and the Family, Improving Child
Rearing and Preventing Delinquency, Family Policy Studies Centre, Paper no.16, London.
Wieviorka, M. (1994) ‘Racism in Europe: Unity and Diversity’, in A. Ratsani and S.
Westwood (eds) Racism, Modernity, Identity on the Western Front, Cambridge: Polity
Press.
 Introduction: beyond criminology 25

Wilson, J. and Kelling, G. (1982) ‘Broken Windows: The Police and Neighbourhood
Safety’, Atlantic Monthly (March): 29–38.
Wilson, J. Q. (1975) Thinking About Crime, New York: Basic Books.
—— and Herrnstein, R. (1985) Crime and Human Nature, New York: Simon and
Schuster.
Wilson, W. J. (1987) The Truly Disadvantaged: The Inner-city, the Underclass and Public
Policy, Chicago: University of Chicago Press.
—— (1994) Citizenship and the Inner-city Ghetto Poor, in B. van Steenbergen The Condi-
tion of Citizenship, London: Sage Publications.
—— (1996) When Work Disappears, New York: Knopf.
Young, I. (1990) Justice and the Politics of Difference, New Jersey: Princeton University
Press.
Young, J. (1986) ‘The Failure of Criminology: The Need for Radical Realism’, in R.
Matthews and J. Young (eds) Confronting Crime, London: Sage.
—— (1999) The Exclusive Society, London: Sage.
1 Identity, community and social
exclusion
Jock Young

Introduction
For the majority of social commentators, community safety programmes to
control crime and anti-social behaviour are an exercise in the obvious.
‘Community’ is a wholesome, homogeneous entity waiting to be mobilised,
‘safety’ is a risk free end state – the self-evident goal of public desire – whilst
‘crime’ and ‘anti-social behaviour’ are entities which are readily recognisable by
any decent citizen. Indeed, the trio – community, safety and crime – become
elided together, each term defining each other. Thus the intense, socially rich
interacting community is seen to be the very antithesis of crime and is indeed
the place and source of all safety.
Nowadays, as Adam Crawford puts it, ‘the attraction of the notion of
“community” unites and transcends the established … political parties’ (1997:
45). He goes on to write that ‘community … is cleansed of any negative or crim-
inogenic connotations and endowed with a simplistic and naïve purity and
virtue’ (ibid.: 153). Thus problems of crime and disorder are ascribed to a malaise
of community and their solution is seen as the regeneration of community. Such
a perspective can readily be seen as an aspect of modernity, in particular the
social engineering characteristic of the post-war period. This was a period
wherein slums were cleared, towns were planned and ‘rational’ communities
were constructed – the apotheosis being the ‘new town’ developments, the more
mundane and commonplace being the modernist tower blocks interspersed with
family maisonettes and bleak gardens characteristic of the post-war city.
It has been the role of critical criminology to stand outside such conventional
wisdoms and to problematise each of these seeming certainties. Community
becomes seen as a reality lacking coherence, a rhetoric invoked to ensure the
domination of one group over another, a key instrument of social exclusion.
Safety becomes a curtain for fear, the miscalculation of risks, the denial of
excitement and the excuse for xenophobia towards outgroups. Finally, crime and
anti-social behaviour are social constructs which are bestowed by the powerful
on disapproved behaviour but with no essential reality or core. In fact, their
definitions are frequently contested within localities and adherence to criminal
values can even be a dominant aspect of community.1
 Identity, community and social exclusion 27

My concern in this chapter is not to reiterate these doubts but rather to

point to the fashion by which the massive changes which have occurred in the
last third of the twentieth century, in the basic building blocks of society –
employment, the family and community itself – have transformed and magni-
fied such doubts. For the shift from the modernity of the post-war period with
its high (male) employment, intense communities and stable families, to the
conditions of late modernity has been associated with a remarkable problemati-
sation of each area.

The coordinates of order: class and identity in the late

twentieth century
First, before mapping the terrain of late modernity, I wish to briefly establish
coordinates. Nancy Fraser, in her influential Justice Interruptus: Critical
Reflections on the Post-Socialist Condition (1997), outlines two types of politics:
those centring around distributive justice and those centring around the justice
of recognition – that is class politics and identity politics.2 She points to the rise
in prominence of the latter – a phenomenon which I will attempt to explain
later in this chapter. However, what concerns me here is to develop and explain
this distinction as a basis for an analysis of social order and political legitimacy.
In my book The Exclusive Society (1999), I point to the two fundamental prob-
lems in a liberal democracy. First is the need to distribute rewards fairly so as to
encourage commitment to work within the division of labour. Second is the
need to encourage respect between individuals and groups so that the self-
seeking individualism characteristic of a competitive society does not lead to a
situation of war of all against all. Individuals must experience their rewards as
fair and just and they must feel valued and respected.
I now wish to develop this distinction between the sphere of distribution and
that of recognition. Central to distributive justice is the notion of fairness of
reward, and in modern capitalist societies this entails a meritocracy in which
merit is matched to reward. Recognition involves the notion of respect and
status allocated to all, but if we stretch the concept a little further, it also
involves the notion of the level of esteem or social status being allocated justly.
Indeed, both the discourses of distributive justice and recognition have the
notion of a basic equality (all must receive a base level of reward as part of being
citizens), but on top of this rather than a general equality of outcome: a hier-
archy of reward and recognition dependent on the individual’s achievements.3
What terms are we to use when distributive justice or recognition is wanting?
When material reward is unjustly allocated we commonly use the term ‘relative
deprivation’, when recognition is denied someone we call this ‘ontological inse-
curity’. But let us explore these further. We can talk of two aspects of unfairness
– deprivation and insecurity, and two dimensions – economic and ontological.
This gives us four bases of disaffection (see Figure 1.1).
How does this help inform us as to the genesis of crime and punishment? First, a
major cause of crime lies in deprivation. That is, very frequently, the combination
28 Jock Young

Figure 1.1 The dimensions of disaffection

of feeling relatively deprived economically (which causes discontent) and

misrecognised socially and politically (which causes disaffection). The classic
combination is to be marginalised economically and treated as a second-rate
citizen on the street by the police. Second, a common argument is that widespread
economic and ontological insecurity in the population engenders a punitive
response to crime and deviancy (see for example Luttwak 1995; Young 2000).
As we shall see in the process of the transition from modernity to late
modernity, powerful currents shake the social structure transforming the nature
of relative deprivation, causing new modes of misrecognition and exclusion,
whilst at the same time being accompanied by widespread economic and onto-
logical insecurity. The purchase of each of these currents impacts differentially
throughout the social structure by each of the prime social axes of class, age,
ethnicity and gender.

The journey into late modernity and mapping the terrain

How can long-term purposes be pursued in a short-term society? How can

durable social relations be sustained? How can a human being develop a
narrative of identity and life history in a society composed of episodes and
fragments? The conditions of the new economy feed instead on experience
which drifts in time, from place to place, from job to job.
(Sennett 1998: 26–7)

If one wishes to travel anywhere one must know what terrain one has to travel
over, what means of travel are available and last, and by no means least, where
one wants to go. In reality over the last twenty years the terrain, the structure of
society, has radically changed, effective means of intervening in the social world
have profoundly altered and, most important, the metropolis of possibility and
desire in which we find ourselves has altered beyond recognition.
The last third of the twentieth century witnessed a remarkable transforma-
tion in the lives of citizens living in advanced industrial societies. The Golden
Age of the post-war settlement, with high employment, stable family structures
and consensual values underpinned by the safety net of the welfare state, has
been replaced by a world of structural unemployment, economic precariousness,
 Identity, community and social exclusion 29

a systematic cutting of welfare provisions and the growing instability of family

life and interpersonal relations. And where there once was a consensus of
value, there is now burgeoning pluralism and individualism (see Hobsbawm
1994). A world of material and ontological security from cradle to grave has
been replaced by precariousness and uncertainty, and where social commenta-
tors of the 1950s and 1960s berated the complacency of a comfortable ‘never
had it so good’ generation, today they talk of a ‘risk society’ where social
change becomes the central dynamo of existence and where anything might
happen. As Anthony Giddens puts it: ‘to live in the world produced by high
modernity has the feeling of riding a juggernaut’ (1991: 28; see also Beck 1992;
Berman 1983).
Such a change has been brought about by market forces which have system-
atically transformed both the sphere of production and consumption. This shift
from Fordism to post-Fordism involves the unravelling of the world of work
where the primary labour market of secure employment and ‘safe’ careers
shrinks and the secondary labour market of short-term contracts, flexibility and
insecurity increases, as does the growth of an underclass of the structurally
unemployed. It results, in Will Hutton’s catchphrase, in a ‘40:30:30 society’
(1995) – 40 per cent of the population are in tenured secure employment, 30
per cent in insecure employment, and 30 per cent are marginalised, idle or
working for poverty wages.
Second, the world of leisure is transformed from one of mass consumption to
one where choice and preference is elevated to a major ideal, and where the
constant stress on immediacy, hedonism and self-actualisation has had a
profound effect on late modern sensibilities (see Campbell 1987; Featherstone
1985). These changes, both in work and in leisure, characteristic of the late
modern period, generate a situation of widespread relative deprivation and
heightened individualism. Market forces generate a more unequal and less meri-
tocratic society and encourage an ethos of every person for him- or herself.
Together, these create a combination which is severely criminogenic. Such a
process is combined with a decline in the forces of informal social control, as
communities are disintegrated by social mobility and left to decay as capital
finds more profitable areas to invest and develop. At the same time, families are
stressed and fragmented by the decline in community systems of support, the
reduction of state support and the more intense pressures of work (see Currie
1997; Wilson 1996). Thus, as the pressures which lead to crime increase, the
forces which attempt to control it decrease.
The journey into late modernity involves both a change in perceptions of
the fairness of distributive justice and in the security of identity. There is a shift
in relative deprivation from being a comparison between groups (what
Runciman [1966] calls ‘fraternal’ relative deprivation) to that which is between
individuals (what Runciman terms ‘egoistic’ relative deprivation). The likely
effect on crime is, I would suggest, to move from a pattern of committing
crimes outside of one’s neighbourhood onto other richer people to committing
crimes in an internecine way within one’s neighbourhood. In other words, the
30 Jock Young

frustrations generated by relative deprivation become focused inside the

‘community’ rather than, as formerly, projected out of it.
Thus Roger Hood (Hood and Jones 1999) and his associates, in their fasci-
nating oral history of three generations in London’s East End, find a sharp
difference between the 1980s generation, who left school in the late 1970s and
1980s, and those of the 1930s and 1950s. Again and again, the older respon-
dents talk of their lack of fear of crime in the earlier periods, and make the
distinction between the tolerated crime committed against shops, warehouses,
rich outsiders and the anathema of crime against one’s family, friends or neigh-
bours. For the 1980s generation, all of this had changed:

crime had formerly been seen as an activity on the margins of everyday

relationships, something that a few segregated members of the community
did to outsiders: the bosses and impersonal companies. It was now
perceived to be increasingly an internal threat. Indeed some young men, as
well as women described it as pervasive within their neighbourhoods.
(Hood and Jones 1999: 154–5)

But it is also in the realm of identity that relative deprivation is increased

and transformed. For here, on one side, you have raised expectations: the spin-
off of the consumer society is the market in lifestyles. On the other side, both in
work and in leisure, there has been a disembeddedness. That is, identity is no
longer secure; it is fragmentary and transitional – all of which is underscored by
a culture of reflexivity which no longer takes the world for granted. The iden-
tity crisis permeates our society. As the security of the lifelong job and the
comfort of stability in marriage and relationships fade, as movement makes
community a phantasmagoria where each unit of the structure stays in place but
each individual occupant regularly moves, where the structure itself expands
and transforms and where the habit of reflexivity itself makes choice part of
everyday life and problematises the taken for granted – all of these things call
into question the notion of a fixed, solid sense of self. Essentialism offers a
panacea for this sense of disembeddedness.

The identity crisis and the attractions of essentialism

In The Exclusive Society (1999) I discuss the attractions of essentialism to the
ontologically insecure and denigrated. To believe that one’s culture, ‘race’,
gender or community has a fixed essence which is valorised and unchanging is,
of course, by its very nature the answer to a feeling that the human condition is
one of shifting sands, and that the social order is feckless and arbitrary. To
successfully essentialise oneself it is of great purchase to negatively essentialise
others. That is to ascribe to the other either features which lack one’s own values
(and solidity) or which are an inversion of one’s own cherished beliefs about
one’s self. To seek identity for oneself, in an essentialist fashion, inevitably
involves denying or denigrating the identity of others.
 Identity, community and social exclusion 31

Crime and its control is a prime site for essentialisation. Who, by definition,
could be a better candidate for such a negative ‘othering’ than the criminal and
the culture that he or she is seen to live in? Thus the criminal underclass,
replete with single mothers living in slum estates or ghettos, drug addicts
committing crime to maintain their habit and illegal immigrants who commit
crime to deceitfully enter the country (and continue their lives of crime in
order to maintain themselves), have become the three major foci of emerging
discourses around law and order of the last third of the twentieth century. These
types can be summed up as the welfare ‘scrounger’, the ‘junkie’ and the ‘illegal’.
This triptych of deviancy, each picture reflecting each other in a late modern
portrait of degeneracy and despair, comes to dominate public discussion of social
problems. As the discourse develops, their ontologies become distinct and
different from ‘normal’ people, their social norms absent or aberrant, their
natures frequently racialised and rendered inferior. Crime, a product of our
society, becomes separated from the social structure: it is viewed as a result of
distinct aetiologies, it embodies differing values, it emanates from distinct and
feared areas of the city. It is these areas that are contrasted with the organic
community where social trust and harmony are seen to reside.

The organic community

The community of the post-war period was characterised by a sense of perma-
nence and solidity. Placed there by the needs of capital around large-scale
manufacturing industry or labour fixed to land for centuries, it involved the
following characteristics:

• intergenerationality;
• an embeddedness of the individual in locality;
• intense face-to-face interaction;
• much direct information with regard to each other;
• high level of informal social control;
• provision of a localised sense of identity.

The obverse of the organic community is the anomic community, the locality
without norms and the inevitable consequence of this is seen to be the prolifer-
ation of crime and anti-social behaviour. This formulation is wrong on two
scores. First, there is no direct one-to-one relationship between strong commu-
nity and crime. Second, even slum estates have strong social networks. Thus
the organic community could easily support criminal values. As Adam
Crawford nicely puts it:

The logic behind this association between the lack of ‘organized’ commu-
nity and crime is that, conversely, more community equals less crime. …
This benevolent understanding of community is highly misleading. In some
instances ‘community’, i.e. its communal normative values, itself may be
32 Jock Young

the source of criminogenic tendencies. Recent British research into crim-

inal subcultures has reiterated the long established criminological truism
that the collective values of a community may serve to stimulate and
sustain criminality.
(Crawford 1997: 153)

The recent research of Sandra Walklate and Karen Evans (1999) in Salford
corroborates Crawford’s comments.4 But the important conclusion from this is
that although intense community does not necessarily inoculate against crime,
it does exert considerable levels of social control – which may influence law-
abiding or criminal behaviour.
Conversely, the image of the sink estate, the inner-city ghetto or the satel-
lite slum is often seen as in a state of anomie. The immune system of informal
control is down and gross social pathology abounds. But even a brief visit to
one of these estates quickly dissipates such a notion. Mothers congregate
around the local school at delivery and collecting times, they discuss the
teachers, their kids, and this and that. Older children hang out at the entrance
to the tower blocks, or along the road by the shops, men meet in the pubs and
betting shops. Networks can be seen everywhere; divided, of course, by all the
axes of age, gender, ethnicity and economic situation, but certainly no lack of
an informal system or anomie in a sense of normlessness, or ‘Broken Windows’
because no one knows each other (Wilson and Kelling 1982). Ironically,
because of structural unemployment, poverty and both geographical and social
immobility, the stigmatised estates of the so-called underclass allow greater
social interaction than on middle-class dormitory estates on the edge of cities.
Here, there is usually a quick turnover of residents, dual time-consuming
careers of husband and wife, the bussing of children to relatively distant schools
and little time except at the weekend to interact with anyone – even with one’s
own family.
The work of Janet Foster has helped to dispel such images of the normless
community. In her study of the London housing estate ‘Riverside’, she notes:

The differing types and composition of networks on Riverside aptly demon-

strate that different forms of neighbourliness can occur in the same setting
simultaneously among different groups of people. These networks played an
important role in an environment where Asian, white, black and
Chinese/Vietnamese households alike lived with an underlying suspicion
about their neighbours and expressed concerns about their safety.
The very existence of these networks suggests that popular perceptions
of council housing – and the very poorest and most difficult-to-let estates in
particular – as alienating environments in which there is little tenant inter-
action and support, are too simplistic. Instead of condemning such places as
‘dreadful enclosures’ (Damer 1974) we should look more closely at the
patterns of interaction between tenants in these contexts. We need to
understand more about how they are characterised, in what ways they influ-
 Identity, community and social exclusion 33

ence tenants’ perceptions of the estates on which they live, and how
different individuals and groups, who have had little or no choice in their
housing allocation, manage to co-exist.
(Foster 1997: 126)

Crime and community: the one-to-one fallacy

The argument that there is no one-to-one relationship between strong commu-
nity and little crime often teeters on the edge of suggesting that there is no
relationship. This crucially mistakes the situation and it is worth turning for a
moment to fundamentals. Crime occurs when there is motivation to commit
crime (because of structural reasons such as those which engender relative
deprivation) and where there are insufficient countervailing norms (chiefly
informal, although enforced, legal norms are a vital backup). The major site of
such informal norms is the community. Yet, of course, there are communities
that have norms which are supportive or at least tolerant of crime and thus do
not act in such a controlling fashion. Thus crime will occur when you have
motivation and no community or where you have motivation and some level of
support for criminal activity. Crime will be much less prevalent where there is
no motivation even though there might be a lack of community or where there
is strongly enforced communal norms against crime. In the Barbican in London
or in the tower blocks of the Upper West Side of Manhattan, there may be little
concept of community, yet no conventional crime may occur. In contrast, in the
past, many poor communities were carriers of strong norms which stressed
honesty and respectability, and this was underscored by intensive day-to-day,
face-to-face interaction and underwritten by the disciplines of locally-based
work and the fear of losing one’s job. Hence crime, despite severe deprivation,
tended to be low in these communities.
If, then, we are to make anything of the finding that the organic community
sometimes supported crime and that conversely areas without a hint of
Gemeinschaft, often have low crime rates – it is merely to register that there is
no logical necessity of countering crime in this time-honoured fashion. The
communitarian notion that the most effective means of tackling crime is to
reconstitute the community (see Etzioni 1997) is, therefore, not only nostalgic
(in the sense of a politics which is backward looking and difficult to achieve)
but unnecessary. We know that communities of lightly engaged strangers can
have low crime rates and, as we shall see, we know also that the phrase ‘lightly
engaged strangers’ does not imply an atomistic and anomic society. What
demands our attention is how the territory of community has changed and how
we can find new and appropriate ways of fostering civility.
In late modernity, we face severe problems: the organic community is in
decline, whilst the rise in change in the nature of relative deprivation gives
greater motive to commit crime of a more internecine nature. Thus both the
rise in motivation and the decline in informal control give rise to higher crime
rates of a type which further fragments the already weakened community. The
34 Jock Young

desire to ‘maintain order amongst lightly engaged strangers’ has, therefore, an

urgency and this is further extended by the progressive demands of the new
social movements to improve upon levels of civility in the areas of gender and
ethnic relations and between diverse sexual, religious, and other subcultures
which have flourished in late modernity.

The search for community

One of the unintended consequences of modern capitalism is that it has

strengthened the value of place, aroused a longing for community. All the
emotional conditions we have explored in the workplace animate that
desire: the uncertainties of flexibility; the absence of deeply rooted trust
and commitment; the superficiality of teamwork; most of all, the spectre of
failing to make something of oneself in the world, to ‘get a life’ through
one’s work. All these conditions impel people to look for some other scene
of attachment and depth.
Today, in the new regime of time, that usage ‘we’ has become an act of
self-protection. The desire for community is defensive, often expressed as a
rejection of immigrants or other outsiders – the most important communal
architecture being the walls against a hostile economic order. To be sure, it
is almost a universal law that ‘we’ can be used as a defense against confu-
sion and dislocation. Current politics based on this desire for refuge takes
aim more at the weak, those who travel the circuits of the global labor
market, rather than at the strong, those institutions which set poor workers
in motion or make use of their relative deprivation.
(Sennett 1998: 138)

Here, Sennett notes that modern capitalism drives people to seek identity in
community. But Hobsbawm counters that it is precisely the self-same late
modern capitalism which has destroyed community and rendered destitute the
bank of social trust that underwrote it. The local community becomes increas-
ingly more invoked as a place of identity and moves to become a major part of
the rhetoric of political mobilisation just at the time that it is transforming. In a
well-known passage, Hobsbawm writes:

We have been living – we are living – through a gigantic ‘cultural revolu-

tion’ an extraordinary dissolution of traditional norms, textures and values,
which left so many inhabitants of the developed world orphaned and
bereft. … Never was the word ‘community’ used more indiscriminately and
emptily than in the decades when communities in the sociological sense
become hard to find in real life. Men and women look for groups to which
they can belong, certainly and forever, in a world in which all else is
moving and shifting, in which nothing else is certain.
(Hobsbawm 1996: 40)
 Identity, community and social exclusion 35

The organic community is, however, in decline, affected by both the globalisa-
tion of the economy and of culture. Manufacturing industries shrink and in
many instances disappear leaving areas bereft of work, service industries prolif-
erate often with small sizes, commuting increases to and from work, local
cultures become less self-contained and more penetrated by the global. They
become, in Giddens’ (1990) evocative term, ‘phantasmagoric’, constituted by
the ghostly presence of distant influences.
I argue that this paradox of a search for identity in community when organic
community is failing, and whose failure to provide tradition and embeddedness
is a core reason for the search for identity, is not nearly as cataclysmic as either
Sennett or Hobsbawm maintain. In part, this is because they cannot envisage
the notions of association and trust outside of the image of the face-to-face,
organic community.

The privileging of community

Iris Marian Young has been one of the most ardent advocates of urban life as an
ideal and, at the same time, an effective critic of the privileging of the organic
community:

Theorists of community privilege face-to-face relations because they

conceive them as immediate. Immediacy is better than mediation because
immediate relations have the purity and security longed for in the
Rousseauist dream: we are transparent to one another, purely copresent in
the same time and space, close enough to touch, and nothing comes
between us to obstruct our vision of one another.
This ideal of the immediate copresence of subjects, however, is a meta-
physical illusion. Even a face-to-face relation between two people is
mediated by voice and gesture, spacing and temporality. As soon as a third
person enters the interaction the possibility arises of the relation between
the first two being mediated through the third, and so on. The mediation of
relations among persons by the speech and actions of other persons is a
fundamental condition of sociality. The richness, creativity, diversity, and
potential of a society expand with growth in the scope and means of its
media, linking persons across time and distance. The greater the time and
distance, however, the greater the number of persons who stand between
other persons.
(Young 1990: 233)

Young has no doubt about the oppressive nature of many relationships in

modern urban societies and the real dangers of the city nor that the closest rela-
tionships are those of immediate intimates. But in its anonymity, the city allows
deviance, freedom to develop (which I discuss in more detail below), its differ-
ence offers a frisson of excitement and entertainment, its access to such a huge
bulk of people via the mediation of telephone and mass transit allows for the
36 Jock Young

creation of vibrant new communities of difference. Let us first of all examine

what changes have occurred in the late modern community.

The deterritorialisation of community

The use of the term ‘community’ in crime prevention discourse commonly

evokes images of localism and ‘neighbourhood’; of territorially bounded
communities in which it is assumed members also share other dense
cultural and social bonds. But in ‘modern’ Western societies, high and
increasing residential and spatial mobility among large sections of the
population; high levels of functional specialisation mirrored in the spatial
differentiation of social life; and sophisticated mass communications and
transport systems, mean that the centrality of geography and the embed-
dedness of existing social relations give way to the importance of other,
‘deterritorialised’ communities in the lives of many citizens. These are
communities of interest and function related to work, leisure, education,
taste, and so on, which increasingly owe little to geographical locale or resi-
dence for most citizens. These provide individuals with plural communities
and identities, in and through which they live their lives. Crime preven-
tion needs to be considered in relation to these ‘deterritorialised’, as well as
the more familiar residential, communities.
(Hogg and Brown 1998: 6)

Writers in cultural studies note the fashion in which the late modern community has
lost its mooring in the locale – in the coincidence of the social and the spatial. Thus
Mike Tomlinson (1999) talks of its ‘deterritorialisation’, whilst John Thompson
(1995) refers to the notion of ‘despatialised commonality’ and, perhaps a little more
elegantly, Joshua Meyrowitz (1989) talks of ‘the generalised elsewhere’. Important
here is the way in which people through the various media can share experiences
and identity despite the separation of physical distance. This is not to deny locality –
people after all must live somewhere – but it is to point to the diminution and trans-
formation of the local community and the rise of the virtual community.
John Thompson (1995: 85) usefully classifies social interaction into three
groups:

• Face-to-face interaction: the dialogical basis of the traditional community.

• Mediated interaction: which is two-way and dialogical, like face-to-face
interaction, but occurs over space and time by telephone and e-mail.
• Mediated quasi-interaction: the conventional mass media which is mono-
logical, yet where there is audience selection, interpretation and
interaction.

What is new are the developments in the latter two groups. Portable telephone
companies are presently the fastest growing firms and e-mail communication
 Identity, community and social exclusion 37

has become part of everyday life for many people. But even before this, the old-
fashioned landline telephone has had a major impact on people’s lives. Thus
Barry Wellman and other social network analysts have long pointed to the way
in which social technology has liberated people from dependence on spatial
locality. He argues that it makes more sense to perceive of personal communities
and networks rather than communities of neighbourhood (Wellman 1982).
Indeed, as Meyrowitz (1989) wryly comments: ‘access to nonlocal people is
now, via the telephone, often faster and simpler than access to physical neigh-
bors’. Furthermore, within mass media the rise of multi-media and the vast
expansion of choice in radio and television allow the development of niche
audiences and subcultures (see McRobbie and Thornton 1995). Indeed the
mass media take up a surprisingly larger and larger proportion of people’s lives.
In England and Wales, for example, watching television and listening to the
radio constitute 36 per cent of the waking life of the average citizen (figures
from Social Trends 2000).
Thus on one hand the local is penetrated by the global in terms of distant
events, consumer choices and values (such as those emphasising lifestyle choice,
feminism, meritocracy, etc.), and on the other hand, virtual communities
develop on the back of the local which incorporate images, reference groups,
favoured characters and celebrities from a global repertoire and involve both
mass media and mediated interaction.

The community in late modern times

Tim Hope’s (1995) excellent classification of images of community and crime
prevention practices from community as disorganised (in the 1930s), to
community as disadvantaged (in the 1960s) to community as frightened (in the
late 1970s and 1980s) requires further reappraisal by the end of the twentieth
century and the beginning of the twenty-first century. For all of these previous
conceptions perceived a fixed entity which is variously disorganised, disadvan-
taged or frightened and consequently a thing which can be regenerated,
redistributed or mobilised as strategies to regenerate itself and tackle crime.
And, even more recently, the Social Exclusion Unit (1999a) carries the further
image of the excluded community which needs to be included into citizenship
in order to tackle crime and disorders. All of these images encompass facets of
the truth but need to be drastically reviewed in the light of the late modern
community. For the single entity has long gone and any notion of fixity has
disappeared – there is no reified community out there to mobilise or repair, no
fixed thing in need of incorporation. Nor can this situation be solved by
adopting the conventional language of multiculturalism and communitarianism.
Namely, that there is now a series of communities to reinvigorate and galvanise.
That is to replace the fallacy of a single fixed entity with the chimera of the
multiple community. None of this fully captures the fluidity and plurality of late
modernity.
Let us delineate the basic features of the late modern community:
38 Jock Young

1 Difference: it is pluralistic, not just in terms of ethnicity but in terms of age,

gender and class.
2 Fragmentation, cross-cutting and hybridisation: such a pluralism by combination
of ethnicity, age, gender and class offers to create on the one hand fragmen-
tation, and on the other hand, other cross-cutting alliances, for example
gender across ethnicity and class. Furthermore, such subcultures bricolage
from one another, creating hybrids of cross-over and reinterpretation.
3 Intensity: it is pluralistic in terms of intensity. The same locale can contain
high intensity, disorganised and atomised groups. For example, around the
school environment, single mothers can create intense coherent subcultures,
at the same time as unemployed men can be atomistic and withdrawn.
4 Transience: such subcultures change over time in composition, intensity and
coherence. Individual biographies are experienced as shifts backwards and
forwards, from a sense of embeddedness to disembeddedness.
5 Mediated: it is highly mediated, the global penetrates the local, creating a
series of virtual communities some of which are territorialised and rooted in
the locale, and others which are considerably deterritorialised.
6 Actuarial: relationships are wary and calculative because of low information
about a large proportion of the ‘community’.
7 Internecine conflict: wealth and status is perceived as distributed in a chaotic
fashion with no clear rationale or fairness. Relative deprivation, both mate-
rially and in terms of status, is widespread.
8 Reinvention: the history of the ‘community/communities’ is constantly rein-
vented and the boundaries redrawn and redrafted.

This late modern mixture of local and virtual community can clearly enhance
people’s ability to realise identity, yet it is also a prime site of misrecognition,
and a threat to one’s status and identity.

Freedom in the city

We have emerged from a world of neighbours and entered what has increas-
ingly become one of strangers. Here we have the old theme in social
science of a shift from community (crudely, the familiar, interpersonal and
village-centred life of pre-industrialism) to associations which involve the
mixing of people unknown to one another save in specific ways such as bus
conductor, shop assistant, and newsvendor (crudely, the urban-oriented
way of life of the modern). Ever since at least Simmel we have appreciated
how disorienting and also often liberating the transfer from closed commu-
nity to a world of strangers can be. The city may fragment and
depersonalise, but in doing so it can also release one from the strictures of
village life. With the shift towards town life comes about a decline in
personal observation by neighbours and, accompanying this, a weakening
of the power of community controls that are exercised on an interpersonal
 Identity, community and social exclusion 39

basis. Entering urban-industrial life from a country existence one is freed

from the intrusions of local gossip, of face-to-face interactions, from close
scrutiny of one’s everyday behaviour by neighbours. … By the same token,
in the urban realm one can readily choose freedom, to be as private as one
likes, to mix with others on one’s own terms, to indulge in the exotic
without fear of reprimand, to be anonymous.
(Webster 1995: 56–7)

The breakdown of the organic community, the deterritorialisation of the local,

is, on the face of it, an immediate gain in terms of personal freedom. Freed from
the constraints of control of the organic community people become more free to
change. Narrow chauvinisms, conceptions of masculinity rooted in the indus-
trial plant and the local pub, respectabilities which were once policed by gossip
and sanctioning, all begin to crumble.
It is conventional, particularly in liberal political philosophy, to think of that
which is public as being good and that which is private as being concealed and
possibly reprehensive. But such liberalism which sees freedom as bringing
private problems into the light of public debate, however commendable, forgets
the sociology of resistance and subterfuge (see Fraser 1997). For the public
world, whether it is the local community or the wider polity, consists for the
powerless of distinctly unequal partners and fellow citizens who can be poten-
tially both censorious and coercive. Youth culture, for example, would be
moribund and conformist if it did not learn to manoeuvre the restrictions of
family by the device of deviance and half-truth. Teenagers are, in Dick
Hebdige’s marvellous phrase, ‘hiding in the light’ (1988). Similarly, from the
black diaspora (see Gilroy 1993) to the gay community (see Plummer 1995),
subcultures develop in the freedom of the urban landscape and spread into a
virtual community of mass media and cultural artefact. The privacy, therefore,
provided by the late modern city permits the exercise of freedom. Is it not more
possible in this environment, then, to develop genuine identity and a sense of
self than in the stifling atmosphere of the organic community?
Richard Sennett, in The Conscience of the Eye (1991), remarks that ‘deviance
is the freedom made possible in a crowded city of lightly engaged people’. But
he then adds a pessimistic caveat: that the tolerance of diversity is not so much
because the city encourages diversity but rather it permits it because it is merely
indifferent towards it – in other words, it could not care less. And, indeed, if our
mapping of the terrain of late modernity is correct, we can go further. Because
indifference can very easily disintegrate into predation and punitiveness from
both sides of the spectrum of criminal justice. Tony Bottoms and Paul Wiles
capture this well when they write:

Recent technology means that neither time nor space is the fixed frame-
work of our routines. … The result can be globalised cultures, no longer
fixed in time or space, from which we choose and indeed make a series of
different choices. Television representations based on the culture of
40 Jock Young

Australia or west coast America have taken on an autonomous existence

and may be as ‘real’ to some British youth as anything else. Yet geographi-
cally localised cultures, in the sense of ‘community’ and spatially fixed,
institutions, such as churches and families, have been regarded by much
criminological theory as the main defences against crime: hence the
Chicagoans’ concern with community ‘disorganization’ and crime.
(Bottoms and Wiles 1997: 351)

I have argued that there is no one-to-one relationship between the organic

community and the crime-free society. Yet how can we ensure that crime is
diminished rather than enhanced by the freedoms of the city?
Similarly the disembeddedness of the self, the collapse of fixed, pre-destined
narratives of biography, poses both problems and possibilities of progress. The
dangers of essentialism of creating a fixed identity for oneself and a demonisation
of the other is the premise for a punitive and exclusionary response to diversity.
Yet, as we have seen, the freeing up from these constraints also presents precisely
the opposite path: changes which might deconstruct the fixed categories of
gender, age and ethnicity, for example – those which open up rather than fore-
close on freedom. The question, therefore, boils down to how can we achieve and
maintain a civilised society which minimises oppressive and predatory behaviour
towards each other, in a world where the power of the localised organic commu-
nity is diminished and replaced by the new community of late modernity?

From generalised other to generalised elsewhere

Now, physically bounded spaces are less significant as information is able to

flow through walls and rush across great distances. As a result, where one is
has less and less to do with what one knows or experiences. Electronic
media have altered the significance of time and space for social interaction.
(Meyrowitz 1985: viii)

The notion of who you are becomes constructed on a much wider stage in late
modernity. To understand this one must look at the late modern self and its
reference points. Joshua Meyrowitz (1989), in a seminal article, develops the
work of Charles Cooley and George Herbert Mead on the generalised other.
The self according to the symbolic interactionist tradition is given reality by its
reflections in the significant others around us. Cooley calls this ‘the looking
glass self’, and Mead terms it ‘the generalized other’ – we see ourselves through
our perceptions of others’ perceptions of us. According to Meyrowitz, the rela-
tive decline in community and the rise in the media has:

Extended the generalized other so that those who we perceive as significant

others are no longer only the people we experience in face-to-face interac-
tion within the community. People from other communities and localities
 Identity, community and social exclusion 41

also serve as self-mirrors. The ‘mediated generalized other’ weakens (but

surely does not eliminate) our dependence on locality and on people in it
for a sense of self.
(Meyrowitz 1989: 327)

We can expand this idea a little further, since the series of other reference points
against which we judge ourselves fairly or unfairly treated in comparison with
others, widen out as does our knowledge of what is fairness and unfairness and its
distribution. Reference groups are much less attached to locality in late modernity,
for instance, as Bottoms and Wiles (1997) pointed out, the culture of Australia or
the west coast of America can be as ‘real’ to British youth as anything else.
How does this relate to crime and its control? What has happened is that
people’s notion of their self (and hence their sense of shame, of losing self-
respect when certain norms are transgressed), the actual norms themselves – the
informal mores which structure behaviour, the feelings of discontent which
provide the wellsprings of criminality and the vocabularies of motive and justi-
fying circumstances of crime – to a greater extent than ever before are a product
of discourses which are of a global rather than a local nature.
Relative deprivation, for example, is global in its comparison points: aspira-
tions jump frontiers, discourses about crime (including notions of fear, risk and
danger) are a free-floating commodity of a world media, the informal mores of
everyday life (including the introduction of new and more or less stringent defini-
tions of deviance) are constituted within public cultures which are global in their
reach. Indeed, the twin impact of cultural and economic globalisation penetrates
the local, moulding aspirations and changing opportunities, so that the causes of
crime can invariably be tied to global processes. All that is consistently local
about crime itself (and here only conventional crimes) is its actual impact – the
burglary of the dwelling, the violence on the street, the brutality in the home.
But let us return to the pressing problem: the maintenance of order among
lightly engaged strangers. In late modernity the generalised other immediately
presented to the self in the organic community by the reactions of neighbour
and local friend becomes potentially more remote. Deviance is less likely to be
directly observed, there are more nooks and crannies in the contemporary
world, but the weakening of the local is accompanied by the rise of the ‘gener-
alised elsewhere’: our notions of respectable and proper behaviour are then a
product of a wider interaction and a more public discourse. Thus our feelings of
shame and guilt are no longer solely constituted by the face-to-face encounter,
but shame and guilt scarcely vanishes. Indeed it might well be argued that in a
society where identity is both precarious and sought after, the need for the
recognition of social worth is all the more pressing.

On the edge: the test case of the Philadelphian underclass

In The Exclusive Society, I examine the notion that the outcasts of US society
are a distinct and localised entity. In particular I looked at Carl Nightingale’s
42 Jock Young

brilliant ethnography of the ghetto of Philadelphia. The conservative image is a

spatially segregated underclass, whose values are different from the wider
society, who have evolved a highly differentiated and dysfunctional way of life.
They are a caste apart, a disgrace to the American dream (see, for example,
Murray 1984). A common liberal image is not dissimilar to the conservative
one: the underclass are the excluded, the people whom the economy has left
behind, whose social behaviour is dependent and aberrant. The problem, they
believe, is to resocialise them, retrain them, get them to work and back into
society (see for example Social Exclusion Unit 1999a and 1999b).
Nightingale’s research confounded these images. For here in the ghetto, at
the bottom of this most segregated society, instead of a separate culture, what he
found was the apotheosis of America. He discovered a community fully
immersed in the American dream, a culture hooked on Gucci, BMW, Nike.
They watched television eleven hours a day, shared mainstream culture’s obses-
sion with violence, backed, at the time of the study, Bush’s involvement in the
Gulf War, lined up outside the cinemas, worshipped success, money and status,
and, in a perverse way, even shared the racism of wider society. The problem of
the ghetto was not so much the process of it being simply excluded but rather
one which was all too strongly included in the culture but, then, systematically
excluded from its realisation. All of this is reminiscent of Merton – but where, in
a late modern context, the implosion of the wider culture on the local is dramat-
ically increased. We have a process which I have likened to a bulimia of the
social system: a society which choruses the liberal mantra of liberty, equality and
fraternity, yet systematically in the job market, on the streets, in the day-to-day
contacts with the outside world, practices exclusion. It brands as ‘losers’ those
who had learnt to believe that the world consisted of ‘winners’ and ‘losers’.
None of this is to suggest that there is not a subcultural diversity within late
modern societies, but this is what it is – subcultural. Cultures criss-cross with
each other, hybridise, transform and constantly change. They are not separate
in any essentialist fashion. Nor, for that matter, are they separate in such a strict
spatial sense as is frequently suggested. Thus Zygmunt Bauman writes of
Washington, DC:

One difference between those ‘high up’ and those ‘low down’ is that the
first may leave the second behind – but not vice versa. Contemporary cities
are sites of an ‘apartheid à rebours’: those who can afford it, abandon the
filth and squalor of the regions that those who cannot afford the move are
stuck to. In Washington D.C. … there is an invisible border stretching
along 16th Street in the west and the Potomac river in the north-west,
which those left behind are wise never to cross. Most of the adolescents left
behind the invisible yet all-too-tangible border never saw downtown
Washington with all its splendours, ostentatious elegance and refined plea-
sures. In their life, that downtown does not exist. There is no talking over
the border. The life experiences are so sharply different that it is not clear
what the residents of the two sides could talk to each other about were they
 Identity, community and social exclusion 43

to meet and stop to converse. As Ludwig Wittgenstein remarked, ‘If lions

could talk, we would not understand them’.
(Bauman 1998: 86)

This eloquent expression of the dual city thesis is wrong, not in its sense of divi-
sion, but in its sense of borders. For the borders are regularly crossed and the
language spoken on each side is remarkably similar. The most obvious flaw in
this argument is that of gender: maids, nurses, clerical staff move across borders
into work everyday. Women, as William Julius Wilson argues in When Work
Disappears (1996), are more acceptable to the world outside of the ghetto than
their male counterparts. It is after all ‘home boys’ who stay at home. But bell-
hops, taxi drivers, doormen, maintenance men regularly ply their way across the
invisible borders of Washington, DC. It is not, therefore, just through television
that the sense of relative deprivation of the poor is heightened; it is in the
direct and often intimate knowledge of the lives of the affluent.

Including the excluded

The decline of old industries and the shift to an economy based on knowl-
edge and skills has given rise to a new class: a workless class … Today the
greatest challenge for any democratic government is to refashion our insti-
tutions to bring this new workless class back into society and into useful
work, and to bring back the will to win.
(Tony Blair, speech at the Aylesbury Estate, Southwark, 2 June
1997, cited in Peck 1999)

In Britain the work of the Social Exclusion Unit (1999a), set up under the New
Labour administration, explicitly regards its task as the integration of what it
sees as socially disorganised communities into the mainstream. This task is
explicitly aimed at reducing the levels of crime and disorder within society as a
whole. So, once again, we have a notion of the dual city. But the political aim
here is overtly inclusionist and can fittingly be seen as a social democratic
attempt to tackle the correlate of social problems occurring in the poorest part
of society. Here a key concept is the notion of the welfare dependency of the
poor and the emancipating effect of work in transforming their lives.

Table 1.1 The binaries of social exclusion

Society at large The underclass
The unproblematic The problem
Community Disorganisation
Employment The workless
Welfare dependency Independence
Stable family Single mothers
The natives The immigrants
Drug free Illicit drug use
44 Jock Young

The danger of the concept of social exclusion is that it carries with it a series
of false binaries: it ignores the fact that problems occur on both sides of the line,
however much one has clusters in one area rather than another and, more
subtly, it conceals the fact that the ‘normality’ of the majority is itself deeply
problematic.
Thus in the first respect, as we have seen, unemployment, poverty and
economic insecurity is scarcely unknown outside the designated areas – indeed
quantitatively they are overall more prevalent in the supposedly secure majori-
tarian heartlands of society than they are in the selected minority of ‘excluded’
areas. And the same, of course, is true of illicit drug use, community disorganisa-
tion, unstable family structures, etc. In the case of the notion of ‘the normal
majority’ it assumes that, in this world, class differentials are somehow insignifi-
cant, that paid work is an unambiguous benefit, that ‘stable’ family life is
unproblematic, licit psychoactive drug use is less a problem than illegal drug
‘abuse’, and so on. Furthermore, it assumes that the transition from the socially
excluded to the socially included majority via the vehicle of work will miracu-
lously solve all these problems.
The centrality of work to the process of inclusion cannot be overemphasised.
Being a member of society, being included, means being in work and useful work
means paid work, labour sold in the market place. As Ruth Levitas clearly puts it:

Under cover of a concern with ‘social exclusion’, and a rhetoric of soli-

darity, society dissolves into market relations. The importance of unpaid
work to the maintenance of social life and human relationships is ignored.
The possibility of integration into society through any institution other
than the labour market has disappeared. There is no such thing as society –
only individual men and women and their jobs.…
The concept of social exclusion as it is currently deployed places people
either inside or outside mainstream society, synonymous with outside the
labour market. The concept works both to devalue unpaid work and to
obscure the inequalities between paid workers – not to mention the
inequalities between paid workers and a property-owning class who can
afford not to work at all, but who are apparently not among the ranks of the
socially excluded. But then they do have a relationship with the labour
market: it just happens to be one of exploitation.
(Levitas 1996: 12, 19)

From community to public sphere

Having discarded the notion of a series of organic communities either actually
in existence or, as in the communitarian dream, to be greatly regenerated and
refurbished, as nostalgic and impractical can we, therefore, substitute a series of
virtual communities with some territorial basis – a multiculturalism of a late
modern sort? Thus we have the gay community, the Sikh community, the Irish
community, the black community, women, etc. Such a formulation is a currency
 Identity, community and social exclusion 45

of contemporary politics and the media: events are publicly examined and
debated by turning to representatives of various ‘communities’. There can be no
doubt that such a formulation has some foundation. A whole series of what
Nancy Fraser calls ‘subaltern public spheres’ occur where genuine debates occur
and which can by careful argument and presentation influence the debate
within the more general public sphere. A key example which Fraser gives, is
that of second wave feminism which as a new social movement has elaborated
extensive networks, journals, activist groups and discussion centres. Further-
more, such activism has produced broad debate in the wider public sphere over
a whole series of issues concerned with crimes against women: sexual harass-
ment, rape, domestic violence, etc., many of which has resulted in changes in
public attitudes and a broad raft of legislation. But it would be wrong to see such
sections of the population as late modern equivalents of the organic community.
Nancy Fraser, in her essay ‘Sex, Lies and the Public Sphere’ (1997), discusses
the 1991 struggle over the confirmation of Clarence Thomas as an associate
justice of the US Supreme Court who was nominated as only the second
African American on the Court in US history. Thomas was accused of sexual
harassment by Anita Hill, a black female law professor who had served as
Thomas’ assistant at the Equal Employment Commission in the 1980s. Fraser
takes us through a fascinating account of the discourses surrounding this
struggle: for discourses of gender, race and class each entered the public arena as
the debate developed. What became clear was that there was no clear line from
women, from blacks, from the middle- or working-class white males. It was not
that self-conscious ‘communities’ of a sort existed, particularly in this context of
American multiculturalism, but that each divided and crossed in their alliances.
A result was what Fraser (1997) called ‘the fracturing of the myth of homoge-
neous ‘ “communities” ’. As a result, she continues, it would be better to
consider:

Replacing the homogenizing ideological category of ‘community’ with the

potentially more critical category of ‘public’ in the sense of a discursive
arena for staging conflicts … In these respects, the concept of a public
differs from that of a community. ‘Community’ suggests a bounded and
fairly homogeneous group, and it often connotes consensus. ‘Public’, in
contrast, emphasizes discursive interaction that is in principle unbounded
and open-ended, and this in turn implies a plurality of perspectives. Thus,
the idea of a public, better than that of a community, can accommodate
internal differences, antagonisms, and debates.
(Fraser 1997: 118)

We start then from noting how the notion of ‘community’ has changed remark-
ably in late modernity but, further to this, that what takes its place is not simply
a series of discrete multicultural communities with both local and virtual
dimensions, which both criss-cross and are contested. The community loosens
46 Jock Young

its mooring in the locality and the various public discourses no longer have any
one-to-one relationship to a specific section of the population.

Conclusion
It is now time to bring the threads of this discussion together. The problem of
crime is inevitably one of order, to tackle crime we must, therefore, involve
the politics of distribution and the politics of recognition. We must, in short,
intervene both on a material and a symbolic level. This chapter has been
concerned largely with the latter although it has been argued that without
some form of redistribution any considerable reduction in crime is unlikely.
The significance of the symbolic level has changed remarkably with the tran-
sition to late modernity. First, a more individualistic society generates greater
and greater demands for self-actualisation and recognition. Second, the
increased sense of disembeddedness makes, at the same time, a sense of secure
identity more and more precarious. Third, a potent solution to this ontolog-
ical uncertainty is that of essentialism. Fourth, such a fake sense of solidity is
more easily achieved by denigrating others. And finally, such a dehumanisa-
tion of others can be a potent facilitator both of crime (particularly violence)
and a punitive attitude towards the criminal. It is therefore crucial that we
attend to the problems of identity, arguing for policies which ensure a sense of
self-worth and actualisation yet which do not rest upon the fake premises of
essentialism where others are systematically denigrated and then abused. Not
the least reason for this is that such a sense of identity brings with it the
informal norms which apportion praise and shame and help control predatory
crime towards others.
Nancy Fraser, in Justice Interruptus (1997), develops an extremely useful
typology of the politics of reform based on the two dimensions of redistribution
and recognition. Reform, she argues, must recognise the necessity of changes in
both these areas, assuaging the failings of distributive justice and misrecognition
and devaluation. But to this dichotomy she adds a further distinction: between
the politics of affirmation and the politics of transformation. Affirmative poli-
tics merely involves the surface transfer of resources without changing the basic
underlying divisions whereas transformative politics seek to eliminate the basic
underlying structures of injustice (see Young 1999; Mooney 2000). Thus in the
area of redistribution affirmative remedies involve, for example, coercing the
underclass into the labour market at extremely low wages. Their underclass
position is merely reproduced this time within the lower reaches of the market
place (see Levitas 1996). This dragooning of people from one category of exclu-
sion to another (‘getting the people to work’, as the Social Exclusion Unit
[1999a] put it, with its cheerless double entendre) is experienced all too
frequently, not as inclusion but as exclusion, not as the ‘free’ sale of labour but
as straightforward coercion. Relative deprivation would, of course, not be solved
by such ‘inclusionary’ politics and the sources of discontent which are liable to
generate high crime rates would be unabated. Transformative redistribution, on
 Identity, community and social exclusion 47

the other hand, would involve such measures as retraining so that jobs could be
gained and then rewarded on a meritocratic basis – thus putting a genuine
element of equality into equal opportunity policies, the recognition of non-paid
work (e.g., child rearing, caring for ageing parents) as of vital importance for
social reproduction, the creation of viable childcare infrastructures for women
with children, and the enforcement of a minimum wage on a level which allows
the individual an existence which is neither demeaning nor severely straitening
in circumstance. Above all, it would not fetishise paid work – it would not view
such work as the vital prerequisite for full citizenship, for acceptance and inclu-
sion in society.
An affirmative politics of recognition does not question the various essen-
tialisms of difference. That is, in the case of conventional multiculturalism,
what is stressed is the need for the positive recognition of various groups on
equal terms, for example: Irish, African-Caribbean, gays, women, etc. In
contrast, transformative politics seek to break down and destabilise the cate-
gories by questioning the very notion of fixed identity and essence. Thus the
invented notion of tradition is challenged, the overlapping, interwoven nature
of what are supposedly separate cultures stressed, and the ambiguity and blurred
nature of boundaries emphasised. Diversity is encouraged and, where non-
oppressive, celebrated, but difference is seen as a phenomenon of cultures in
flux not essences which are fixed.
In the case of crime and punishment, the critique of essences both in crim-
inal victimisation and in punishment is a high priority. The category of hate
crimes must be widened out in the realisation that a considerable proportion of
acts of violence involve vocabularies of motive which debase and dehumanise
the victim (see Young 1990). Thus not only crimes against gays and blacks, but
against women, the elderly, the poor, etc. In terms of our response to crime it is
vital that the essentialism which runs through the discourses about crime and
its causes is thoroughly debunked. Important, here, is to confront and shatter
the triptych which locates crime spatially and socially in three loci – the under-
class, the drug user and the immigrant. Such a combination, portrayed as
interdependent and very frequently racialised, is presented as the major source
of crime and disorder in our society.
Against this we must emphasise that crime occurs throughout the structure
of society and that its origins lie not in a separate aetiology but in the structure
of society and its core values. The identification of a distinct criminal class is an
endeavour bound to failure. Politicians forget this at their peril. As I write, the
British Prime Minister Tony Blair roundly castigates drunken hooligans one
week and calls for robust legislation to bring them under control when in the
next week his own 16-year-old son is arrested for drunkenness in London. A
year previously, the Home Secretary, Jack Straw, famous for his tough on crime
approach and the appointment of a drug Czar, is awakened from sleep by a tele-
phone call from the police informing him that his son has been arrested for
selling drugs. As the perceptive journalist Joan Smith (2000) put it:
48 Jock Young

The Government’s responses are off the cuff and authoritarian … Again
and again it reveals an us-and-them mentality as though there are only two
Britons: decent God-fearing folk whose only transgression is the occasional
parking ticket and a violent, anti-social sub-class whose members habitually
exploit drugs and alcohol and deliberately go out deliberately looking for
trouble.
(Smith 2000: 13)

At the start of this chapter I discussed the need to chart the journey into late
modernity. It was necessary to examine the terrain, chose our means of travel
and be clear as to our destination. We have seen how the terrain has changed
dramatically: employment, family, community – the structure of society, has
become less secure, boundaries blur, identities are less and less fixed, place and
social category become less determinate in prescribing behaviour, vocabularies
of motive lose their mooring in discrete parts of the structure – we have entered
the period of what Bauman (2000) graphically calls ‘liquid modernity’. And this
terrain has become a more risky place both in terms of crime and disorder and
in terms of demonisation and scapegoating. On the one hand, the organic
community has diminished and fragmented becoming much less capable of
controlling the rising tide of crime and discontent, whilst on the other hand,
community itself becomes reinvented as a mythical Gemeinschaft which serves
to exclude and essentialise others.
Yet the organic community of the past is in terminal decline, although not
into a black hole of atomistic individuals devoid of trust as the more dystopian
of commentators would have it. For it has reformed into the late modern
community: virtual and mediated in part and global in its reach which touches
down at the local sometimes substantially but always intermittently in the biog-
raphy of each individual. It is in this new community, the ‘generalised
elsewhere’, where old identities are discarded and new identities reconstructed.
It is here both in the more general public sphere and the host of subaltern
publics where concepts of appropriate and inappropriate behaviour, and the
criteria of social worth and opprobrium are regularly debated. Never have so
many people looked at so many people: never in human history has there been
such a degree of reflexivity about human behaviour. The paradox of identity, of
course, is that it is precisely at the point in history where it becomes most
apparent as a social construction that there is a widespread desire for essence
and fixity. The importance of the new social movements here is vital. For they
become the key axes of the debate about gender, ethnicity, or relations with
other species and the environment. The worry, of course, even here, is the
tendency to essentialise whether it is anti-racists who give credence to ‘race’,
radical feminists who essentialise masculinity and femininity or gay rights
activists who begin to believe in the biological basis of homosexuality.
Furthermore, as Fraser (2000) points out in a recent essay, it is necessary to
recognise the need to tackle the institutionalised patterns which maintain such
essences. For example, the institutionalised racism within police practice which
 Identity, community and social exclusion 49

causes such disproportionate focusing on African-Caribbeans, the Irish, etc. (see

also Mooney and Young 2000).
The terrain has changed and, of course, with it the available means of
change. Thus as Hans Hofman (1996) pointed out, the worthy social demo-
cratic critiques of society which link crime and punitiveness to lack of stabile
employment, community and family life assume that we can nostalgically bring
these entities of the 1960s back into existence by an act of political will. We
have seen how, in the case of community, this is an implausible dream, possible
only for a minority. Artificially created communities, such as Disney’s new town
Celebration in central Florida (Ross 1999), are the exceptions which prove this
rule. But this is true of the other institutional areas. Take paid work as an
example, an important site both of distributive justice and of identity. Herein,
as Andre Gorz trenchantly puts it:

Is an enormous fraud. There is not and never will be ‘enough work’

(enough paid, steady, full-time employment) for everyone any longer, but
society (or, rather, capital), which no longer needs everyone’s labour, and is
coming to need it less and less, keeps on repeating that it is not society
which needs work (far from it!), but you who need it, … Never has the
‘irreplaceable’, ‘indispensable’ function of labour as the source of ‘social
ties’, ‘social cohesion’, ‘integration’, ‘socialization’, ‘personalization’,
‘personal identity’ and meaning been invoked so obsessively as it has since
the day it became unable any longer to fulfil any of these functions …
Having become insecure, flexible, intermittent, variable as regards hours
and wages, employment no longer integrates one into a community, no
longer structures the daily, weekly or annual round, or the stages of life, and
is no longer the foundation on which everyone can base his/her life project.
The society in which everyone could hope to have a place and a future
marked out for him/her – the ‘worked-based society’, in which he/she could
hope to have security and usefulness – is dead. Work now retains merely a
phantom centrality: phantom in the sense of phantom limb from which an
amputee might continue to feel pain.
(Gorz 1999: 57–8)

Work in the sense of that which involves self-realisation and creativity, is not,
of course, dead but secure, paid, full-time employment for life is considerably
diminished and where it exists does not have this quality. Hence the title of
Gorz’s book ‘Reclaiming Work’. Work, like the community, needs to be refor-
mulated if we are to seek to provide the basis of identity and social worth.
Lastly, let us look at our final destination. The transition to late modernity is
one which involves the most dramatic changes in the fabric of society. Anthony
Giddens (1990) describes it like a juggernaut sweeping all solid institutions
aside, Francis Wheen (1999), in his recent biography of Marx, wryly notes how
the images of globalisation and the metaphor of the ‘melting’ of all that seemed
solid in the Communist Manifesto is prescient more of the present time than in
50 Jock Young

the nineteenth century. Indeed there is more than a slight resonance of today
when we read of a world where there is ‘uninterrupted disturbance of all social
conditions, everlasting uncertainty and agitation … fixed, fast-frozen relations,
with their train of ancient and venerable prejudices and opinions are swept
away’. Such changes with heavy irony swept away the fossilised state socialist
regimes of Eastern Europe just as they transform our lives in the West.
Community, work, the family – all the major institutions of social order, face a
transformation. Whether this is in the direction of greater equality and a sense
of self-worth or towards inequality and essentialism is the central hub of the
politics of the future.

Notes
1 And, of course, such a process of deconstruction has been characteristic of critical
analysis throughout the humanities and the social sciences. What is worth noting is
that the period in which critical criminology emerged in the late 1960s and early
1970s was at the cusp of change within which dramatic transformations were occur-
ring throughout the social fabric of advanced capitalist societies. The early critiques
were, so to speak, a harbinger of what was to come and what, at that time was a rele-
vant yet perhaps technically peripheral critique of modernity, has become a pressing
and key concern of any social analysis of late modernity.
2 In fact she points to a shift from one to the other so that ‘the struggle for recognition
is fast becoming the paradigmatic form of political conflict in the late twentieth
century’ (Fraser 1997: 11). Her predictions as to the decline of the social democratic
politics of class are undoubtedly incorrect: what her own work vividly shows is that
the discourses of both class and identity politics co-exist even in the United States
(see, for example, her remarkable essay on the Clarence Thomas case in her book
Justice Interruptus [1997]), as also does her discussion of methods of social transforma-
tion lean heavily on the co-existence of both dimensions (of which more later).
Indeed, her distinction between the politics of distribution and recognition and the
strategies needed to overcome maldistribution and misrecognition are of great use
and versatility.
3 I have paired the concepts of distributive justice and relative deprivation and
misrecognition and ontological deprivation. Fraser’s distinction echoes the tradi-
tional Weberian distinction between class and status and it is significant that
Runciman (1966) in his original, although largely unexplored, discussion of depriva-
tion utilised as I am doing here relative deprivation on both these dimensions. In a
recent essay, Fraser (2000) acknowledges the Weberian influence on her work.
4 Of course, all of these authors rediscover that which was a mainstay of traditional
subcultural theory. Such is the amnesia in criminology that the pivotal work of
Cloward and Ohlin (1961), particularly in their discussion of the organised and
disorganised slum, is not built upon.

References
Back, L. (1996) New Ethnicities and Urban Culture, London: UCL Press.
Bauman, Z. (1998) Globalization, Cambridge: Polity Press.
—— (2000) Liquid Modernity, Cambridge: Polity Press.
Beck, U. (1992) Risk Society, London: Sage.
Berman, M. (1983) All That’s Solid Melts Into Air, London: Verso.
 Identity, community and social exclusion 51

Bottoms, A. E. and Wiles, P. (1997) ‘Environmental Criminology’, in M. Maguire, R.

Morgan and R. Reiner The Oxford Handbook of Criminology, 2nd edition, Oxford:
Clarendon Press.
Campbell, C. (1987) The Romantic Ethic and the Spirit of Modern Consumerism, Oxford:
Blackwell.
Castells, M. (1994) ‘European Cities, the Informational Society, and the Global
Economy’, New Left Review, 204 (March–April): 19–35.
Cloward, R. and Ohlin, L. (1961) Delinquency and Opportunity, London: Routledge and
Kegan Paul.
Crawford, A. (1997) The Local Governance of Crime, Oxford: Oxford University Press.
Currie, E. (1997) ‘Market Society and Social Disorder’, in B. MacLean and D. Milanovic
(eds) Thinking Critically About Crime, Vancouver: Collective Press.
Damer, C. (1974) ‘Wine Alley: The Sociology of a Dreadful Enclosure’, Sociological
Review, 22: 21–48.
Davis, M. (1990) City of Quartz, London: Vintage.
Elias, N. (1982) State, Formation and Civilisation: The Civilising Process, trans. E. Jephcott,
Oxford: Blackwell.
Etzioni, A. (1997) The New Golden Rule, London: Profile Books.
Featherstone, M. (1985) ‘Lifestyle and Consumer Culture’, Theory, Culture and Society,
4: 57–70.
Foster, J. (1997) ‘Challenging Perceptions of “Community” and Neighbourliness on a
Difficult-to-Let Estate’, in N. Jewson and S. MacGregor (eds) Transforming Cities,
London: Routledge.
Fraser, N. (1997) Justice Interruptus: Critical Reflections on the Post-Socialist Condition, New
York: Routledge.
—— (2000) ‘Rethinking Recognition’, New Left Review, 2/3: 107–20.
Giddens, A. (1990) The Consequences of Modernity, Cambridge: Polity Press.
—— (1991) Modernity and Self-Identity, Cambridge: Polity Press.
Gilroy, P. (1993) The Black Atlantic, London: Verso.
Gitlin, T. (1995) The Twilight of Common Dreams, New York: Henry Holt.
Gorz, A. (1999) Reclaiming Work: Beyond the Wage-Based Society, Cambridge: Polity
Press.
Harvey, D. (1997) ‘Contested Cities: Social Process and Spatial Form’, in N. Jewson and
S. MacGregor (eds) Transforming Cities, London: Routledge.
Hebdige, D. (1988) Hiding in the Light, London: Routledge/Comedia.
—— (1990) ‘Fax to the Future’, Marxism Today (January): 18–23.
Hobbs, D. (1988) Doing the Business, Oxford: Clarendon Press.
Hobsbawm, E. (1994) The Age of Extremes, London: Michael Joseph.
—— (1996) ‘The Cult of Identity Politics’, New Left Review, 217: 38–47.
Hofman, H. (1996) ‘Kritische Criminologie en Preventie in het Licht van een Postmo-
derne Conditie’, Tijdschrift voor Sociale Wetenschappen, 41(2): 109–205.
Hogg, R. and Brown, D. (1998) Rethinking Law and Order, Annandale, NSW: Pluto.
Hood, R. and Jones, K. (1999) ‘Three Generations: Oral Testimonies on Crime and
Social Change in London’s East End’, British Journal of Criminology, 31(1): 136–60.
Hope, T. (1995) ‘Community Crime Prevention’, in M. Tonry and D. Farringdon (eds)
Building a Safer Society: Strategic Approaches to Crime Prevention, Crime and Justice.
Crime and Justice 19, Chicago: University of Chicago Press.
Hughes, G. (1998) Understanding Crime Prevention, Buckingham: Open University Press.
Hutton, W. (1995) The State We’re In, London: Cape.
52 Jock Young

Jacobs, J. (1961) The Death and Life of Great American Cities, New York: Random House.
Levitas, R. (1996) ‘The Concept of Social Exclusion and the New Durkheimian Hege-
mony’, Critical Social Policy, 16(46): 5–20.
Luttwak, E. (1995) ‘Turbo-Charged Capitalism and its Consequences’, London Review of
Books, 17(21): 6–7.
McRobbie, A. and Thornton, S. (1995) ‘Rethinking Moral Panic for Multimediated
Social Worlds’, British Journal of Sociology, 46(4): 559–74.
Matthews, R. (1988) ‘Reassessing Informal Justice’, in R Matthews (ed.) Informal Justice?,
London: Sage.
Meyrowitz, J. (1985) No Sense of Place: The Impact of Electronic Media on Social Behaviour,
New York: Oxford University Press.
—— (1989) ‘The Generalized Elsewhere’, Critical Studies in Mass Communication, 6(3):
326–34.
Mooney, G. and Danson, M. (1997) ‘Beyond Culture City: Glasgow as a Dual City’, in
N. Jewson and S. MacGregor (eds) Transforming Cities, London: Routledge.
Mooney, J. (2000) Gender, Violence and the Social Order, London: Macmillan.
Mooney, J. and Young, J. (2000) ‘Policing Ethnic Minorities’, in B. Loveday and A.
Marlow (eds) Policing After the Stephen Lawrence Inquiry, Lyme Regis: Russell House.
Morley, D. and Robins, K. (1995) Spaces of Identity, London: Routledge.
Murray, C. (1984) Losing Ground, New York: Basic Books.
Nightingale, C. (1993) On the Edge, New York: Basic Books.
Peck, J. (1999) ‘New Labourers? Making a New Deal for the “Working Class” ’, Environ-
ment and Planning Government and Policy, 17: 345–72.
Plummer, K. (1995) Telling Sexual Stories, London: Routledge.
Robins, D. (1992) Tarnished Vision, Oxford: Oxford University Press.
Ross, A. (1999) The Celebration Chronicles, New York: Balantine.
Runciman, W. (1966) Relative Deprivation and Social Justice, London: Routledge and
Kegan Paul.
Schelsky, H. (1957) ‘Ist die Dauerreflektion institutionalisierbar?’, Zeitschrift für Evange-
lische Ethik, 1: 153–74.
Sennett, R. (1991) The Conscience of the Eye, London: Faber.
—— (1998) The Corrosion of Character, New York: W. Norton and Co.
Smith, J. (2000) ‘Mr Blair gets a wake-up call from the real world’, Evening Standard, 6
July: 13.
Social Exclusion Unit (1999a) Bringing Britain Together: A National Strategy for Neigh-
bourhood Renewal, London: The Stationery Office.
—— (1999b) Teenage Pregnancy, London: The Stationery Office.
Social Trends (2000) 2000 Edition, London: The Stationery Office.
Thompson, J. (1995) The Media and Modernity, Cambridge: Polity Press.
Tomlinson, M. (1999) Globalization and Culture, Cambridge: Polity Press.
Walklate, S. and Evans, K. (1999) Zero Tolerance or Community Tolerance, Aldershot:
Ashgate.
Webber, J. (1992) ‘Individuality, Equality and Difference: Justification for a Parallel
System of Aboriginal Justice’, in R. Silverman and M. Nielsen (eds) Aboriginal Peoples
and Canadian Criminal Justice, Toronto: Butterworths.
Webster, F. (1995) Theories of the Information Society, London: Routledge.
Wellman, B. (1982) Studying Personal Communities in East York (research paper no. 128),
University of Toronto: Centre for Urban and Community Studies.
Wheen, F. (1999) Karl Marx, London: Fourth Estate.
 Identity, community and social exclusion 53

Willis, P. (1990) Common Culture, Milton Keynes: Open University Press.

Wilson, J. Q. and Kelling, G. (1982) ‘Broken Windows’, Atlantic Monthly (March):
29–38.
Wilson, W. J. (1996) When Work Disappears, New York: Knopf.
Young, I. M. (1990) Justice and the Politics of Difference, Princeton, NJ: Princeton Univer-
sity Press.
Young, J. (1992) ‘Ten Points of Realism’, in J. Young and R. Matthews (eds) Rethinking
Criminology, London: Sage.
—— (1999) The Exclusive Society, London: Sage.
—— (2000) Globalization, Chaos and the Narcissism of Minor Differences, Middlesex
University: Centre for Criminology. Paper first given at The American Society of
Criminology, Toronto, 1999.
2 Joined-up but fragmented
Contradiction, ambiguity and
ambivalence at the heart of
New Labour’s ‘Third Way’
Adam Crawford

New Labour’s Third Way

The ‘Third Way’ has become a term which has both sociological (academic)
currency as an analysis of the ‘modern times’ in which we live (Giddens 1998;
2000) and political currency (Blair 1998) as a banner under which an evolving
set of ‘new’ political ideas – with associated programmes of action – coalesce.
The Third Way refers to the construction of a political agenda in response to,
what are perceived to be, profound changes unfolding across the world.
Traditional politics, it is argued, are inadequate to respond to contemporary
needs, including required responses to cultural diversity, scientific and technolog-
ical change, globalisation and ecological concerns. Contemporary conditions, it
is also argued, have caused a transformation in, and blurring of, traditional polit-
ical boundaries, as many issues cut across the Left–Right divide. Hence, the
Third Way seeks a social analysis and political response which goes ‘beyond the
Left and Right’.
Crudely put, the Third Way as identified by Anthony Giddens and Tony Blair
represents a quest to transcend social democracy, on the one hand, which looks
to the state for answers, and neo-liberalism, on the other hand, which looks to
the market for solutions. It is argued that, whilst neo-liberals historically have
wanted to shrink the state, social democrats have been keen to expand it. What
is necessary, rather, is to reconstruct a ‘new democratic state’ (Giddens 1998:
70). The reformed state, it is argued, will establish a new relationship between
risk and security, on the one hand, and individual and collective responsibility,
on the other.
The Third Way accepts anxieties and worries about civil decline as both real
and tangible within many sectors of contemporary societies, not just the inven-
tions of conservative politicians or the media. Hence, the loss of a sense of
community, weakening sociability, urban decline, fear of crime and the break-up
of marriages and families are seen as real problems needing to be addressed by
the Third Way. However, as Giddens suggests, ‘it is just as wrong to reduce civic
decline to economics, as the old left often did, as to deny the influence of
poverty and underprivilege’ (1998: 79).
Whilst the over-interventionist welfare state has removed individual
 Joined-up but fragmented 55

responsibility by (inadvertently) producing a ‘culture of dependency’, it is

argued, the neo-liberal marketisation of everyday life, in which the ‘public
good’ has become defined as merely the outcome of individual market-driven
decisions and choices, has left little sense of, or space for, public responsibility
and civic duty.

The life of any family and any community depends on accepting and
discharging the formal and informal obligations we owe to each other. The
politics of ‘us’ rather than ‘me’ demands an ethic of responsibility as well as
rights.
(Blair 1998: 14)

In sum, the Third Way is premised upon the short-comings of traditional social
democratic ideals that the ‘good life’ can be delivered through the (Keynesian)
welfare state, whilst seeking simultaneously to tame, ride and steer the appar-
ently ‘inevitable’ juggernaut-like pressures of globalisation, including the free
flow of capital, finance, trade, technology and information. Alongside these
global pressures is a growing salience of ‘place’ and locality. Global pressures, it
is argued, are refracted and interpreted through local meanings, identities and
sensibilities (Giddens 1991). As a result, there appears to be an increasingly
profound relationship between globalised conditions and local circumstances in
which the nature of the nation-state’s capacity to exercise political control is
called into question. In this sense, community safety both reflects a field of
policy that it is saturated with wider concerns about contemporary social
change and is a referent for an unfolding form of governance.

The Third Way for crime control

In the field of crime control and criminal justice policy the Third Way has been
captured by the infamous phrase, first used by Tony Blair in an interview for
BBC Radio 4 in 1993, namely: ‘Tough on crime; tough on the causes of crime’.
This catchphrase evocatively suggested a break with a ‘soft on crime’ stance
traditionally associated with Labour. Downes and Morgan (1997) note that, for
a Party historically tainted with many ‘hostages to fortune’ with regard to law
and order (albeit paradoxically with a record in government which was no
worse than that of the Conservative Party), this shift constituted a potential
political vote winner. Moreover, it marked a ‘watershed’ in so far as it broke
with the discourses associated both with the traditional Left of societal responsi-
bility for crime and with that of the Right which emphasised the purely
individual responsibility for crime (Downes 1998: 192). In so doing, it borrowed
selectively upon the pragmatism of Left Realist criminology in its proclamation
to take crime and the victims of crime seriously, focusing upon the ‘lived reali-
ties’ of ordinary people within a framework of moral responsibility (Young
1992). In the politics of the Third Way, obligations are both social and indi-
vidual. Giddens reduces this to a further catchphrase: ‘One might suggest as a
56 Adam Crawford

prime motto for the new politics, no rights without responsibilities’ (Giddens
1998: 65).
The New Labour government, elected in May 1997, wasted little time in
identifying and selecting crime control as a major plank of public policy upon
which its fortunes in office would hinge and upon which it should be judged. It
clearly sees ‘flagship’ pieces of legislation such as the Crime and Disorder Act
(1998) and the Youth Justice and Criminal Evidence Act (1999) as key
elements in the implementation of Third Way politics in the field of law and
order. The Home Secretary, Jack Straw, has suggested that the statutory commu-
nity safety partnerships established by the Crime and Disorder Act (1998)
alongside measures such as the anti-social behaviour orders (ASBO) and child
curfews will help bind together and empower communities to fight crime.
Through these and associated measures he declared:

We are trying to develop the concept of ‘the Active Community’ in which

the commitment of the individual is backed by the duty of all organisations
– in the public sector, the private sector and the voluntary sector – to work
towards a community of mutual care and a balance of rights and responsi-
bilities.
(Straw 1998: 16–17)

The proactive prevention of crime is a central and recurring theme in New

Labour’s policies. The duty on local authorities and the police to establish and
promote community safety partnerships and to put in place crime and disorder
strategies, has been acknowledged by many commentators as possibly the most
important elements of the recent legislation. It represents a fundamental
acknowledgement that the causes of crime lie far from the traditional reach of
the criminal justice system. This is a theme which also resonates with the trans-
formations to the youth justice system introduced by the 1998 Act and
extended by the 1999 Act. Youth justice now has the prevention of youth
crime as its over-riding central aim (s. 37). Moreover, the new community
safety partnerships and strategies recognise the need for social responses to
crime which reflect the nature of the phenomenon itself and its multiple aeti-
ology through a multi-agency perspective. This approach is reinforced through
reforms to youth justice which have introduced youth offending teams and
other proposals such as the joint local approaches to truancy (s. 16 of the 1998
Act). Moreover, the establishment of inter-departmental bodies at central
government level constitutes an important new addition to the institutional
landscape with which to advance ‘joined-up government’. The Social Exclusion
Unit (SEU), for example, has established ‘action teams’ to develop a national
strategy to tackle social exclusion and launched the New Deal for
Communities, which funds programmes of intensive regeneration of targeted
small neighbourhoods (SEU 1998).
Importantly, section 17 of the Crime and Disorder Act (1998) imposes a
duty on local authorities to consider the crime and disorder implications of their
 Joined-up but fragmented 57

various functions and the need to do all that they reasonably can to prevent
crime and disorder in their area, thus requiring local authorities to anticipate
the potential crime consequences of their policies. As the Home Office
Consultation Document noted, this will ‘give the vital work of preventing
crime a new focus across a very wide range of local services … putting crime and
disorder considerations at the very heart of decision making, where they have
always belonged’ (Home Office 1997: para. 33). Consequently, some commen-
tators have gone so far as to suggest that like ‘a wolf in sheep’s clothing’ this
section could come to constitute one of the most important parts of that Act
(Moss and Pease 1999).
Moreover, through the required community consultation processes in the
implementation of community safety strategies and community representa-
tion in local partnerships there is the potential to encourage a stronger and
more participatory civil society. This theme is continued through the restora-
tive justice elements in the youth justice reforms and particularly the
involvement of voluntary panel members on youth offending panels (under
the 1999 Act, currently being piloted), which attempt to reproduce forms of
‘community conferencing’ at the heart of the youth justice system in England
and Wales.
Together, many of the reforms inspired by Third Way politics potentially
afford a more holistic and problem-oriented approach to crime prevention and
community safety (Crawford 1998a). As such, they may allow a fundamental
shift in the way we govern crime and its prevention. Moreover, they challenge,
both implicitly and explicitly, many of the taken-for-granted modernist assump-
tions about professional expertise, hierarchical competence and specialisation,
as well as state paternalism and monopoly. In their place, the new politics offers
more plural understandings of and social responses to crime, drawing together a
variety of organisations and stakeholders, in the public, voluntary and private
sectors as well as from among relevant community groups in ways which are
problem-focused rather than defined according to the means most readily avail-
able for their solution. As such, they may allow a systemisation and
co-ordination of effort, expertise and information and a democratisation of
control through greater community empowerment.
Nevertheless, in this chapter I want to focus upon a number of sites of
tension, ambiguity and ambivalence in both Third Way politics and the manner
in which they are being implemented. I do so because it is in these conflicts and
tensions that we find some of the key determinants of contemporary law and
order policy and practice, as well as central components guiding the future
shape of crime control.

Central ideas at the heart of Third Way politics

Within Third Way politics, notably the crime and disorder elements, we can
identify three broad dominant influences, namely: communitarianism; manage-
rialism and a partnership approach. This is not to suggest that these are the only
58 Adam Crawford

dynamics which infuse New Labour’s crime control policies (see Downes 1998;
Newburn 1998) nor do they exhaust the range of influences upon change, but it
is my contention that they are key elements within these, both of specific
import and worthy of particular attention. Moreover, they expose and reveal
some of the central and broader tensions, ambiguities and ambivalences in
Third Way thinking more generally.

Communitarianism
As the above quotes already suggest, Third Wayism draws upon a communi-
tarian philosophy, notably that espoused by Amitai Etzioni (1993; 1997a),
whose work has explicitly sought to influence (Etzioni 1997b), and has
impressed, the Labour leadership (Blair 1995; 1996). Such ideas have been
advanced in Britain by the Demos think-tank (see Leadbeater 1996) and such
commentators as Henry Tam (1998). This influence expresses itself in the
diverse appeals to ‘community’ as a focus of moral renewal. As Etzioni
suggests:

Communities often have strong moral voices and hence can help maintain
a social order that draws significantly on value commitments … communi-
ties also share sets of values and reaffirm them, encourage members to abide
by these values, and censure the members when they do not.
(Etzioni 1997a: 123)

This form of ‘communitarian moralism’ asserts the need to restore communities

and their moral voices, requiring a greater emphasis upon individuals’ responsi-
bilities towards, rather than rights over, their communities. As such, it attempts
to reactivate institutions of civil society – notably communities, schools and
families – into forms of vibrant social control. It is concerned with a ‘remorali-
sation of justice’, through community as moral order (Etzioni 1997c).
Communitarianism is premised upon the notion that individuals are members
of one another, they are ontologically embedded in a social existence. Thus, we
gain our initial moral commitments from the communities into which we are
born, which over time are reinforced by other forms of community membership.
For communitarians ‘the social’ and the ‘community’ are no longer necessarily
complementary aspects of the same broad rationality of rule, but constitute
different and potentially ‘competing problematics of government’ (O’Malley
and Palmer 1996: 140). As Etzioni notes: ‘Society, as a community of communi-
ties, should encourage the moral expectation that attending to welfare is the
responsibility of the local community’ (1993: 146). Such communitarian ideas
infuse New Labour’s thinking and policies, as Jack Straw notes: ‘Concepts such
as community and personal responsibility, which have so long been buried in
the futile arguments between Left and Right, are at the centre of everything we
do’ (1998: 17).
 Joined-up but fragmented 59

Managerialism
Second, there is both an explicit and implicit managerialism at the heart of
New Labour’s reforms. In many ways, recent policy initiatives have intensified
the New Public Management (NPM) reforms introduced under previous
Conservative governments since the mid-1980s. These reforms have seen a
significant restructuring of the public sector and the role of the state, as well as
the introduction of a new regulatory style and managerial culture. NPM reforms
represent a cluster of ideas and strategies which variously have sought: to hive
off certain traditional aspects of public service delivery to the private sector;
introduce private sector management methods to the public sector; flatten
bureaucratic hierarchies; measure performance by results set against clear objec-
tives; disaggregate separable functions into quasi-contractual or quasi-market
forms; introduce purchaser/provider distinctions; open up provider roles to
competition between agencies and private interests; and advocate a ‘closeness
to the customer’ (see Hood 1991; 1995; Clarke and Newman 1997).
Under New Labour the managerialist urges to objective setting, performance
measurement and output-fixation, notably through ‘league table’ comparison,
has reached a new zenith. This is particularly apparent with regard to crime
control in the Crime and Disorder Act (1998) and the managerial processes
introduced to implement the local crime and disorder strategies. Lord Warner, a
senior policy adviser to the Home Secretary during the conception and passage
through Parliament of the Crime and Disorder Act (and now the Chair of the
Youth Justice Board of England and Wales), described the Act as reflecting ‘a
much more managerial feel about how you implement election promises’
(Warner 1998).

Partnerships, networks and joined-up government

A third influence, as already noted, is the appeal to ‘joined-up’ and ‘holistic’
government through partnerships with cross-cutting approaches to problem
solving. This has implications both within the state (as a critique of ‘depart-
mentalism’) and across the traditional state-civil society divide. With regard to
the former concern, the government declared in a recent White Paper, entitled
‘Modernising Government’, that ‘To improve the way we provide services, we
need all parts of government to work together better. We need joined-up
government. We need integrated government’ (Cabinet Office 1999: 5). With
regard to the latter concern, according to Giddens, ‘state and civil society
should act in partnership, each to facilitate, but also to act as a control upon,
the other’ (1998: 79). Hence, a partnership approach reverses the dominant
assumption of a single agency (notably police) ‘solution’ to crime. The appeal to
a partnership approach to community safety is premised upon the belief that
crime prevention lies beyond the competency of any one single agency.
Traditional segmented and compartmentalised social responses fail to provide
coherence, co-ordination and synergy. Partnerships, by contrast, propose an
60 Adam Crawford

holistic approach to crime and disorder which is problem-focused rather than

bureaucracy-premised.

The Third Way recognises the limits of government in the social sphere,
but also the need for government, within those limits, to forge new partner-
ships with the voluntary sector. Whether in education, health, social work,
crime prevention or the care of children, ‘enabling’ government
strengthens civil society rather than weakening it, and helps families and
communities improve their own performance … New Labour’s task is to
strengthen the range and quality of such partnerships.
(Blair 1998: 14)

Networks of interlaced agencies drawing from the public, private and volun-
tary sectors are heralded as alternatives to a primacy upon bureaucracies and
markets as the bedrocks of the new governance.

Sites of contradiction, ambiguity and ambivalence

It is my contention that these three dynamics, as well as being important in
their own right in understanding the direction of Third Way policies, constitute
what I have elsewhere described as ‘an unholy Trinity’ (Crawford 1998b: 242).
These visions are deeply embedded within the politics of the Third Way but
clash in many important respects. In so doing, they yield and reflect a number
of important sites of contradiction, ambiguity and ambivalence all of which
have significant implications for the future. I will consider a number of these in
turn under the following seven aphorisms:

• joined-up but fragmented

• arm’s length but hands on
• wide-angled but tunnelled vision
• growing demands for trust and the institutionalisation of distrust
• co-operation and negotiation in a cold climate of competition
• nostalgia disguised as modernisation
• ambivalent political responses

Joined-up but fragmented

At the heart of appeals to partnerships and policy networks is an ambiguity as to
what partnerships entail and their purpose, inclusiveness, responsibilities,
working relations and lines of accountability. The discussion tends to treat part-
nerships as if the public sector, voluntary organisations, private businesses,
communities and groups are undifferentiated clusters of organisations, as if they
present the same issues and opportunities as well as difficulties. There is little
sense of the diverse priorities and forces as well as the plural traditions, cultures
and practices which differentiate such clusters of interests. Little concern is
 Joined-up but fragmented 61

given to the problematic task of managing such networks, particularly in light

of the reality that conflicts are overlain by very different power relations and
access to resources (both human and material). It would appear that there is
little consideration as to what partnerships, as a Third Way between the state
and the market, actually entail. Giddens, for example, identifies but fails to
resolve this problem. He suggests that:

Reinventing government certainly sometimes means adopting market-solu-

tions. But it also should mean reasserting the effectiveness of government
in the face of markets.
(Giddens 1998: 75)

Far from wrestling with this conundrum, Third Way proponents sidestep it,
leaving a vacuum too easily filled with today’s dominant managerialist dogma,
which often resembles little more than the ‘marketisation of the public sector’,
through the injection of private sector management ideas, such as those popu-
larised by Osborne and Gaebler (1992).
More fundamentally, the ‘unholy Trinity’ has encouraged a pluralisation of
services and service providers. Managerialist reforms and appeals to community
involvement in policing and crime control have enticed new actors and active
citizens into what was once perceived as the sole responsibility of the public
police. As a consequence, we have witnessed the fragmentation and dispersal of
policing and control. Fragmentation, itself, has been encouraged by the intro-
duction of privatisation and quasi-markets through purchaser/provider splits
where services could not easily be privatised. We can no longer speak (if we
ever could) of a state monopoly in crime control and policing (Jones and
Newburn 1998). They have become shared, as diverse agencies, organisations,
groups and individuals are implicated in these tasks. This has produced a
complex array of networks, partnerships, interlaced alliances of organisations
and active citizenry that transcend the ‘public’ and ‘private’ spheres constituting
hybrid mixes of plural agencies, places and functions.
In many senses, the establishment of the new community safety partnerships,
by the Crime and Disorder Act (1998), will encourage and entrench this
process of fragmentation of service delivery (and service providers) at a local
level. The logic of the Act and the Guidance (Home Office 1998) which
accompanies it is to encourage a pluralisation of local service providers and
networks. It does so, not only by transcending the traditional workings of
particular local agencies, but also by opening up new policy arenas, encouraging
partnerships across and between the public, voluntary and private sectors. In
addition, the emphasis upon a locally-grounded, problem-solving methodology
encourages new local players to enter the field. However, the (unintended but
logical) consequence of this is to exacerbate co-ordination and undermine the
effectiveness of steering mechanisms.
This process of fragmentation itself has generated a demand for co-ordina-
tion and the conditions for the emergence of ever newer networks and
62 Adam Crawford

partnerships which attempt to draw together the diverse services in new

‘tangled webs’ (Charlesworth et al. 1996). The task of co-ordination is itself
undermined by the expanding complex of networks. As Rod Rhodes notes with
irony: ‘the competition-based fragmentation which undermined existing
networks created pressures to form new networks which, in turn, undermined
the competitive rationale of marketisation’ (2000: 353). Perversely, government
policy has encouraged fragmentation, not only by fostering the disaggregation of
purchaser/provider roles, sub-contracting and the introduction of quasi-markets,
but also through its response to earlier fragmentation in the form of new
networks. This has produced what the Social Exclusion Unit, in its report into
services for socially deprived neighbourhoods in Britain, referred to as ‘initia-
tive-itis’ (SEU 1998: 38). As Rhodes also suggests: ‘Such trends make steering
[by central government] more difficult, so the mechanisms for integration
multiply’ (2000: 349).
In the community safety field the new partnerships in England and Wales
must jostle for their own position on an already crowded terrain which includes
Drug Action Teams (DATs), Drug Reference Groups, Youth Offending Teams
(YOTs), Child Protection Committees, and Single Regeneration Budget (SRB)
partnerships, as well as in some areas, New Deal for Communities pilots, Sure
Start initiatives, Crime Reduction programme pilots and pathfinder projects
and health, education and employment action zones. In addition, the commu-
nity safety strategies are required to share a local arena with statutory demands
for a plethora of other local management ‘plans’: for health, education, housing,
policing and youth justice as well as police force objectives and probation
strategies which are clearly related but often remain uncoordinated. Moreover,
there is often no hierarchy of control or clear lines of accountability between
the various services and networks: they lack overall coherence and correspon-
dence. It is unsurprising, therefore, that according to Home Office research
evaluating recent youth justice changes introduced by the Crime and Disorder
Act 1998, ‘the experience of YOT managers in the pilot areas who have
responded to the range of requests for “joined up government” at the local level
has highlighted an apparent lack of co-ordination at national government level’
(Hines et al. 1999: 7).
The massive transformation of inter-organisational relations that a partner-
ship approach and ‘joined-up government’ entail have in part been hindered,
rather than advanced, by a managerialist culture. The baronial interests that
historically have structured agencies and organisations within and around the
criminal justice system have been disrupted only in part by managerialist
reforms. In the late 1980s Joanna Shapland (1988) likened the integration of
victims into criminal justice to a medieval baron system, whereby each baron or
criminal justice fiefdom jealously guards its piece of criminal justice processing,
only negotiating reluctantly with others. The particular difficulty for victims,
she argued, is that their needs span several fiefdoms who rarely communicate
with one another. As a consequence, trying to effect real change, even with
regard to simple and uncontroversial needs of victims, becomes heavily prob-
 Joined-up but fragmented 63

lematic. The same analogy applies equally well to community safety – and other
so-called ‘wicked issues’ – which, as the Morgan Report reminded us, ‘is a
peripheral concern for all the agencies, and a truly core activity for none of
them, even those agencies which explicitly include crime prevention within
their objectives such as the police and the probation service’ (1991: 15, para.
3.15). However, the managerialist reforms which increasingly have pervaded
criminal justice organisations with their emphasis upon performance measure-
ment, financial control and responsiveness to ‘customers’ for whom services are
delivered have not disrupted this image significantly and perversely may have
strengthened it (Shapland 2000). They have strengthened intra-organisational
control but often at the expense of concerns for inter-organisational relations (a
point to which I return below), thus exacerbating fragmentation. The supreme
paradox of managerialist reforms is that they are increasingly undermined by
plural partnerships and networks while simultaneously weakening the effective-
ness of the partnerships and networks that they spawned. They induce a
landscape which is simultaneously joined-up but fragmented.

Arm’s length but hands on1

As already noted, one of the primary managerialist techniques has been the
separation of purchaser and provider functions with regard to public services.
The rationale behind this has been to redraw the role of government to that of
policy-making which should be disentangled from the administration and oper-
ational delivery of services. This is a relationship in which government ‘steers’
by leading and setting agendas, as well as catalysing and facilitating change. It
leaves the task of ‘rowing’, on the other hand, the ‘doing’ of things, to others. It
is a relationship of ‘less government but more governance’ (Osborne and
Gaebler 1992). As the state attempts to ‘steer’ more and ‘row’ less it authorises,
licenses, audits and inspects the doings of others. Hence, they necessitate
processes of control and verification which allow those that ‘steer’ to monitor
and correct the activities of those that ‘row’ (ibid.: 32). This shift in regulatory
style demands new flows of information to fill the gulf created by the redrawing
of state functions. Contracts, performance indicators, audits and inspections are
some of the practical tools used to deliver this relationship of ‘governing at a
distance’. This ‘revolution’ in the relationship between state and civil society
suggests a situation in which the former asserts a form of control through the
setting of norms and the correction of deviations from them.
However, in practice there are difficulties implicit in observing effort and
performance and in obtaining information. For example, the experience from
the UK in health service reforms suggests that: ‘The purchaser is often depen-
dent on the provider for knowledge of what has been done, or even what should
be done, so that information becomes a key battleground in service manage-
ment’ (Deakin and Walsh 1996: 37). The lesson is that delivery systems once
set up may begin to exhibit a life of their own, which become increasingly diffi-
cult to rein back or control. This internal momentum can end up pulling
64 Adam Crawford

contract-based systems in directions that were not always anticipated by those

that set them up. Furthermore, partnerships once established tend to become
increasingly autonomous and resist central guidance. Hence, in practice, it is
often harder to distinguish those who are ‘rowing’ from those who are ‘steering’.
Inadvertently, government may be producing ‘self-steering networks’ (Rhodes
1997: 110).
Often the implication of these harsh realities for governments is that they
tend to fall back upon interventionist strategies. Fuelled by the pluralisation and
fragmentation, outlined above, a perception on the part of central government
can develop for the need to maintain a ‘hands on’ approach to the networks and
partnerships it has helped spawn. In this context, as Rhodes notes, for govern-
ment ‘ “hands off ” is the hardest lesson of all to learn’ (2000: 361).
This ambiguity for central government is illustrated in the content of the
Home Office (1998) Guidance on Statutory Crime and Disorder Partnerships. The
government has preferred to leave local partnerships to decide the content of
strategies. It ‘deliberately avoids attempting to define the terms ‘‘crime’’ or
‘‘disorder’’ and goes on to declare that ‘‘within reason, nothing is ruled out and
nothing is ruled in’’ ’ (Home Office 1998: para. 1.43).2 Nevertheless, the
Guidance goes on to identify a variety of issues which local partnerships would
do well to consider. More recently, however, the government has retreated from
its abstentionist position and been highly critical of the new community safety
partnerships for not sufficiently complying, in its view, with the guidance.
Moreover, the Home Secretary has chastised local authorities and the police for
under-utilising certain powers given to them by the CDA, such as the contro-
versial anti-social behaviour orders (ASBOs) and youth curfews.3 At the Local
Government Association conference in Bournemouth in June 2000, Jack Straw
criticised councils for not making greater use of ASBOs. He claimed ASBOs to
constitute a ‘powerful weapon’ which ‘should be used swiftly where circum-
stances demand it, not just against the very hard cases of unacceptable
behaviour’ (Independent, 29 June 2000). Moreover, he suggested that councils
should ignore concerns over conflicts with the recent Human Rights Act
(which incorporates the European Convention on Human Rights into UK law)
commenting that ‘no drunk on the street who has his can of Tenants lager
taken off him is going to take us to the court in Strasbourg’ (The Times, 29 June
2000). This ‘hands on’ central government interventionism runs counter to the
spirit of locally sensitive and problem-oriented approaches adopted by commu-
nity safety partnerships.
Networks and partnership arrangements require ‘hands off’ management and
‘governance at a distance’ which are often undermined by the practice of inter-
vention on the part of government. This can produce significant conflicts and
tensions between local initiatives and central steering. The reality of trying to
separate purchaser/provider functions has been much more complex than
managerialist gurus suggest, often fusing and blurring roles and responsibilities,
and producing forms of government which are simultaneously ‘at arm’s length
but hands on’.
 Joined-up but fragmented 65

Wide-angled but tunnelled vision

The innovation of managerial reforms in many ways runs counter to the elabo-
ration of a partnership approach. As one political commentator notes, these
reforms:

had one overriding failure – they left government less, rather than more,
able to solve the important, ‘wicked’ problems that most concern electors:
how to cut crime. … The vertical links between departments and agencies
in any one field and professional groups such as the police, teachers, doctors
and nurses are strong. The horizontal links are weak or non-existent.
(Perri 6 1997: 9–10)

Managerialist reforms have served, perversely, to increase the isolation and intro-
spection of many criminal justice agencies and other public sector organisations
drawn into community safety partnerships. They have done so in four principal
ways. First, managerialist reforms focus upon efficiency, economy and value for
money, on hierarchical control and on the clear distribution of authority and
responsibility (Rhodes 1997: 55). They are primarily concerned with internal re-
organisation. Managerialist reforms encourage an intra-organisational focus that
pays little attention to the task of managing inter-organisational relations. Scant
regard is afforded to the more complex process of negotiating shared purposes,
particularly where there is no hierarchy of control. Managerialism may suit hier-
archical line management structures but is largely inappropriate for managing
horizontal inter-organisational networks. There are fundamental tensions
between a ‘partnership’ approach and aspects of the intra-organisational focus of
managerialist reforms.
Second, the intra-organisational focus on ‘outputs’ and performance
measurement can make agencies concentrate their energies upon their core
tasks and activities at the expense of peripheral ones. Managerialist reforms
focus attention upon narrowly construed service delivery to ‘customers’ of a
particular segment of criminal justice at a given time and place within the
process, rather than upon cross-cutting, horizontal accountabilities and respon-
sibilities. Where a partnership approach envisages corporate decision-making
on behalf of a number of actors, managerialism conceives of compartmentalised
and independent responses coupled with an intra-organisational focus on objec-
tives and results. The emphasis upon intra-organisational performance
measurement fails to recognise that partnerships inherently blur the boundaries
between, and the roles of, the incorporated organisations, such that individual
contributions cannot easily be separated off and disentangled from each other
without undermining the nature of that contribution.
The recent development of ‘corporate’ performance indicators has gone
some way towards tying diverse agencies into the production of collective
goods which transcend the competency or capacity of individual agencies –
such as community safety. Yet, this does not resolve two dilemmas. First, there
are problems that arise with regard to any conflicts between intra-organisa-
66 Adam Crawford

tional and inter-organisational performance indicators and where priorities

should lie. Second, there are significant problems with regard to account-
ability for corporate performance indicators, given the very nature of
partnerships, their multiple layers of authority and complex interdependency.
Joint and negotiated decisions tie the various parties into corporate policy and
outcomes but often fail to identify lines of responsibility. Institutional
complexity further obscures who is accountable to whom, and for what. This
gives rise to what Rhodes has identified as ‘the problem of many hands, where
so many people contribute that no one contribution can be identified; and if
no one person can be held accountable after the event, then no one needs to
behave responsibly beforehand’ (1996: 663). As authority is ‘shared’ it
becomes difficult to disentangle and can become almost intangible. As Peters
comments:

Under the present conditions authority can hardly be located any more: it
has become shared, parcelled, delegated, conditioned, subjected to approval
and correction … In a word: authority has become elusive.
(Peters 1986: 34)

Third, managerialist reforms place a growing emphasis upon organisationally

defined outputs as distinct from outcomes. ‘Outputs’ are service activities
whereas ‘outcomes’ are the impacts or consequences (intended or unintended)
of these outputs on the wider environment. ‘Outputs’ frequently refer to inter-
nally defined organisational goals over which organisations have considerable
control. These may depart significantly from ‘outcomes’ – the effects of an
output, or set of outputs, upon the community. Hence, output measurement and
outcome evaluation are not the same thing. Given the control that organisa-
tions can assert over defining their own outputs there are questions to be asked
about the validity of output measurement as a central aspect in the monitoring
of community safety. The concern is that the gulf between the two may grow
larger under managerialist pressures. There is a danger that ‘outputs’ may take
precedence over ‘outcomes’, so that social goals are eclipsed by organisational
ones. This can express itself as ‘measure fixation’ whereby greater concentration
is given to the measure, rather than the service which the measure is intended
to signify.
Fourth, while it is clear that managerialist reforms have generated much
more information about public organisations and services and the way in
which they work, the quality, value and utility of the information produced
remain uncertain. There is a danger that the application of performance
cultures to public sector organisations drawn into community safety partner-
ships focuses attention upon ‘target management’ rather than upon the quality
of performance itself. This can produce the unwelcome dangers of ‘gaming’
through the strategic management of information as well as the manipulation
and obfuscation of data, all of which can result in the demoralisation of staff
and public cynicism (see Tilley 1995). More fundamentally, it can encourage a
 Joined-up but fragmented 67

concentration upon ‘easy’ or ‘soft’ targets: those which are more likely to yield
positive results, consequently diverting attention and resources away from (and
possibly writing off all together) more problematic areas or intractable crime
problems.
In sum, managerialist reforms are essentially hierarchical and as such antithet-
ical to the management of complex networks: they emphasise inward-looking
and myopic forms of performance measurement and output fixation which often
work against holistic governance. As a consequence, they often fail to address
‘wicked issues’, focusing upon narrowly defined, measurable elements of a
problem. They are simultaneously wide-angled but tunnelled vision.

Growing demands for trust and the institutionalisation of

distrust
The construction and maintenance of trust, as I have argued elsewhere
(Crawford 1997: 112–23; 1998a: 179–80), is a central dynamic in effective
community safety partnerships. This is particularly evident where structural
conflicts exist, where partners are steeped in divergent cultural ideologies and/or
where there is a legacy of mistrust or misunderstanding. Fostering relations of
trust between partner agencies, as well as among and between lay representa-
tives, can take time and the investment of considerable effort. As such,
partnerships have ‘life-spans’ during which they adapt and change. They need
to develop and mature as circumstances alter. As trust develops and partner-
ships become more established different ways of managing conflict may become
more appropriate. A crucial element in establishing trust relations is making
people aware of the limitations of their own and other agencies’ contribution,
such that they neither try to ‘do it all’ nor do they have unrealistic expectations
of what others can deliver. Consequently, there is a need for mutual respect for
different types of contributions.
The crucial nature of inter-organisational trust is increasingly recognised
within government guidance and by influential policy bodies such as the Audit
Commission. In a management paper, the logics of which inform the Safety in
Numbers Report, the Audit Commission declared: ‘For most partnerships,
building trust between partners is the most important ingredient in success’
(1998: 26, para. 69). And yet, the wider context of managerialist reforms have
tended to problematise trust, most notably trust in professional expertise.
Managerialist reforms have introduced various ‘rituals of verification’: contracts,
performance indicators, audits and inspections. Most notably, as Michael Power
(1997) has shown, the audit has become a central instrument in the transforma-
tion of modes of governance and control. As the state has redrawn its functions
and become increasingly committed to a supervisory role, audit and accounting
practices have assumed a more prominent place. The ‘hollowing out’ of the
state by NPM generates a demand for audit and other forms of inspection to fill
the void. As Power notes:
68 Adam Crawford

Audit has become a benchmark for securing the legitimacy of organisa-

tional action in which auditable standards of performance have been
created not merely to provide for substantive internal improvements to the
quality of service but to make these improvements externally verifiable via
acts of certification. As the state has become increasingly and explicitly
committed to an indirect supervisory role, audit and accounting practices
have assumed a decisive function. The state cannot play this indirect role
without assuming the efficacy of these practices at the foot of a regulatory
hierarchy.
(Power 1997: 10–11)

Despite its technical discourse, the audit is also a system of values and goals
which are inscribed in the official programmes which both demand and
encourage it. It helps shape public conceptions of the problems for which it is
the perceived solution. As such, ‘it is constitutive of a certain regulatory or
control style which reflects deeply held commitments to checking and trust’
(ibid.: 7). Audit is not merely a technical solution to a problem: ‘it also makes
possible ways of redesigning the practice of government’ (ibid.: 11). Audits have
as their primary object organisations and their sub-systems of control, as such
audits are concerned with the control of control (ibid.: 128).
It is unsurprising, therefore, that the community safety partnerships have, at
their very heart, a process of audit. In many senses, this process differs from
traditional financial audits, but nevertheless the core idea of independent
checking through informational evidence remains a key factor. Moreover, the
audit process incorporates wholesale the language of managerialism, thus
constructing afresh the nature of local crime problems: how they should be
measured and managed. It seeks to shape public expectations, redefine success
and failure and construct new forms of authority and legitimacy.
Appeals to ‘community participation’, ‘active citizenry’ and ‘consumer
sovereignty’ also imply a scepticism of professional expertise. They feed off
critiques of ‘self-serving professional élites’ and claim to represent a challenge to
the ‘cosy cultures of professional self-regulation’ (Power 1997: 44). Here lies an
appeal to open up a new circuit of power between the demands of communities
and individuals as against those of criminal justice agencies. Where once we
were told to ‘leave it to the professionals’ now we are enjoined to active partici-
pation in a ‘self-policing society’ (Leadbeater 1996). As concerns over
‘vigilantism’ have demonstrated, this circuit of power is a problematic and
volatile one.
Paradoxically, technologies such as audits and inspections are contrived, in
part, in the hope of restoring trust in organisational competence. Moreover,
they presuppose and necessitate trust in themselves as instruments of verifica-
tion and control. And yet, rather than resolving trust deficits they tend to
multiply and disperse these into the fabric of the organisational environment.
They disrupt trust relations by replacing traditional forms of trust in the profes-
sional use of discretion with forms of audit and inspection. Relations of trust in,
 Joined-up but fragmented 69

and between, functionaries which have traditionally allowed large scope for
policy discretion have been significantly curtailed in recent years by managerial
reforms. They have encircled trust, by reducing the boundaries of discretion and
rendering it subject to inspection and budgetary constraint.
As such, managerialist reforms institutionalise forms of distrust. However,
recourse to audits, performance measurement, contracts and inspections does
not necessarily destroy trust – as some have argued (Broadbent et al. 1996) – but
rather reconfigures it. Trust and the contractualisation of relations are not
mutually exclusive. In many senses, contracts presuppose a certain degree of
trust. The new governance does not obliterate trust altogether, in fact it necessi-
tates that trust is placed in new ‘guardians of trust’: auditors and inspectors. But
this is a different form of trust, involving a ‘second order’ relationship.
Ultimately, however, these systems of control are themselves the subject of
questions as to their competence. The trust that new technologies demand must
itself be the subject of ‘verification’ (Crawford 2000b).
Whilst it is the government’s explicit intention that the audit process should
be a developmental one, in that the Crime and Disorder Act (1998) establishes
a strategy cycle in which the audit process is revisited every three years, there
are legitimate concerns that in practice this may be undermined by competing
pressures. These arise in large part because the veracity of the audit process
becomes bound up with target setting and performance measurement. It would
seem that the quality of the data which informed the first round of audits
published in early 1999 was both limited and variable. Nevertheless, the very
process of striving to meet the targets set on the basis of the audit baseline lends
a legitimacy to the audit itself and shrouds it from critical gaze. For, to question
the baseline is to question the process, and hence the very purpose of measure-
ment. Rather than providing a foundation for informed dialogue and
negotiation, the audit process demands that its efficacy is trusted. The concern
is that in the ‘audit society’ the strategy cycles will become less a process of
institutional learning and problem-identification or merely the basis for discus-
sion but more, in Power’s words, ‘ritualised practices of verification whose
technical efficacy is less significant than their role in the production of organi-
sational legitimacy’ (Power 1997: 14).
Moreover, the reality is that, once set up, baselines and audit processes exert
a dynamic of their own. Continuity and comparability often rein back innova-
tion and development. These problems are compounded by difficulties of
measurement and the inadequacies of information available. Setting meaningful
targets and determining performance indicators for dealing with ‘disorder’ or
‘anti-social behaviour’ are inherently problematic. Public definitions of disorder
and community safety are inconsistent. Different audiences define the same
behaviour differently. Furthermore, many of the neighbourhoods with high
levels of crime and incivilities are inscribed by a general lack of consensus about
such issues. In England and Wales, the Crime and Disorder Act (1998) provides
no statutory definition of disorder nor is it defined in the accompanying
Guidance (Home Office 1998). In practice, no standard systems for monitoring
70 Adam Crawford

anti-social behaviour exist. In particular, there is no formal process for making

decisions about what counts as an instance of anti-social behaviour.
Consequently, the increased monitoring of non-criminal anti-social activity –
through nuisance-related calls to the police and complaints to housing and
environmental health departments and the development of more sophisticated
methods of measurement – may actually result in an increase in recording.
Expanding the focus of official regulation into non-criminal incidents,
perversely, may produce an apparent increase in the phenomenon itself, at least
in the short to medium term.
As Mollie Weatheritt noted some time ago with regard to performance
measurement in policing, that it

is primarily a way of formulating and asking further questions and

promoting a dialogue with service providers about the prospects for service
improvement. That dialogue may well turn out to be no more than a sterile
tussle over disputed meanings.
(Weatheritt 1993: 41)

In sum, managerial reforms exhibit factors which make trust more valuable
and simultaneously render it more difficult to sustain. They corrode trust and
encourage distrust at the moment in which trust is perceived to be a precious
commodity – as the basis for communal self-regulation, managing networks and
inter-organisational negotiation – such that the increasing demand for trust
coexists alongside the institutionalisation of distrust. Moreover, the new tech-
nologies which seek to respond to trust deficits themselves demand that their
efficacy is trusted, whilst generating information which is of limited use.

Co-operation and negotiation in a cold climate of competition

Since its inception in 1982, the Audit Commission has played a central role in
the spread of managerialist ideas throughout public sector organisations, with a
particularly significant impact on the field of crime control and policing. Along
with the National Audit Office and Her Majesty’s various Inspectorates, the
Audit Commission has been crucial in encouraging the spread of objective-
driven performance and generating a performance conscious culture, most
notably within senior management, but ultimately throughout public sector
organisations.4 This ‘value for money’ approach has encouraged competition
and comparison within and between public sector agencies.
In mid-1997, after the publication of the highly influential report into the
youth justice system, Misspent Youth (Audit Commission 1996), the Audit
Commission announced an eighteen-month national investigation and analysis
of ‘community safety’. The subsequent report Safety in Numbers (Audit
Commission 1999) brings the Audit Commission’s influential version of
managerialism to the heart of community safety debates. In place of the rather
crude totem of ‘value for money’ it advances the notion of ‘best value’ as a key
 Joined-up but fragmented 71

element of managerial accountability. The logics of the report correspond with,

and are inscribed into, government policy and guidance (see the Local
Government Act 1999). The Audit Commission (1999: 23) defines ‘best value’
as requiring:

• challenging authorities to be clear and open on their service objectives,

• consulting local people on these objectives,
• comparing performance with other authorities,
• competing – demonstrating that services are competitive against alternative
providers in the market place.

Hence, competition and comparison are seen as key criteria through which to
require authorities to demonstrate improvements in services. And yet, unlike
the first two criteria they appear particularly problematic with regard to commu-
nity safety partnerships.
First, there is a contradiction between competition between service
providers, on the one hand, and the process of managing networks of service
providers – notably on the basis of negotiated equilibrium – on the other hand.
Managerialist reforms by creating quasi-markets and injecting competition into
community safety can serve to produce new sites of inter-agency conflict
(Raine and Willson 1997: 88). They encourage low levels of interdependence,
which may leave inter-organisational networks unstable. Moreover, competi-
tion can serve to undermine both reciprocity and trust. The cultures conducive
to managing collaborative partnerships and constructive networks are rooted
more in the identification of shared values, mutual trust and the respect of
common principles and diverse contributions than in competitive market
mentalities or managerial performance comparison. The skills necessary for co-
ordinating and steering horizontal networks are more concerned with
enhancing commitment through management by negotiation. Rhodes (2000:
355–6) likens such skills to those of ‘diplomatic virtues’ which are far removed
from those of the ‘head-kicker macho-manager’, the ‘individualist entrepreneur’
or the ‘bureaucrat’.
Second, comparison as an element of competition can introduce negative
dynamics into community safety given the capacity of crime and disorder to
divide populations and excite fears and anxieties. Moreover, there are signifi-
cant difficulties in drawing meaningful comparisons given the very different
nature of local crime and disorder problems in differing communities. Certain
forms of crime and associated problems are heavily concentrated in specific
geographic areas. Furthermore, there are dangers that community safety is not
a rationalistic ‘zero sum’ endeavour. Safety is relative and relational: one area’s
sense of security and ‘place’ within the city, may be defined, in large part, in
contradistinction to the perception of other areas’ insecurity. In other words,
to feel safe in your own area may necessitate the identification of other, less
safe areas or urban ‘badlands’. Hence, competition and comparison in commu-
nity safety may serve to fuel dynamics of social exclusion and ghettoisation.
72 Adam Crawford

Comparison, in a context of appeals to community as a force for moral

order and social control, can promote a disaggregating social force. An
emphasis upon comparison can encourage boundary disputes and the forma-
tion of new communities around defensiveness and exclusivity, increasing
social polarisation. There are dangers that contemporary quests for security
will encourage the flight of people and capital out of certain localities as they
become associated with, and stigmatised by, high levels of recorded crime and
disorder. This can produce increasing social dislocation of a fundamentally
spatial nature, in which disorder is a dynamic in a vicious spiral of decline and
‘market residualisation’. The concern is that ‘security differentials’, fuelled by
crude area-based comparisons, may become increasingly significant character-
istics of wealth and status as communities solidify around defensive exclusivity
(Crawford 1997).
This can be further exacerbated by the central paradox in crime prevention:
the existence of an often inverse relationship between activity and need with
regard to community crime prevention. Years of research have shown that
community responses to crime are easiest to generate in exactly those areas
where they are least needed and hardest to establish where the need is greatest.
This paradox is compounded by problems of spatial displacement. As such, the
role of community as a force for social cohesion can be undermined by the
increasing extent of social and geographic dislocation, as a result of which the
wealthy retreat into disaggregated, defended communities of affluence and with-
draw from participation in, and contribution to, public services. In this context,
the image of communities as exclusive islands of self-regulation raises concerns
about the relationship between community and social justice. Partnerships, in
this context need to address the vexed question: to what extent is local commu-
nity safety a public good rather than a club good serving the interest of its
members?
If community safety partnerships are to play a role in reconstructing the
institutional bedrock of inner-city communities by assisting in building institu-
tions of social solidarity, governments – both central and local – will need to
provide the conditions under which genuine partnerships and networks can
flourish. This requires policies and an infrastructure which foster reciprocity
and interdependence between organisations and interests, not insularity and
competition. The challenge for government is to cultivate such conditions and
to nurture new forms of co-operation and shared values, rooted in mutual
acceptance of difference and reciprocity on the basis of inter-organisational
trust.
Hence, (quasi-)markets, competition and crude comparisons can inject
pernicious dynamics into inter-community and inter-agency relations. They can
encourage boundary disputes and the formation of new communities around
defensiveness and exclusivity, increasing social dislocation and polarisation. As
such, environments of co-operation and negotiation, so necessary for construc-
tive partnership and inter-community relations, must exist in an unfavourable
cold climate of competition.
 Joined-up but fragmented 73

Nostalgia disguised as modernisation

Inherent in much of the appeal to community in Third Way politics, under the
communitarian influence, is a nostalgia which harks to a different age of tradi-
tion and authority. At a psychic level, it allows people to delve into the
nostalgic ‘imagined communities’ of tradition. Underlying policy initiatives
around community crime prevention is the prevailing idea that crime results
from a failure or breakdown of community life. Appeals to community assume
that there is a direct relationship between a lack of ‘community’ and the exis-
tence of high levels of crime. Disorganised communities – which lack a sense of
mutuality and cohesiveness, where people are unwilling to intervene in support
of communal order as ‘no one cares’ – are associated with high levels of crime
and, inversely, low crime areas are associated with well-organised and cohesive
communities. Yet, contemporary communities are not synonymous with social
order. ‘Organised’ communities can produce both disorder and foster high levels
of crime. Collective values of a community may serve to stimulate and sustain
criminality. Communal values themselves can be crimogenic. Moreover, strong
communities can be pockets of intolerance and prejudice. Nevertheless, appeals
to community retain a profound emotional legitimacy in that they hold out the
fantasy of genuine human identity, connectedness and reciprocity precisely at a
moment in time at which they appear most absent.
This ‘wilful nostalgia’ (Simon 1995) has been particularly notable in the
debate about ‘zero tolerance’ policing. The idea of ‘zero tolerance’ – warmly
embraced by the New Labour government – offers a strategy through which to
re-assert sovereignty, impose discipline and order and reclaim the streets from
the deviant (Dennis 1997). ‘Zero tolerance’ policing evokes a nostalgic reasser-
tion of moral authority through more aggressive and assertive strategies (see
Bratton 1997). Wilson and Kelling’s (1982) ‘Broken Windows’ thesis, from
which zero tolerance draws its intellectual legitimacy, implicitly evokes a
nostalgia for a secure and consensual world of the past. Signs of disorder are the
enemy of community. Disorder, it is argued, violates a community’s expectations
of what constitutes appropriate civil behaviour (see Kelling in this volume).
What is deemed necessary, therefore, is for the community to reassert its
‘natural forces’ of authority and control. The role of the police and other formal
agencies is to support a community’s political and moral authority. To do so, the
police must accommodate those ‘natural forces’.
This nostalgic re-imagining of community appears to run in stark contrast to
the ‘modernisation offensives’ of New Labour and Third Way politics, for whom
‘modernisation’ stands as a short-hand for the need to embrace the ‘inevitable’
consequences of contemporary (global) conditions. Furthermore, the settled
and stable communities idealised by communitarianism are often the enemies of
innovation and experimentation and offer little space for creativity and diver-
sity. Community as a force for conforming can serve to undermine the quest for
novel, non-hierarchical modes of policy-making. There is a fundamental differ-
ence between restoring communities and transforming them into innovative
and democratic institutions (Crawford and Clear 2001). Communities are not
74 Adam Crawford

the utopias of egalitarianism, which some might wish, but are hierarchical
formations, structured upon lines of power, dominance and authority.
Communities are intrinsically exclusive – as social inclusion presupposes
processes of exclusion – and may solidify and define themselves around notions
of ‘otherness’ potentially infused with racialised overtones. Challenging and
disrupting established community order, its assumptions and power relations
may be a more fundamental aspect of a progressive politics. As such, nostalgic
appeals to community as the basis of moral order may often be incongruous with
the radical transformations necessary to realise modern institutions of gover-
nance through partnerships.
In short, it is debatable the extent to which the Third Way quest for commu-
nity is a reworking of tradition or a force for democratic renewal. Rather, it may
simply constitute nostalgia disguised as modernisation.

Ambivalent political responses: moral scolding, punitive

populism and the rationalistic urge to manage
Finally, there are deep tensions within New Labour’s policies regarding the rela-
tionship between, on the one hand, developments in community safety and the
preventive strategies associated with them and, on the other hand, the growth of
‘populist punitiveness’ through an increased resort to strategies of penal exclusion
(Bottoms 1995: 39–41). These tensions are apparent in the ‘tough on crime, tough
on the causes of crime’ catchphrase and operate at a number of different levels.
First, the (re)moralising urge of communitarianism runs in conflict to the
administrative impulse of managerialism. Where the former calls for a new
moral crusade upon which to reconstruct civil society, the latter hails adminis-
trative efficiencies and cost effectiveness. There are significant tensions
between the amoral market place of competition as the dynamic engine of
change and the moral authority at the centre of communitarianism: between
the instrumental and the moral elements of control. This ambiguity is seen, on
the one hand, in the punitive tone and expressive ‘moral scolding’ which insists
upon individual and family responsibility for offending and, on the other hand,
in the rationalistic desire to manage crime at a minimal cost (Brownlee 1998).
These contrasting visions awkwardly cohabit the Crime and Disorder Act
(1998) and associated policy strategies.
Second, the normalisation and de-dramatisation of crime co-exist with the
continued demonisation of cultural scapegoats: crime is both an ordinary part of
everyday life to be the subject of routine precautions, avoided and prevented as
well as a metaphor for urban and civic decline. At one moment the offender is a
self-maximising and autonomous rational actor no different in his/her choices
from the rest of us, the next minute, the offender is a depraved member of an
‘underclass’ operating within a wholly different moral frame. The tone shifts
from an emphasis upon ‘de-differentiation’ – namely that crime is a usual and
normal aspect of modern life and that criminals are essentially ‘like us’ – to a
pre-occupation with pathological differences and ‘otherness’.
 Joined-up but fragmented 75

David Garland (1996; 1999) relates this duality of ‘the commonplace and
the catastrophic’ to much deeper (structural) interpretations of crime in late
modernity, as a product of a decentring of state governance of crime. These two
visions Garland refers to as the ‘criminology of the self’ and the ‘criminology of
the other’ respectively:

One criminology de-dramatises crime, allays disproportionate fears, and

promotes routine preventative action. The other demonises the criminal,
arouses popular fears and hostilities, and strives to enlist support for drastic
measures of control.
(Garland 1999: 354)

Not only are these antagonistic accounts the product of dilemmas of contempo-
rary state governance, as Garland suggests, but they also find a particular
expression in ambiguities within Third Way politics.
This tension helps to explain the moral vacillation in the apparently contra-
dictory processes of ‘defining deviancy down’ (Moynihan 1993) and ‘defining
deviancy up’ (Krauthammer 1993) which are simultaneously found in the
Crime and Disorder Act (1998). In one instance, we see a normalisation of
‘crime as everyday life’, whereby in response to high crime rates government
relaxes notions of deviance and allows previously deviant behaviour to become
‘acceptable’ or even ‘normal’, thus limiting the level of demand placed upon
criminal justice systems. This is to be found in governmental strategies through
the spread of situational crime prevention whereby individuals are increasingly
responsible for their own personal safety and the security of their possessions. In
another instance, the increasing concern for incivilities, anti-social behaviour,
disorder and quasi- or sub-criminal activities constitute an inverse process of
‘defining deviancy up’. Here, the previously ‘normal’ is declared deviant and the
deviant is unmasked residing within the normal. Perversely, this can have the
consequence of increasing the demand for crime control and judicial interven-
tions. Janus-faced, the Third Way expresses two distinct visions of crime, the
criminal and the role of criminal justice.
More fundamentally, these tensions express themselves in ambivalent polit-
ical responses. Strategies of responsibilisation through appeals to community
involvement and plural partnerships beyond the state co-exist with declarations
of sovereign authority, notably through punitive sentiments as evidence of the
state’s commitments and strengths. However, this is also to be found with regard
to policing. At one moment, crime rates are portrayed as beyond the compe-
tence of the police alone, as under the new rhetoric ‘we are all responsible for
crime’. In a different moment, however, the government as part of its Crime
Reduction Strategy has set targets for the police, notably with regard to
domestic burglary, to reduce crime by certain measurable amounts within a
certain time-span, as if the police’s capacity for action in this field were
somehow restored. As Garland argues, state sovereignty over crime is simultane-
ously denied – as being ‘beyond the state’ – and symbolically reasserted –
76 Adam Crawford

through periodic episodes of hysterical and populist denials of the state’s limita-
tions (1996: 462). The limitations of traditional criminal justice – police,
prosecution and punishment – are recognised in certain instances only to be
discounted or ignored in others. This dualistic denial and recognition produce
volatile shifts in the state’s presentation of its own capacity for effective action
in crime control.
This ambivalence also expresses itself in the uncertain integration of victims
within criminal justice (Crawford and Goodey 2000). The victim has become a
pivotal figure within contemporary criminal justice and yet victim input
remains circumscribed and ambiguous. The recent concern for victims, their
rights and needs, has found inconsistent political expression. Victims needs are
concurrently embraced and ignored, as illustrated by the integration of victims
within the post-sentencing process through victim contact work by the proba-
tion service (see Crawford and Enterkin 1999). The interests of victims have
simultaneously been used in the service of severity, in the service of offenders
and in the service of systems’ efficiency (Crawford 2000a: 292). The awkward
position of the victim within the complex balance between state/offender and
public/private interests expresses itself in the uncertain relation between passion
and reason as well as between moral punitiveness and instrumental offender
management in Third Way politics.
On Garland’s (1996) account, punitive rhetoric and policy are as much a
product of problems of state sovereignty and legitimation as they are a rational
response to the problems of crime. And yet, he implies that this oscillation and
ambivalence are a product of late modern conditions which produce dilemmas
for state governance. Whilst this broader canvas is insightful and helps us locate
the manner in which similar issues affect different societies it does not accord
sufficient attention to the different cultural and political expressions that these
contradictions can take. The manner in which ambivalence is played out in the
policy field takes on a political form. Hence, these contradictions are also a
product of ‘political incoherence’ (O’Malley 1999), notably, although not
exclusively, the ‘unholy Trinity’ – of appeals to communitarian ideals, gover-
nance through partnership and managerialist implementation – at the heart of
Third Way politics. As I have argued, it is through the analysis of these more
specific aspects of political incoherence that we can identify the sites of contra-
diction, ambiguity and ambivalence around which the future shape and
direction of community safety is likely to unfold.

Notes
1 I borrow this phraseology from Taylor’s (1997) article of the same title.
2 An alternative and altogether more cynical interpretation of this ambiguity lies in
the unwillingness of central government to be explicitly too prescriptive in order to
avoid the counter-accusations from local government, police and other agencies that
if central government wish to require them to engage in specified tasks then dedi-
cated funding should be made available, in order to fulfil such a requirement. The
Crime and Disorder Act (1998) whilst imposing a statutory duty upon local authori-
 Joined-up but fragmented 77

ties and the police to establish community safety partnerships did not provide any
new or dedicated resources.
3 In the first fourteen months since they came into effect in April 1999, there had
been no youth curfews implemented and only eighty ASBOs taken out in England
(The Times 29 June 2000). This low use rate has been perceived to be problematic by
the government, which is strongly encouraging greater usage (SEU 2000: 41). The
government now believes that ASBOs should not be viewed as a method of ‘last
resort’, which appears to depart significantly from their initial intentions (Labour
Party 1995).
4 The status and role of bodies like the Audit Commission in recent years have
become increasingly ambiguous, albeit with considerable influence. Its work often
bears little correspondence with traditional depiction of the word audit. Its role
frequently appears more consistent with that of a catalyst for organisational change.
As Day and Klein (1990) suggest, the Audit Commission is in the position of ‘a
policeman constantly tempted to turn consultant’.

References
Audit Commission (1996) Misspent Youth: Young People and Crime, London: Audit
Commission.
—— (1998) A Fruitful Partnership: Effective Partnership Working, London: Audit
Commission.
—— (1999) Safety in Numbers: Promoting Community Safety, London: Audit Commis-
sion.
Blair, T. (1995) ‘The Rights We Enjoy Reflect the Duties We Owe’, The Spectator
Lecture, 22 March.
—— (1996) New Britain: My Vision of a Young Country, London: Fourth Estate.
—— (1998) The Third Way: New Politics for the New Century, London: Fabian Society.
Bottoms, A. E. (1995) ‘The Philosophy and Politics of Punishment and Sentencing’, in
C. Clarkson and R. Morgan (eds) The Politics of Sentencing Reform, Oxford:
Clarendon, 17–49.
Bratton, W. J. (1997) ‘Crime is Down in New York City: Blame the Police’, in N. Dennis
(ed.) Zero Tolerance: Policing a Free Society, London: Institute for Economic Affairs, 29–42.
Broadbent, J., Dietrich, M. and Laughlin, R. (1996) ‘The Development of Principal-
Agent, Contracting and Accountability Relationships in the Public Sector:
Conceptual and Cultural Problems’, Critical Perspectives in Accounting, 7(4): 259–84.
Brownlee, I. (1998) ‘New Labour – New Penology?: Punitive Rhetoric and the Limits of
Managerialism in Criminal Justice Policy’, Journal of Law and Society, 25(3): 313–35.
Cabinet Office (1999) Modernising Government, Cm 4310, London: HMSO.
Charlesworth, J., Clarke, J. and Cochrane, A. (1996) ‘Tangled Webs? Managing Local
Mixed Economies of Care’, Public Administration, 74: 67–88.
Clarke, J. and Newman, J. (1997)The Managerial State, London: Sage.
Crawford, A. (1997) The Local Governance of Crime: Appeals to Community and Partner-
ships, Oxford: Clarendon.
—— (1998a) Crime Prevention and Community Safety: Politics, Policies and Practices,
Harlow: Longman.
—— (1998b) ‘Community Safety and the Quest for Security: Holding Back the
Dynamics of Social Exclusion’, Policy Studies, 19(3/4): 237–53.
—— (2000a) ‘Salient Themes Towards a Victim Perspective and the Limitations of
Restorative Justice’, in A. Crawford and J. Goodey (eds) Integrating a Victim Perspec-
tive within Criminal Justice, Aldershot: Ashgate, 285–304.
78 Adam Crawford

—— (2000b) ‘Situational Crime Prevention, Urban Governance and Trust Relations’,

in A. Von Hirsch, D. Garland and A. Wakefield (eds) Ethical and Social Perspectives
on Situational Crime Prevention, Oxford: Hart Publishing, 193–213.
—— and Clear, T. R. (2001) ‘Community Justice: Transforming Communities Through
Restorative Justice?’, in G. Bazemore and M. Shiff (eds) Restorative Community
Justice: Repairing Harm and Transforming Communities, Cincinnati: Anderson Publica-
tions, 127–49.
—— and Enterkin, J. (1999) Victim Contact Work and the Probation Service: A Study of
Service Delivery and Impact, Leeds: CCJS Press.
—— and Goodey, J. (2000) (eds) Integrating a Victim Perspective within Criminal Justice,
Aldershot: Ashgate.
Day, P. and Klein, R. (1990) Inspecting the Inspectorates, London: Joseph Rowntree Foun-
dation.
Deakin, N. and Walsh, K. (1996) ‘The Enabling State: The Role of Markets and
Contracts’, Public Administration, 74: 33–48.
Dennis, N. (1997) (ed.) Zero Tolerance: Policing a Free Society, London: Institute for
Economic Affairs.
Downes, D. (1998) ‘Toughing it Out: From Labour Opposition to Labour Government’,
Policy Studies, 19(3/4): 191–8.
Downes, D. and Morgan, R. (1997) ‘Dumping the Hostages to Fortune?: The Politics of
Law and Order in Post-war Britain’, in M. Maguire, R. Morgan and R. Reiner (eds)
The Oxford Handbook of Criminology, Oxford: Clarendon Press, 87–134.
Etzioni, A. (1993) The Spirit of Community: Rights, Responsibilities and the Communitarian
Agenda, New York: Simon & Schuster.
—— (1997a) The New Golden Rule: Community and Morality in a Democratic Society,
New York: Harper Collins.
—— (1997b) ‘Tony Blair: A communitarian in the making’, The Times, 21 June.
—— (1997c) ‘Community watch’, Guardian, 28 June.
Garland, D. (1996) ‘The Limits of the Sovereign State: Strategies of Crime Control in
Contemporary Society’, British Journal of Criminology, 36(4): 445–71.
—— (1999) ‘The Commonplace and the Catastrophic: Interpretations of Crime in Late
Modernity’, Theoretical Criminology, 3(3): 353–64.
Giddens, A. (1991) The Consequences of Modernity, Cambridge: Polity Press.
—— (1998) The Third Way: The Renewal of Social Democracy, Cambridge: Polity Press.
—— (2000) The Third Way and Its Critics, Cambridge: Polity Press.
Hines, J., Holdaway, S., Wiles, P., Davidson, N., Dignan, J., Hammersley, R. and Marsh,
P. (1999) Interim Report on Youth Offending Teams, London: Home Office.
Home Office (1997) Getting to Grips with Crime: A New Framework for Local Intervention,
London: Home Office.
—— (1998) Guidance on Statutory Crime and Disorder Partnerships, London: Home
Office.
Hood, C. (1991) ‘A Public Management for all Seasons?’, Public Administration, 69: 3–19.
—— (1995) ‘ ‘‘The New Public Management’’ in the 1980’s: Variations on a Theme’,
Accounting, Organisations and Society, 20(2/3): 93–110.
Jones, T. and Newburn, T. (1998) Private Security and Public Policing, Oxford: Clarendon
Press.
Krauthammer, C. (1993) ‘Defining Deviancy Up’, The New Republic, 22 November:
20–5.
 Joined-up but fragmented 79

Labour Party (1995) A Quiet Life: Tough Action on Criminal Neighbours, London: Labour
Party.
Leadbeater, C. (1996) The Self-Policing Society, London: Demos.
Morgan, J. (1991) Safer Communities: The Local Delivery of Crime Prevention Through the
Partnership Approach, London: Home Office.
Moss, K. and Pease, K. (1999) ‘Crime and Disorder Act 1998: Section 17 a Wolf in
Sheep’s Clothing?’, Crime Prevention and Community Safety, 1(4): 15–19.
Moynihan, D. P. (1993) ‘Defining Deviancy Down’, American Scholar, 62, Winter:
17–30.
Newburn, T. (1998) ‘Tackling Youth Crime and Reforming Youth Justice: The Origins
and Nature of New Labours Policy’, Policy Studies, 19(3/4): 199–212.
O’Malley, P. (1999) ‘Volatile and Contradictory Punishment’, Theoretical Criminology,
3(2): 175–96.
—— and Palmer, D. (1996) ‘Post-Keynesian Policing’, Economy and Society, 25(2):
137–55.
Osborne, D. and Gaebler, T. (1992) Reinventing Government: How the Entrepreneurial
Spirit is Transforming the Public Sector, Reading, Massachusetts: Addison-Wesley.
Perri 6 (1997) Holistic Government, London: Demos.
Peters, A. A. G. (1986) ‘Main Currents in Criminal Law Theory’, in J. M. van Dijk, C.
Haffmans, F. Ruter and J. Schutte (eds) Criminal Law in Action: An Overview of
Current Issues in Western Societies, Arnhem: Gouda Quint, 19–36.
Power, M. (1997) The Audit Society, Oxford: Oxford University Press.
Raine, J. W. and Willson, M. J. (1997) ‘Beyond Managerialism in Criminal Justice’,
Howard Journal, 36(1): 80–95.
Rhodes, R. A. W. (1996) ‘The New Governance: Governing without Government’,
Political Studies, 44: 652–67.
—— (1997) Understanding Governance: Policy Networks, Governance, Reflexivity, and
Accountability, Milton Keynes: Open University Press.
—— (2000) ‘The Governance Narrative: Key Findings and Lessons from the ESRC’s
Whitehall Programme’, Public Administration, 78(2): 345–63.
Shapland, J. (1988) ‘Fiefs and Peasants: Accomplishing Change for Victims in the Crim-
inal Justice System’, in M. Maguire and J. Pointing (eds) Victims of Crime: A New
Deal?, Milton Keynes: Open University Press, 187–94.
—— (2000) ‘Victims and Criminal Justice: Creating Responsible Criminal Justice Agen-
cies’, in A. Crawford and J. Goodey (eds) Integrating a Victim Perspective within
Criminal Justice, Aldershot: Ashgate, 147–64.
Simon, J. (1995) ‘They Died with Their Boots On: The Boot Camp and the Limits of
Modern Penality’, Social Justice, 22(1): 25–49.
Social Exclusion Unit (1998) Bringing Britain Together, a National Strategy for Neighbour-
hood Renewal, London: Cabinet Office.
—— (2000) Anti-Social Behaviour, Report of Policy Action Team 8, National Strategy
for Neighbourhood Renewal, London: Cabinet Office.
Straw, J. (1998) ‘Building Social Cohesion, Order and Inclusion in a Market Economy’,
Paper presented to the Nexus Conference on Mapping out the Third Way, 3 July.
Tam, H. (1998) Communitarianism: A New Agenda for Politics and Citizenship, Hound-
mills, Basingstoke: Macmillan.
Taylor, A. (1997) ‘ ‘‘Arm’s Length But Hands On’’. Mapping the New Governance: The
Department of National Heritage and Cultural Politics in Britain’, Public Administra-
tion, 75(3): 441–66.
80 Adam Crawford

Tilley, N. (1995) Thinking About Crime Prevention Performance Indicators, CPU Paper no.
57, London: Home Office.
Warner, N. (1998) ‘The Crime and Disorder Bill’, presentation to the ESRC Research
Seminar on Crime Prevention and Community Safety, King’s College, London, 27
March.
Weatheritt, M. (1993) ‘Measuring Police Performance: Accounting or Accountability?’,
in R. Reiner and S. Spencer (eds) Accountable Policing: Effectiveness, Empowerment
and Equity, London: Institute for Public Policy Research, 24–54.
Wilson, J. Q. and Kelling, G. (1982) ‘Broken Windows: The Police and Neighbourhood
Safety’, Atlantic Monthly, March: 29–37.
Young, J. (1992) ‘Ten Points of Realism’, in J. Young and R. Matthews (eds) Rethinking
Criminology: The Realist Debate, London: Sage, 24–68.
3 Evaluation and evidence-led
crime reduction policy and
practice
Nick Tilley

Introduction: the evaluation and evidence-led agenda

Evidence and evaluation for policy and practice are in vogue. The British Prime
Minister, Tony Blair, has said that what counts is what works, and that the
current Labour government is committed to evidence-led policy. The Crime
and Disorder Act (1998) requires partnerships to assemble evidence to deter-
mine priorities for their strategies; it also requires them to monitor and evaluate
their efforts to tackle local crime and disorder problems. The government’s
three-year £250 million Crime Reduction Programme, announced in 1998 and
begun in 1999, follows from a Home Office report for the Comprehensive
Spending review (Goldblatt and Lewis 1998), summarising the evidence on
what had been found to work, or to be promising, in reducing crime. Ten per
cent of the £250 million is due to be spent on evaluation research. This is a
huge investment by the usual standards of social science. Recent reports of Her
Majesty’s Inspectorate of Constabulary (HMIC) on police work in crime reduc-
tion and of the Audit Commission on Local Authority Community Safety work
stress the need for evaluation and evidence, but bemoan its typical absence
(HMIC 1998; Audit Commission 1999). Moreover, HMIC and the Audit
Commission themselves root their inspections in part on evidence from
previous evaluation studies, and collect additional evidence on effectiveness
and efficiency as a further basis for reports and recommendations.
Traditional patterns of service delivery are no longer to be taken for granted.
Evidence-based justifications are required. The probation service is clearly being
steered towards a set of interventions for which there is some evidence of effec-
tiveness, and probation services are being asked to measure their own
effectiveness (Chapman and Hough 1998). The popularity of problem-oriented
policing springs in part from a similar sense that traditional policing will no
longer do (Leigh et al. 1996; 1998). Evidence is needed about the nature and
distribution of the problems currently being faced by the police, and it is then
to be collected to see whether efforts to target those problems are effective.
The evaluation agenda includes a strong economic emphasis on value for
money (Dhiri and Brand 1999; Stockdale et al. 1999). The ‘Best Value’ move-
ment is about re-examining services to try to get more benefit at less cost,
82 Nick Tilley

making use of evidence about inputs, outputs and outcomes (Leigh et al.
1999). An aspiration of the Crime Reduction Programme is to assemble
evidence of what works most cheaply to deliver effective crime reduction. This
is a critical agenda: it suggests a move from the use of professional authority or
of tradition as a warrant for action. Instead, policy and practice will be
dictated by hard evidence about what works and at what cost. The develop-
ments in relation to crime reduction are consonant with those presently
washing through the criminal justice system, and indeed through much social
policy and practice. Health, education, welfare, employment, economic regen-
eration, housing, and so on, are all being subjected to similar disciplines of
evaluation and evidence.
It clearly makes sense for evidence and evaluation to play a part in the devel-
opment of policy and practice. The issue that is addressed in this chapter,
however, is how they should do so. In the first part of this chapter I wish to
express doubts about four propositions:

1 that policy and practice can be led by evidence;

2 that random controlled trials are the gold standard for all evaluations;
3 that the University of Maryland overview of evaluation studies on crime
prevention provides reliable guidance for policy and practice; and
4 that cost-benefit analyses of measures can provide sound guidance for the
allocation of resources for crime prevention.

In the second part of the chapter I present what I consider to be a ‘realistic’ way
of taking forward evidence-based policy and practice.

Four doubts

Evidence-led policy and practice

Regardless of its surface appeal, evidence-led policy faces a range of method-
ological, practical and ideological problems. Together they cast doubt on the
project of evidence-led policy and practice.

Methodological issues
All evidence, including that from recorded crime, crime surveys, records of
convicted offenders, accounts of emotional responses to crime, etc., is socially
constructed (see Bottomley and Pease 1986; Maguire 1997). Data on crime
reflect the way they have been assembled. This is not simply a matter of data
sometimes being improperly, and hence corrigibly, collected. Rather, data are
always a product of the social processes through which they are generated (see
Cicourel 1964). Moreover, the range of available evidence, however
constructed, is always circumscribed. Clearly, all facts about everything cannot
be assembled – data collection is, and has to be, selective (Ragin 1994). Policy
 Evidence-led crime reduction policy 83

that is led by available evidence risks focusing on the easily measurable simply
because it is easily measurable.
Evidence does not speak for itself. Reasoning, imagination and values are
needed to draw conclusions from it (see Weber 1949). Evidence can be
collected and used to test, and potentially to falsify, conjectures about patterns,
though even here tested conjectures are not proved – they remain fallible
(Popper 1972) and are often disputed. Even in the less contentious world of
natural sciences and mathematics, it is only over time that a consensus may
emerge (Lakatos 1978).
If we assume that evidence does corroborate conjectures about patterns, this
does not mean that these patterns will be sustained over time. The social world is
subject to continuous change. One source of change is the understanding of actors
in the social world. In the case of crime prevention the preventer adapts to the
offender and the offender to the preventer. Each innovates and creates an
unstable environment for the other (Ekblom 1997). Policy and practice may need
to anticipate events on the basis of informed theory, rather than wait for evidence
that patterns are already present (see Foresight Crime Prevention Panel 2000).

Practical issues
Even if the methodological difficulties could be overcome, there are significant
practical obstacles to evidence-led policy and practice. No government can be
without a policy on crime; no criminal justice system can be suspended pending
evidence. Workers within the existing institutions cannot do nothing whilst
waiting for evidence to lead them. Much policy and practice has to proceed
without a research lead. Policies and practices cannot disregard public opinion
(Tilley and Laycock 2000). Politicians need to attend to the views of the elec-
torate, even when those views are not congruent with what the available
evidence suggests. Furthermore, elections create a periodic imperative for politi-
cians to show results. This, of course, opens a window for evidence to play a
part. It is liable, however, to lead to policies and practices offering relatively
quick fixes to take precedence over measures whose effect is only likely to be
felt in the longer term.
Established policies and practices will often constitute obstacles to evidence-
led changes, especially where there is already a public commitment to them, or
where there are vested interests in maintaining them (Weiss 1976).

Ideological problems in evidence-led policy and practice

Crime is a moral issue. Definitions of crime, and policies and practices to
respond to it, involve judgements of value. An appeal to evidence can be a way
of evading or passing on responsibilities that require those judgements of value.
Moreover, it is not only the ends that require ideological debate, so too do the
means. Efficacy is not the only criterion for judging the means to attain a given
end. The means, too, require that value questions be considered.
84 Nick Tilley

This section, it should be stressed, has expressed doubts about evidence-led

policy and practice – the notion that policies and practices can simply be
derived from evidence. It is not intended to cast doubt on the possibility (or
desirability) of policies and practices being informed by evidence. The distinc-
tion may seem a fine (and trite) one, but there are real issues involved. Crime
and disorder partnerships can be persuaded to try to assemble and warehouse
more and more data in the belief that they will tell them what to do. But the
data cannot and will not do so. The notion that evidence can lead policy may
also tempt politicians to hide behind evidence where real issues of value and
political judgement are at stake.

‘Experimental’ methods of evaluation

The classic purpose of evaluations is to find out what works and what does not
work, so that failures are not reproduced and the need to ‘reinvent the wheel’ is
avoided. The conventional gold standard for evaluation studies to sort the
programme wheat from the programme chaff is the randomly controlled trial
(RCT), where subjects are allocated blindly and randomly to experimental and
control groups.
In RCTs there is usually an experimental group, which is subject to the
‘treatment’, and a control group, which is not. Both are measured before and
after the intervention. Any difference in the change between the experimental
and control groups is attributable to the intervention. It is important that the
treatment is applied in a standard way so that the experimenter can be certain it
was this that had the effect. Of course, in more complex experiments control
and experimental groups can proliferate to try to deal with Hawthorne effects,
and with the potential independent effects of before measurement and any
interaction it might have with treatment. In the medical version of this method
of evaluation, a placebo may be given to the control group for the potential
effect of treatment per se, rather than the specific treatment applied. In social
experiments this is more difficult, but rather than no treatment, it may involve
an alternative treatment that might otherwise be provided for the subjects.
Moreover, there may have to be efforts to keep the experimental and control
groups separate to avoid potential contamination.
In this chapter, I wish to use the example of mandatory arrest for domestic
violence to reduce repeat incidents as a case study of RCTs for three reasons: first,
because the research is well documented; second, because of the high quality of
the studies; and finally, because there have been several replications, a welcome
but relatively uncommon feature of evaluation studies (Sherman 1992).
Mandatory arrest for domestic violence is probably the most evaluated
crime reduction initiative to make use of RCTs. The original Minneapolis
Domestic Violence Experiment found arrest (for ‘minor assaults which make
up the bulk of police calls to domestic violence’) the most effective of three
standard methods used by the police to reduce domestic violence. The other
two responses in the Minneapolis study were counselling or sending assailants
 Evidence-led crime reduction policy 85

away from the home for several hours. In simple misdemeanour domestic
assaults, cases were referred at random to one of the three responses. A six-
month follow-up checked the frequency and seriousness of any future
domestic violence. As ever, implementation and data collection were less
than perfect. Yet both official records and victim interviews showed that
repeat domestic violence was lowest for those where arrest had been the allo-
cated response. Official records revealed that there had been repeat incidents
in 10 per cent of arrest cases, in 19 per cent of the advice cases, and in 24 per
cent of the ‘send suspect away’ cases. On the basis of their findings, and
notwithstanding clearly stated caveats, the authors of the report note that:
‘the preponderance of evidence in the Minneapolis study strongly suggests
that the police should use arrest in most domestic violence cases’ (Sherman
and Berk 1984).
Domestic violence is of significant social concern and, unsurprisingly, on the
basis of the Minneapolis conclusions many US police departments adopted an
arrest policy. Sherman (1992) notes that 10 per cent of cities with a population
of over 100,000 made arrest the preferred police choice in 1994, 43 per cent in
1986, and 90 per cent in 1988. Further RCTs were conducted in other cities in
the US (Sherman 1992) with mixed results: in three cities mandatory arrest
increased repeat domestic violence, but in another three cities the level of
violence was reduced. This invites the question: why did the impact of manda-
tory arrest vary? The RCTs themselves can throw little light on this. They were
designed to see whether or not the initiatives worked in reducing repeat
domestic violence. Collectively they show that mandatory arrest can reduce
domestic violence, but it can also increase it. It is shown to be effective, but the
problem is that its effects vary.
Following the series of RCTs on mandatory arrest and domestic violence
there is some plausible post-hoc speculation to explain the mixed findings and
some (non-experimental) evidence is adduced for it. That is, it is believed that
arrest may anger unemployed people in marginal communities, whose disposi-
tion to violence is thereby increased. In middle-class communities, however, it
is believed that it shames people who are employed, and they are then chas-
tened and deterred (Sherman 1992). The structure of these explanations brings
out something of the weakness of the RCT method in so far as they highlight
the fact that the same measure may work in different ways and thus have
different effects according to variations in the circumstances of those involved.
The preoccupation of RCTs with net effects obscures this general feature of
interventions (see also Pawson and Tilley 1997: 34–46).
RCTs are not always practicable. If areas are the unit of analysis, for example,
it may not be possible to find and fund sufficient numbers for random allocation.
Random selection of experimental areas is not always politically viable.
Isolating/insulating members of experimental and control groups may not be
achievable. And so on. In these circumstances, the next best quasi-
experimental methods are conventionally preferred. These involve identifying
‘matched’ control and experimental groups to maximise their comparability.
86 Nick Tilley

This approach falls short of the standards of random allocation since there may
be significant differences between the experimental and the control groups,
which would automatically be eliminated through random allocation.
Ray Pawson and I have looked at the best British study undertaken using this
approach (for a detailed discussion see Pawson and Tilley 1994). The study, by
Bennett (1991), reports the evaluation of a police patrols initiative aiming to
reduce the fear of crime. This initiative was tried because some American evalu-
ation research had found that contact patrols had been associated with a fall in
fear of crime. Bennett reports on two experimental sites, in London and in
Birmingham, and two composite controls, chosen because of their high crime
rates and ‘visual indicators’ of disorder including graffiti, broken windows, crim-
inal damage and litter. He monitored implementation in the experimental areas
by measuring time spent by the police in the area and the contacts made by the
police with the public. He found no fall in fear of crime in the experimental
areas compared to the control areas. Other output and outcome measures, for
example those relating to recent sightings of police, perceptions of the area, and
informal social control practices, produced a mixed picture in the two sites. Thus
Bennett’s is a simple, if expensive, before-and-after experimental/composite
control comparison study with programme activity-level monitoring. It is a
classic of its type. Unfortunately, at its conclusion we know little of how the
programme was delivered or experienced. Encounters are more than just making
contact: what is said, who is spoken to, how interaction is managed are impor-
tant elements of a study, and Bennett’s report lacks this kind of information. Our
knowledge of the estates is reduced to their comparability in a number of superfi-
cial respects. The quality of community life and organisation is not considered
(except for an unremarked variation in measurements of ‘level of satisfaction’
and ‘sense of community’ in the two experimental sites, not relevant in what
were composite controls). There is no explanation for the variations in outcome,
nor for the differences between the British and American results. At the end of
the initiative we are little the wiser, other than knowing that in these two places
the intended fear reduction outcome was not achieved. We could certainly not
conclude that in another place police contacts patrol would not lead to a reduc-
tion in fear of crime.
The underlying problems here repeat those of the mandatory arrest series of
studies. How programmes have effects varies by the circumstances in which
they are introduced. This will mean that different outcomes will occur.
Sophisticated though they are in technical terms, Bennett’s quasi-experimental
methods railroad over this.

The University of Maryland overview of crime prevention

evaluation studies
The apotheosis of the experimental approach to evaluation is a review of studies
of what works in crime prevention conducted for US Congress by the
University of Maryland. The findings are posted on the Internet (Sherman et al.
 Evidence-led crime reduction policy 87

1997). Sherman and his associates use a scale from 5 to 1, where, notwith-
standing the disastrous effects they may have as evidenced by the mandatory
arrest for domestic violence débâcle, the highest score – ‘5’ – is given to RCTs.
The next ‘best’ studies most closely corresponding to RCTs are awarded 4, and
so on. The interventions are then classified under four headings: ‘what works’,
‘what doesn’t work’, ‘what’s promising’ and ‘what’s unknown’. The result is in
my opinion dangerous nonsense.
In recent accounts of this mega overview of evaluations, the Maryland group
has used the same headings, but added interesting comments, which seemingly
acknowledge some of the problems. For example, under ‘what works’, they say:

There are programs that we can be reasonably certain prevent crime or

reduce risk factors for crime in the kinds of social context in which they
have been evaluated and for which the findings can be generalised to
similar settings in other places and times.
(Sherman et al. 1998: 6)

Here they recognise that the positive outcomes which programmes bring about
are contingent on context, that where programmes are found to have been
effective this does not mean that they can be expected to have the same results
elsewhere. The effectiveness is dependent on circumstances. The same results
will only be brought about in similar settings. This suggests that identifying
what is salient in the setting is crucial if transferable lessons are to be achieved.
What counts as a ‘similar’ setting is not self-evident. Here, then,‘what works’ is
adjusted and qualified to become ‘what works in (saliently) similar settings’.
Over the issue of ‘what doesn’t work’, however, context-dependency is curi-
ously forgotten. Under the heading ‘what doesn’t work’, the Maryland group says:

These are programs that we are reasonably confident from available

evidence fail to prevent crime or reduce risk factors for crime, using the
identical scientific criteria used for deciding what works.
(Sherman et al. 1998: 6)

No reference is made here to context-dependence, though the authors give no

reason for omitting it. Yet if ‘what works’ depends on a conducive context, and
it is accepted that failure using the same measure might occur in a non-
conducive context, it then follows that a trial that has failed may have been
undertaken in one of these non-conducive settings. A re-trial in a different
setting that is suited to the measure introduced might be followed by a
successful outcome. Indeed no number of failures could show that the measure
could never be successful. It might simply be that no trial has occurred in a suit-
able setting.
In general terms with regard to their taxonomy, the Maryland group
concedes that:
88 Nick Tilley

The weakest aspect of this classification (into ‘what works’ etc.) is that
there is no standard means of establishing external validity: exactly what
variations in program content and setting might affect the generalisability
of findings from evaluations. In the current state of science, that can be
accomplished only by the accumulation of many tests in many settings with
all major variations on the program theme. None of the programs reviewed
in this report have accumulated such a body of knowledge so far. The
conclusions drawn in the report about what works and what doesn’t should
be read, therefore, as more certain to the extent that all conditions of the
programs that were evaluated (e.g. population demographics, program
elements, social context) are replicated in other settings. The greater the
difference on such dimensions between evaluated programs and other
programs using the same name, the less certain the application of this
report’s conclusions must be.
(Sherman et al. 1998: 6)

This is fine as far as it goes, but it begs questions about salience of similarity and
difference. It is impossible to replicate ‘all conditions’. This would require the
same people, same houses, same time, same geography, same age, same external
political events, same offender populations, etc. Such a requirement is utopian
and therefore absurd. A judgement has to be made about what matters in terms
of similarity, both in terms of the intervention measure itself and in terms of the
contexts in which it is introduced. The key question, therefore, is: what ‘popu-
lation demographics, program elements, social context’, etc. need to be the
same, and why? And this is precisely the issue which the experimental approach
consistently misses or fudges.
Though the Maryland group do not say so in so many words, once it is
conceded that context is crucial for the effects produced, the only conclusion
that can be drawn from their review is that there is quite compelling
evidence that some measures can be associated with intended outcomes. In
relation to any given measure it is necessary to have only one strong study to
establish this. Any study which does not find that the intended outcome has
followed cannot conclude that the measure cannot produce that outcome.
What goes for intended outcomes goes, of course, also for unintended
outcomes of interventions. For the same reasons these too will be contextu-
ally contingent.
Instead of the bold headings used in the Maryland overview, which proclaim
‘what works’, ‘what doesn’t work’, ‘what’s promising’ and ‘what’s unknown’, the
authors’ concessions suggest that more accurate headings would be more
circumspect. They might read ‘what’s been found to work somewhere at some
time’, ‘what’s not yet been found to work anywhere, but might do so in the right
conditions’, ‘what seems to work somewhere’ and ‘what might work somewhere
at some time’. Less snappy to be sure, but less misleading, too. The worry is that
the Maryland study headlines will be taken at face value and that their proper
qualifications will be disregarded
 Evidence-led crime reduction policy 89

The Maryland group’s concessions about the shortcomings of their classifica-

tions of evaluations echo weaknesses already mentioned. They accept that past
performance is not a good guide to future performance, that performance in one
place is not an adequate guide for performance in another place, and the reason
for this is that places differ in ways which mean the same measure will work
differently in saliently different contexts. This is not a matter that the high-
scoring experimental evaluations reviewed have been good at dealing with.
Overviews of evaluations, including Sherman’s team’s, consistently find
inconsistent findings (Poyner 1993; Martinson 1974; Gendreau and Ross 1987).
Thus if one takes some examples of frequently advocated crime prevention
interventions, contradictory results are the norm. For example, with regard to
lighting, Painter (1995) concluded it did help in crime prevention, but Atkins
et al. (1991) that it did not. With regard to property marking, Laycock (1997)
found it did, Knuttson (1984) found it did not. The Kirkholt burglary reduction
programme was one of the most successful and influential initiatives in Britain,
but efforts to replicate it again produced mixed results (Tilley 1996).
Mixed results are a menace for traditional evaluation, which is concerned
with finding whether measures do or do not work and nothing else. They leave
the policy-maker in a quandary. Traditional experimental evaluation methods,
which consistently produce inconsistent findings, are therefore largely a waste
of time and money. Moreover, they are also a travesty of real experimental
methods, as used in the natural sciences, where experimental control is seldom
effected through random allocation (Tilley 2000).
Those who expect consistent findings misread the logic-in-use of experi-
mental evaluation methods. Experimental methods can only measure the
association that did occur between two variables (intervention/outcome) in a
specific time and place. There can be no ground from this finding alone for
believing that because an association occurred at one place or time it will do so
at some future place and time.

Cost-benefit analysis
Cost-benefit analysis tries to standardise inputs and outcomes and to assign
them a common currency. In this way it is expected to provide an objective,
technical basis for choosing between competing priorities. It assumes that the
causal relationship between measures introduced and outcomes achieved can be
established and quantified with some confidence. This is clearly not generally
the case. For reasons already given, measures produce different outcomes across
space and time, and traditional evaluation methods have been unable
adequately to capture this.
What count as costs and benefits is not self-evident and what is included
may be quite narrowly circumscribed or it may be very wide-ranging. Inputs may
include: funds specifically allocated from the public purse for the programme;
funds allocated for a variety of purposes including those related to the
programme; programme-relevant private sector, individual and household
90 Nick Tilley

expenditure; resources allocated in kind; and volunteer effort. Benefits may

include: savings to the agency responsible for the programme; to insurance
companies; to individuals and households; to businesses; to local authorities;
and to other national bodies. Some financial benefits may be direct, some may
be indirect. Other benefits have to do with quality of life issues. Further benefits
still will be reductions in personal harm.
Where the benefits are non-financial the aim is to give them cash equivalent
values. The technical and conceptual estimation difficulties are overwhelming.
Methods include ‘willing to pay estimates’ – how much would a person, house-
hold or business be prepared to pay to avoid a given harm; and ‘willing to accept
estimates’ – how much would they need to be paid to endure the harm? Both
produce arbitrary sums. The first is limited by what people might have and
therefore could afford at a given point. The second produces ‘infinite amount’
answers (for example, when referring to a fatal accident), and this cannot be
built into cost-benefit calculations (for useful discussions see Adams 1995;
Zimring and Hawkins 1995). As an extreme case, consider murder. In relation
to serial killer Harold Shipman’s middle-aged and elderly victims, the financial
costs of the discovery of their murder will almost certainly have outweighed any
savings from their detection. Indeed, given the age of the victims, it may be that
in cost-benefit terms their murders saved the public purse! Moreover, the
discovery of their murders may well increase the fears that elderly patients have
of visiting their general practitioners. Dependency on cost-benefit analysis
misses the moral point of attending to murder. There is no means of measuring
the value of a prevented homicide or rape. The conventional ‘willing to pay’
and ‘willing to accept’ measurements make no sense. Policy ends require discus-
sion and debate about values.
There seems to be no non-arbitrary way of deciding what to include in costs
or in benefits or in how to make measurements of non-financial benefits. There
are only contestable and modifiable conventions. Moreover, the calculations
have to be applied to estimates of net effects which are unstable due to contex-
tual variation and instability. It is difficult not to conclude that cost-benefit
calculations are a complex chimera, conjured up by clever people given a super-
ficially sensible but ultimately undoable job.
Whilst it may be possible (and useful) roughly to estimate the financial costs
and benefits of past programmes for specific parties, and their net monetary
benefits, overall estimates including monetary and non-monetary elements
appear unintelligible and arbitrary, and are predictably flawed for predictive
purposes. I turn now to an alternative, realistic approach to understanding how
evaluation and evidence might inform policy and practice

Realistic evaluation and evidence for policy and practice

Because of the failures of traditional experimental evaluations, Ray Pawson and
I have formulated an alternative, which we term ‘realistic evaluation’ (Pawson
and Tilley 1997). At the heart of this approach are precisely those matters
 Evidence-led crime reduction policy 91

where traditional, conventional evaluation has failed. Realistic evaluation is

concerned with understanding how programmes and policies bring about
different effects in different places and among different people.
Realistic evaluation is rooted in scientific realism. Scientific realism has a
distinctive account of causation. Rather than understanding cause in terms of
constant conjunctions between observable events (the Humean position
assumed in the experimental approach), cause is understood in generative
terms. We can think of generative causation in terms of a clock. The hands of a
clock move in an orderly way. There is, though, a mechanism that lies behind,
or causes the orderly movement of, the clock’s hands. We understand the move-
ment of the hands by understanding the (hidden) mechanism lying behind or
generating it. For realists science involves explaining regularities by reference to
the mechanisms producing them. In social life the signing of a cheque and the
transfer of money by the teller to the customer in a bank makes sense in terms
of the (unseen) banking system as a whole.
The triggering of mechanisms is not unconditional. It is contingent on an
appropriate context. Think of fireworks: the explosions are only triggered when
the blue touchpaper is lit if the conditions are right. There must be enough
gunpowder, it must be sufficiently compacted, it must be dry enough, etc.
Laboratory science comprises the construction of idealised conditions in which
conjectured mechanisms will be activated.
A social programme describes an effort to effect a change in some pattern of
behaviour. For realists this has to be understood in terms of an alteration in the
mechanisms generating behaviour. Either existing mechanisms generating
present behaviour have to be disabled or some new causal mechanisms have to
be triggered to overcome those that would otherwise operate. The altered
behaviour resulting from changes in patterns of mechanisms triggered comprises
the programme outcomes.
Ray Pawson and I understand the major mechanisms that programmes intro-
duce or disable to relate to the reasoning and resources of individuals.
Reasoning has to do with values, meanings, and estimates of the costs and bene-
fits. Resources have to do with the powers which actors have to choose between
varying courses of action. Thus, programmes do not just happen to participants
externally, as when one billiard ball strikes another. Rather, it is how human
agents engage with programmes that comprises the medium through which they
do or do not have their effects. Therefore, evaluation research is concerned
with unpicking the ways in which a programme changes the mechanisms trig-
gered to alter the pattern of behaviour within given types of context.
For the realistic evaluator the key evaluation question is: what works for
whom in what circumstances, and how? The traditional evaluation question,
‘what works?’ (which is often assumed to amount to, ‘what always works?’,
because of constant conjunction assumptions) is misconstrued. Programmes do
not work unconditionally and useful answers to evaluation questions entail
penetrating the contexts for the activation of varying causal mechanisms to
produce changed patterns of behaviour. We take evaluation to require a theory
92 Nick Tilley

to explain or predict how changes introduced by an intervention alter the

balance of causal mechanisms at work and lead to a change in behaviour. As
already indicated, whether or not an intervention effects such a change, and
what change is brought about, will depend on the prior conditions or on
context. Some mechanisms may be disabled and some new ones activated as a
result of the intervention. The same programme will be encountered by
different programme participants in different ways, in accordance with varia-
tions in existing bases for their action. A key part of any evaluation will include
specification of key subgroups whose members, it is conjectured, will experience
and thus respond to the same programme in differing ways.
In regard to replication, what counts as the ‘same’ programme is also prob-
lematic. No two programmes are, or can be, identical in every respect (Tilley
1996). A programme theory is needed to specify what a programme is expected
to do in relation to a given target population, to determine what matters from
one programme to another for it to count as the same.
Here, we can use prisons as an example. The traditional evaluation agenda
asks, ‘does prison work?’ Yet we know that prisoners and prisons vary, as do the
communities in which they are located. We know of various mechanisms that
may be triggered by prison – incapacitation, specific deterrence, general deter-
rence, criminal acculturation, non-criminal capacity building, criminality
capacity building, non-criminal opportunity reduction, criminal identity rein-
forcement, etc. In only one of these mechanisms is the subject relatively
passive. Incapacitation happens to individuals; it does not depend much on
their agency for its effects (though its crime preventive impact depends on an
estimate of their exercise of agency were they not incarcerated). The other
mechanisms depend on alterations in choices available for agents, their
apparent consequences and the meanings attached to them by those agents. If
we treat all prisons and prisoners alike, and all social conditions in which
prisons are lodged as the same, we will miss that variation. If we take as our
example groups of prisoners, where net nil effects are found this could reflect
less the causal impotence of prison (or a particular regime) than its capacity to
generate countervailing causal forces working their way through to have
competing impacts on differing inmates.
Many crime prevention measures provide no physical obstacle to crime at
all. CCTV, lighting improvements, alarms, drugs counselling, employment
opportunity creation, etc., all depend on mechanisms affecting choices if they
are to have their effects. Even where programme participants are relatively
passive, mechanism questions arise, but are less concerned with choices. Drugs,
surgery, dietary modifications, for example, involve the operation of biochem-
ical mechanisms. The same basic question still needs to be asked: how, for
whom, and in what contexts will these generate what effects?
Realistic evaluation calls for the formulation, test and refinement of
‘context-mechanism-outcome pattern configurations’ (CMOCs, or ‘seemocks’).
Seemocks comprise models explaining how measures bring about outcomes in
specified sorts of social conditions. They are middle-range theories (Merton
 Evidence-led crime reduction policy 93

1968) in three senses. First, they go beyond the descriptive particulars of

specific interventions at particular places and times, but they do not provide
universal truths for all places and times. Second, they do not provide general
theories of the phenomena being examined. In the case of prisons, they do not
provide a general theory of the prison or of criminal careers, though they may
draw on and inform general theories. Third, they will pay less attention to
broader conditions for social action, and focus more on the locally variable
within that relatively stable set of background circumstances. What is primarily
at issue in the theory is that which is variable within social formations and in
the relatively short term. In relation to prison, this might include, for example,
attributes of inmates, rates of imprisonment, rates of unemployment outside
prison, the level and content of educational provision within prison, the regime
provided within prison, density of prison occupation, prison lay-out, etc.
Seemocks specify what interventions within relatively variable local contextual
conditions trigger which mechanisms with what outcomes, though at their best
they will also make explicit the broader structural arrangements that are taken
for granted but crucial for the potential causal efficacy of the measures which
are introduced.

Doing realistic evaluation

The realistic evaluator’s starting point is to construct the seemock theory or
theories in relation to a programme, after which the next task is to design a
study that will test the seemocks. There are various sources of theory in relation
to a programme. A good starting point will be the architects of the programme.
How do they expect the measures they are planning to change behaviour?
Whose behaviour do they expect to change and whose to remain the same?
What unintended changes might be sparked amongst those touched by the
programme, and how? Practitioners, too, will have ideas about how the
programme is working for whom and why. They will also have an idea about
those for whom the programme is not working and why this is the case.
Practitioners are not, of course, best asked baldly about their implicit theories,
but can usually speak with conviction about cases encountered, and it will be
for the researcher to build a theory out of this and then to check that the practi-
tioner goes along with it. In the event, the theory is often co-produced by a
succession of practitioners’ tales; the realistic evaluator formalises the theory
reflecting the underlying context, mechanism, outcome elements. The practi-
tioner theory, like that of the policy-maker, may well be mistaken, of course.
The job of the realistic evaluator in designing the research is to devise methods
to test it. Other sources of theory include social science literature, previous
evaluation studies, common sense, and so forth. A range of theories can be
developed.
Practitioner and policy-maker theory will often need to be supplemented by
accounts of unintended consequences and unacknowledged conditions, beyond
the practical consciousness of those implicated in programmes. The formalised
94 Nick Tilley

and supplemented middle-range realist theories can then be worked through to

identify appropriate comparison groups within the programme, where detailed
outcome patterns variations would be expected. The upshot will be a tested
theory or series of theories that provides an improved grasp of what worked how
and with whom within the programme.

Policy and practice in the sight of realism

The evaluation agenda sketched out requires that programmes be unpicked,
unpacked and evaluated in terms of sets of variable outcomes. The result will
rarely be that programmes have had and can have no impact along the intended
lines, or that programmes do have and will always have their planned effects.
Programmes, people and social settings are simply not like that.
If the evaluation enterprise is deemed to be about finding out about what
works and spreading it, practitioners may be encouraged to adopt a recipe book
approach, assuming that measures can be pulled off the ‘works’ shelf. Because of
contextual variability and fluidity, however, this is unlikely to maximise effi-
ciency and effectiveness. Ideally, practitioners need to understand seemocks and
be able to think through their own problem situations in terms of the scope
there is to trigger change mechanisms using available intervention measures.
They need then to refine practice through monitoring what is produced. They
need to be (evidence-based) theory learners, users, refiners and developers, not
mechanical deliverers of standard programmes implemented with integrity.
That is, they need to be reflective, evidence-based practitioners.
The rationale for recipe books is that some people can produce palatable
meals more easily with them than without them. Consider the person who
‘knows what they like’ but has never cooked. He knows that he likes gin, sugar,
garlic, aniseed, mushrooms, smoked salmon, custard, ice cream, black sausage
and rice. So he mixes them up and puts them in the oven. That person would be
better off with a recipe book. Recipe books are better than nothing. Moreover,
there are better and worse recipe books. Talented cooks, however, make rather
little use of them. The popular British television programme, Ready, Steady,
Cook, is about guests coming to the show with sets of ingredients, each of which
they presumably like, and presenting them to the professional chef who conjures
up a tasty meal from them. The chef knows enough about the principles of
cooking and the attributes of the ingredients, and has enough evidence of what
kinds of things work in cooking, to be creative in an informed way.
Some current developments appear to treat occupational groups like cooks
‘who know what they like’ but produce ‘dogs’ dinners’ as a consequence. Hence
teachers have pre-programmed and standardised literacy and numeracy hours;
and probation services are being channelled into delivering accredited
programmes with detailed instructions about what is to be done and when.
Probation and community safety workers are to be trained (rather than
educated) into ‘occupational standards’, so that they have ‘core competences’ to
deliver standard ‘approved services’, in the light of the ‘skills maps’ which
 Evidence-led crime reduction policy 95

describe what they are expected to do. In the worst cases the recipe books will
reflect the prejudices of a central authority rather than the practitioner. In the
best cases they may reflect evidence about what has been shown to work in
some context for some subjects. Where untutored, counterproductive discretion
is used then recipe books will be useful and damage will be reduced by requiring
their employment. Where situations are relatively stable (as in the kitchen),
many benefits may follow from the creation of recipe books and requirements
that they be followed. McDonald’s illustrates well the profit to be made by stan-
dardisation. The company’s Hamburger University is presumably a model that
others in the business of vocational training for standardised service delivery
should follow.
Enforcing the adequate may dispense with the truly awful. It will also inhibit
the brilliant. McDonald’s may serve up something better than a dog’s dinner,
but it will be worse than the fashionable London restaurant the River Café.
Enforcing the adequate is in effect to engage in mediocratisation. Evaluators
steering us towards it belong to a new class of evidence-based mediocrats.
In the light of some experience of crime reduction efforts in Britain, there
are examples of dogs’ dinners (Tilley et al. 1999). There are also calls for recipe
books. One role for evaluators might be to contribute to recipe books (and to
help eliminate recipes that don’t work). Another is to help identify CMOCs
that policy-makers and practitioners can use intelligently to determine what
evidence suggests would make sense in differing conditions. Reflective practice
requires the intelligent deployment of principles or theories, which have been
refined through a series of evaluation studies. The elements of theory that will
be transferable/applicable relate to causal mechanisms which may be triggered
and/or inhibited in specified types of contexts to effect changes in behavioural
regularities in the intended direction. Evidence is assembled to arbitrate
between/refine theories for practitioners to apply reflectively.
There are real risks of standardisation and mediocratisation in crime reduction
and in evaluation studies designed to inform standardising and mediocratising
processes. The apparently competent may turn out to be dangerously incompe-
tent. One example of the dangers of crass mediocratisation in crime prevention
comes from just that use of mandatory arrest for domestic violence as a means to
reduce repeat incidents produced in the experiments discussed earlier in this
chapter. Indeed, on the basis of the Minneapolis experiment and the six subse-
quent studies, Sherman (1992) has concluded that mandatory arrest laws should
be repealed, and there should be ‘structured police discretion’. The mandatory
arrest for domestic violence research seems to suggest that informed discretion is
better than either ignorant discretion or substitution of rule-book recipes for it.

Conclusion
Evidence is useful for policy and practice: for informing decisions about priori-
ties and for identifying promising responses to deal with problems. Most social
scientists will welcome the present government’s expressed commitment to see
96 Nick Tilley

more use made of evidence. The evidence-led agenda, however, promises too
strong a role for evidence. Evidence is useful, of course, for checking ideas, but
it is good ideas that have survived critical scrutiny that are needed for effective
policy and practice. The orthodox experimental paradigm for evaluation asks
the wrong question and comes out with misleading and inconsistent answers.
Evaluation studies need to be realistic if they are to be informative. Such studies
require theoretical sensitivity as well as technical skills. The lessons they will
most usefully generate are not crass and unconditional suggestions that
measures will or will not work. What they will do is help develop tested theory
of what works for whom, in what circumstances, and this should help practi-
tioners and policy-makers think through their responses to issues and problems
with a better chance of producing their intended outcomes.
Even with this realistic understanding, however, there are crucial value ques-
tions at stake in policy and practice, and in the theories behind them, that
cannot be sidestepped by reference simply to the authority of empirical social
science.

References
Adams, J. (1995) Risk, London: UCL Press.
Atkins, S., Husain, S. and Storey, A. (1991) The Influence of Street Lighting on Crime,
Crime Prevention Unit Paper no. 28, London: Home Office.
Audit Commission (1999) Safety in Numbers: Promoting Community Safety, London:
Audit Commission.
Bennett, T. (1991) ‘The Effectiveness of a Police-initiated Fear-reducing Strategy’, British
Journal of Criminology, 31: 1–14.
Bottomley, K. and Pease, K. (1986) Crime and Punishment: Interpreting the Data, Milton
Keynes: Open University Press.
Chapman, T. and Hough, M. (1998) Evidence Based Practice: A Guide to Effective Practice,
London: Home Office.
Cicourel, A. (1964) Method and Measurement in Sociology, New York: Free Press.
Dhiri, S. and Brand, S. (1999) Analysis of Costs and Benefits: Guidance for Evaluators,
Crime Reduction Programme, Guidance Notes 1, London: Home Office.
Ekblom, Paul (1997) ‘Gearing Up Against Crime: A Dynamic Framework to Help
Designers Keep Up with the Adaptive Criminal in a Changing World’, International
Journal of Risk, Security and Crime Prevention, 2: 249–65.
Foresight Crime Prevention Panel (2000) Just Around the Corner: A Consultation Docu-
ment, London: Department of Trade and Industry.
Gendreau, P. and Ross, R. (1987) ‘The Revivification of Rehabilitation’, Justice
Quarterly, 4: 349–408.
Goldblatt, P. and Lewis, C. (1998) (eds) Reducing Offending: An Assessment of the
Research Evidence on Ways of Dealing with Offending Behaviour, Home Office Research
Study no. 187, London: Home Office.
Gould, S. J. (1981) The Mismeasure of Man, New York: Norton.
HMIC (1998) Beating Crime, London: Home Office.
Knuttson, J. (1984) Operation Identification: A Way to Prevent Burglaries? Report no. 14,
Stockholm: National Swedish Council for Crime Prevention.
 Evidence-led crime reduction policy 97

Lakatos, I. (1978) The Methodology of Scientific Research Programmes, Cambridge:

Cambridge University Press.
Laycock, G. (1997) ‘Operation Identification, or the Power of Publicity’, in R. Clarke
(ed.) Situational Crime Prevention: Successful Case Studies, New York: Harrow and
Heston.
Leigh, A., Mundy, G. and Tuffin, R. (1999) Best Value Policing: Making Preparations,
Crime Prevention and Detection Series Paper 116, London: Home Office.
——, Read. T. and Tilley, N. (1996) Problem-oriented Policing: Brit Pop, Crime Preven-
tion and Detection Series Paper 75, London: Home Office.
——, Read. T. and Tilley, N. (1998) Brit Pop II: Problem-oriented Policing in Practice,
Police Research Series Paper 93, London: Home Office.
Maguire, M. (1997) ‘Crime Statistics, Patterns and Trends: Changing Perceptions and
their Implications’, in M. Maguire, R. Morgan and R. Reiner (eds) The Oxford Hand-
book of Criminology, 2nd edition, Oxford: Oxford University Press.
Martinson, R. (1974) ‘What Works? Questions and Answers About Prison Reform’,
Public Interest, 35: 22–45.
Merton, R. (1968) Social Theory and Social Structure, New York: Free Press.
Painter, K. (1995) An Evaluation of the Impact of Street Lighting on Crime, Fear of Crime
and Quality of Life, unpublished Ph.D. thesis, Cambridge University.
Pawson, R. and Tilley, N. (1994) ‘What Works in Evaluation Research?’ British Journal of
Criminology 34: 291–306.
—— (1997) Realistic Evaluation, London: Sage.
Popper, K. (1972) Objective Knowledge, Oxford: Clarendon Press.
Poyner, B. (1993) ‘What Works in Crime Prevention: An Overview of Evaluations’, in
R. Clarke (ed.) Crime Prevention Studies, vol. 1, Monsey, NY: Criminal Justice Press.
Ragin, C. (1994) Constructing Social Research, Thousand Oaks, CA: Pine Forge Press.
Sherman, L. (1992) Policing Domestic Violence, New York: Free Press.
—— and Berk, R. (1984) The Minneapolis Domestic Violence Experiment, The Police
Foundation.
——, Gottfredson, D., MacKenzie, D., Eck, J., Reuter, P. and Bushway, S. (1997)
Preventing Crime: What Works, What Doesn’t, What’s Promising: A Report to the United
States Congress. Online. Available: http://www.ncjrs.org/works/index.htm
—— (1998) Preventing Crime: What Works, What Doesn’t, What’s Promising, National
Institute of Justice: Research in Brief, Office of Justice Programs, US Department of
Justice.
Stockdale, J., Whitehead, C. and Gresham, P. (1999) Applying Economic Analysis to Policing
Activity, Crime Prevention and Detection Series Paper 103, London: Home Office.
Tilley, N. (1996) ‘Demonstration, Exemplification, Duplication and Replication in Eval-
uation Research’, Evaluation, 2: 35–50.
—— (2000) ‘Experimentation and Criminal Justice in the United Kingdom’, Crime and
Delinquency, 46(2): 194–213.
—— and Laycock, N. (2000) ‘Joining Up Research, Policy and Practice About Crime’,
Policy Studies, 21(3): 213–27.
——, Pease, K., Hough, M. and Brown, R. (1999) Burglary Prevention: Early Lessons from
the Crime Reduction Programme, Crime Reduction Research Series Paper 1, London:
Home Office.
Weber, M. (1949) The Methodology of the Social Sciences, New York: Free Press.
Weiss, C. (1976) ‘Using Research in the Policy Process’, Policy Studies Journal, 4: 224–8.
Zimring, F. and Hawkins, G. (1995) Incapacitation, New York: Oxford University Press.
4 Crime, community safety
and toleration
Lynn Hancock and Roger Matthews

Introduction
The recently published Home Office Statement of Purpose informs us that its
aim is:

To build a safe, just and tolerant society, in which the rights and responsi-
bilities of individuals, families and communities are balanced, and the
protection and security of the public are maintained.
(Home Office 1998)

This claim appears at first sight to be eminently commendable and there are few
in Britain who, if asked, would not identify with some if not all of these objec-
tives. However, as has been pointed out, rather than strive to develop a more
tolerant society, recent British government policies have been associated with
the ‘regimentation of opinion, moral closure, the repression of dissent and insti-
tutional intolerance’ (Hughes 1998). In contrast to the stated aim of promoting
greater levels of tolerance government policy has expressed a profound intoler-
ance for a range of activities including noisy neighbours, squeegee merchants,
youths hanging around the streets as well as other forms of ‘anti-social’
behaviour (Muncie 1999). Significantly, most of the activities which have
become the object of official intolerance in recent years are, in themselves, rela-
tively innocuous. Nevertheless, we are reminded that we are now engaged in a
‘war against crime’ and a ‘war against drugs’, although it is not clear whether
these are two fronts of the same war or two different wars. Either way, the
language of war presents offenders as the enemy to be defeated or dominated,
rather than helped or tolerated. At the same time the emphasis on zero toler-
ance, or selective intolerance as it has been more accurately described, has been
actively pursued by different agencies and given considerable official support.
Being ‘tough on crime and the causes of crime’ means, according to the Home
Secretary Jack Straw, the instigation of ‘tough and consistent prison sentences
for serious criminals’, and he insists that the ‘purpose of our criminal justice
system is to catch and punish offenders not to make excuses for them’ (Straw
1999). Within this punitive paradigm, the focus is largely on punishment, inca-
 Crime, community safety and toleration 99

pacitation, longer prison sentences and generally getting tough on offenders.

There is little attention given to disadvantage, mitigation, rehabilitation or
toleration.
Government-sponsored campaigns to inform on drug dealers, welfare
scroungers and other ‘anti-social’ groups are designed to encourage citizens to
display increasing intolerance towards certain forms of activity and these have
run alongside anti-sexist and anti-racist campaigns which have called for greater
intolerance of what are increasingly deemed to be unacceptable acts and atti-
tudes. Thus it is apparent that the government is formally calling for greater
tolerance while simultaneously pursuing a strategy of selective intolerance. The
implicit message which is conveyed in recent policy developments in the field
of ‘law and order’ would seem to be that greater intolerance is the route to a
safer and more just society.
How can we make sense of this apparent contradiction? We could simply
dismiss it as an example of hypocrisy or as a consequence of the inevitable gap
between rhetoric and reality, or the mismatch between intentions and
outcomes. Or more cynically we could see it as a product of the ‘New
Managerialism’ which, while espousing the virtues of cost effectiveness, spends
valuable time producing mission statements and statements of purpose that are
either bland, vague or in some cases totally vacuous. However, rather than just
dismiss this statement of purpose or engage in deep cynicism (although this does
have a certain attraction), our intention is to show that embedded in the Home
Office’s Statement of Purpose are some important issues centred around the
notion of tolerance and that these are linked to an unresolved political tension
between liberalism and communitarianism.
Consequently, this chapter has two main aims. First, to articulate the notion
of toleration, which we believe to be a much misunderstood concept, and at the
same time to examine its largely unacknowledged role as centrally organising
concept of sociological investigation. Second, we aim to examine both the
liberal and the communitarian approaches to toleration and attempt to show
that the tension which is evident in current crime control and community
safety policy arises from an uneasy mix of these two political doctrines.

What is toleration?
‘Toleration’ is a term, which like ‘community’, is virtually devoid of any nega-
tive associations. Tolerance suggests respect for others, a recognition of
individual rights, as well as an acceptance of individual difference and social
diversity. It is, however, not an expression of benevolence, but rather it
embodies a sense of disapproval and implies a degree of suffering or sufferance
(Yovel 1998). The working definition of ‘tolerance’ which is widely used is:
‘The deliberate choice not to interfere with conduct or beliefs, with which one
disapproves’. In this way we might distinguish tolerance from indifference and
from those situations where there may be no real choice over whether to take
action or not. Thus in examining the exercise of toleration we need to know
100 Lynn Hancock and Roger Matthews

something about the reasons for intervention and non-intervention (Horton

and Nicholson 1992).
Public tolerance is not the same thing as public opinion. Public opinion, as
expressed in public attitude surveys, asks people what they would advocate
whereas the notion of toleration is concerned with what people will accept or
be willing to endorse under certain circumstances. For example, a public
opinion survey may characteristically ask what length of sentence members of
the public would advocate for a certain type of crime whereas what respondents
might be willing to live with or tolerate may be considerably different (Turner et
al. 1997). Thus the term toleration is much broader than public opinion and
not only involves some consideration of spontaneously expressed viewpoints,
but invites a more detailed consideration of the range and depth of public atti-
tudes.
The notion of toleration, although mainly discussed and refined by political
philosophers, is central to the understanding of crime, law and punishment.
The distinction between crime and incivilities, law and norms, as well as the
application of different sanctions, is conditioned by the prevailing forms of
toleration. Although it is the case that in relation to certain activities such as
rape, murder and robbery there may be a relatively high degree of consensus
that these offences should not be tolerated, the situation becomes more compli-
cated when we consider forms of social disorder or non-criminal activity
(Burney 1999), since those committing lesser transgressions may well claim
some ‘right’ to engage in these activities (e.g., to buy or sell sex, to play music,
to ‘hang around’ on the street), and this right is itself associated with notions of
freedom and autonomy (Mendus 1989; Newey 1999). Inasmuch as the concept
of toleration has featured in the criminological literature it is often presented as
a uni-dimensional expression of free-floating benevolence. However, a review of
the relevant literature suggests that it has a number of important attributes.
First, the notion of tolerance is profoundly social and arises out of social
interaction and in turn serves to regulate that interaction. It involves an expres-
sion of power in that it suggests the ability to limit or influence the activity of
others, and like the concept of power as Michel Foucault (1979) has shown, it is
a relational concept. Just as people do not simply ‘hold’ or ‘exercise’ power over
others, so toleration implies a degree of reciprocity and mutual recognition.
Second, toleration is tied to the material conditions of existence in that the
decisions over the expression of tolerance are conditioned by the structural
location and situational pressures of those expressing tolerance. Thus the class,
gender or ethnic position of those involved in this relation will deeply influence
the nature of tolerance involved and how it is expressed. Third, it is purposeful
and intentional. The expression of toleration is a rational and conscious act even
if it is expressed through inaction. As sociologists have consistently pointed out,
refraining from engaging in an act is no less meaningful than the decision to
engage in one. It is also no less real in its consequences. Fourth, like concepts
such as ‘trust’ or ‘stigma’ the notion of toleration will normally exhibit a
number of different dimensions, which may not in themselves appear as
 Crime, community safety and toleration 101

complementary or consistent. As a multi-dimensional concept it will contain

ambiguities and uncertainties. Thus, any investigation of toleration will require
a multifaceted form of enquiry which will also need to be able to distinguish
between the object of toleration and the mode of expression.
Finally, toleration is an intrinsically moral concept. It implies exercising
judgement, taking responsibility and making choices between that which is
acceptable and unacceptable. It is the case that we may tolerate things which
we believe to be wrong, but this does not preclude the fact that our decision to
tolerate involves an element of moral choice (Bauman 1995).

The centrality of toleration to criminological investigation

A number of well-known criminological terms have all been associated with the
expression of intolerance. These include ‘shaming’, ‘censure’, ‘stigmatisation’
and ‘labelling’, and each of these concepts have been influential in criminology
in different periods (Goffman 1963; Schur 1971; Sumner 1990). Their relation
to the term toleration, which is arguably a more generic concept, is often
alluded to but rarely explained. At a more general level the concept of tolera-
tion is directly and indirectly associated with a number of central concepts in
criminology such as fairness, culpability, equity and responsibility which inform
the meta-theories of crime and justice in both everyday life and among criminal
justice agencies.
Over the past few years a number of criminological texts have touched in
various ways upon the issue of toleration. In particular, those which have
explored issues connected with punishment and sentencing have addressed the
relation between changing penal practices and climates of public opinion in
different periods. These investigations have been encouraged by a sense that
there is considerable agreement within and across advanced western societies in
relation to the appropriate forms and levels of punishment (Doob and Roberts
1988; Garland 1990; Young and Brown 1993).
Among these contributors a number have drawn on Norbert Elias’ (1994)
work on the ‘civilising process’. Elias’ focus is on how people have internalised
social constraint, become more self-disciplined and learned to manage their
feelings in a more discreet way. As a result of growing social interdependency
coupled with the formation of a centralised state authority which claimed a
monopoly on the legitimate use of violence, people in Europe have, he argues,
become less tolerant of overt demonstrations of violence. It has been claimed
that Elias’ work provides a non-economistic account of changing attitudes
towards public demonstrations of violence. In particular, his approach has been
used as a basis for explaining the demise of public executions and the develop-
ment of the modern prison (Franke 1992; Spierenburg 1984). Elias’ approach,
however, while containing suggestive observations on changing forms of tolera-
tion historically, has been charged with vagueness, evolutionism and a
one-dimensional understanding of social change (van Krieken 1989). Most
damaging for the Elisian approach is the claim that it is unable to effectively
102 Lynn Hancock and Roger Matthews

account for the reversals and ambiguities which have been identified in the
deployment of punishment historically and that it is not very useful in
explaining a number of contemporary developments, such as the decline of the
rehabilitative ideal, the emergence of the ‘new penology’ and the ‘new correc-
tionalism’ with its return to boot camps as well as the use of chain gangs
alongside the development of other ‘decivilising’ processes (Pratt 1998;
Vaughan 2000).
A more focused and grounded approach which has linked the issue of tolera-
tion to expressions of punitiveness has been presented by Leslie Wilkins and
Ken Pease (1987) who have suggested that one of the cultural determinants of
society’s penal climate is its relative egalitarianism. That is, they claim that the
greater the society’s tolerance of inequality, the more extreme the scale of
punishment is likely to be. Just as those at the top are seen to receive dispropor-
tionate rewards for their achievement so those at the bottom end of the social
scale who ‘fail’ are punished particularly severely. Thus, countries that have a
highly individualistic and competitive ethos, they argue, and in which there is a
high degree of inequality will tend to be comparatively severe in the use punish-
ment. Mobilising solid comparative evidence, however, to test this hypothesis is
difficult given the limited nature of the available data (Pease 1994). The
United States is the usual point of reference in debates over relative punitive-
ness, particularly since currently it has in excess of two million people
incarcerated. However, while there is evidence that the US tends to impose
longer sentences for a range of crimes, the use of incarceration itself is not, in
general, too dissimilar from European countries (Langham and Farrington
1998). Variations between European countries provide only limited support for
the hypotheses, while it is clear that the use of imprisonment, for example, in
different countries is also likely to be conditioned by other factors such as the
availability of non-custodial sanctions and the degree of social and political
stability. Nevertheless, it remains a relevant and plausible hypothesis.
One of the few attempts to analyse explicitly the role of tolerance in shaping
systems of punishment was the study by David Downes (1988) of the different
use of sanctions in England and Wales and in the Netherlands. Downes saw the
different approaches to the use of imprisonment and alternatives to custody in
these two countries as a function of ‘contrasts in tolerance’. He attributes the
relatively low use of incarceration in the post-war period in the Netherlands to
a greater level of tolerance than existed in England and Wales, although he
points out that Dutch tolerance does not involve a ‘blanket acceptance of all
manner of behaviour and opinion’. Rather, it is a selective strategy based on a
‘politics of accommodation’ which relies less on scapegoating and the exclusion
of deviants and instead provides a more integrative approach. Herman Franke
(1990) in a critical review of Downes’ Contrasts in Tolerance, emphasises that
tolerance is not of one piece and therefore it is inappropriate to talk about a
‘culture of tolerance’. Instead, there is a need, he argues, to examine the history,
the political context, the composition and orientation of criminal justice agen-
cies as well as the role of the media in influencing the exercise of tolerance.
 Crime, community safety and toleration 103

In relation to the construction of crime it has been pointed out by radical

realists that public attitudes, informal normative structures and sanctions are a
critical dimension of the ‘square of crime’ (Lea 1992; Matthews and Young
1992). While it is the case that victims as primary definers play a critical role in
transforming incidents or ‘problematic situations’ into potential ‘crimes’ the
decision to report will not only be conditioned by the gravity of the offence, the
willingness of the police to take it seriously or their ability to do something
about it, but will also be dependent on the level and nature of public tolerance.
The recent shift, for example, in attitudes towards interpersonal violence in
which child abuse, domestic violence, bullying and the like have been ‘redis-
covered’ reflects a changing framework of social attitudes which will affect
individual decisions to report incidents (Pahl 1985; Saraga, 1996). It will also
affect the sanctions imposed and the severity with which different kinds of
offences are viewed. Thus, as Steven Box and Chris Hale (1986) have pointed
out, it is essential when looking at the problem of crime to understand the
social and cultural processes through which different types of activity become
problematic. It is also necessary to understand why it is felt that certain activi-
ties can no longer be adequately dealt with informally and why the sanctions
which were previously deployed to control certain deviant activities are no
longer seen as being adequate. The importance of this point is that we should
not simply see the growing public intolerance of certain types of incivilities or
crimes as a consequence of the growth in crime, but rather regard both the
growth of crime and the changing attitudes to punishment as two sides of the
same process. This relation provides a useful corrective to those criminologists
who see shifts in public tolerance purely as a reaction to higher levels of crime.
Thus we might explain the steady increase in crime over the past two decades
and the intensification and extension of punishment as a product of the twin
effects of shifts in the nature of toleration. By the same token there is a need to
identify those activities which have been effectively decriminalised over time or
which have slipped down the ‘law and order’ agenda.
This two-way development is evident in the Crime and Disorder Act (1998)
which on one side transforms a number of incivilities into criminal offences and
simultaneously involves the intensification of the sanctions relating to each of
these activities. In a similar vein, we have seen in recent years on both sides of
the Atlantic how the ‘war against drugs’ has resulted in the criminalisation of a
range of previously tolerated drug-taking activity and the widespread use of
formal sanctions, particularly imprisonment (Tonry 1995). In this process the
concern across the political spectrum associated with widespread illicit drug use
can effectively be translated into forms of selective enforcement. Thus although
shifts in the nature of public toleration play a critical role in this process, and
may be fuelled by media campaigns, this growing intolerance may be formally
expressed through criminal justice agencies which have their own interests and
organisational constraints. Of central importance, therefore, is the relation
between informal and formal processes of regulation as well as the role of medi-
ating institutions in shaping the nature and expression of toleration.
104 Lynn Hancock and Roger Matthews

If we are correct in our claim that toleration is germane to the whole crimi-
nological enterprise, then why should it be that criminologists have historically
been reticent to engage with this concept? We can suggest three main reasons.
First, the issue of toleration is seen to be located in political philosophy which is
a discipline that remains relatively unexplored by criminologists. Although it is
the case that much of the more interesting criminological work in recent years
has been that which has vicariously lifted ideas and materials from other disci-
plines – particularly sociology, urban studies and cultural studies – successful
forays into the world of political philosophy have been less frequent. Second,
toleration does not lend itself easily to direct measurement, since its exercise
involves inaction as well as action. Since toleration is also a relational, complex
and multi-faced activity its study requires a more imaginative and sophisticated
form of analysis than is normally undertaken by criminologists. Third, an exam-
ination of toleration is likely to draw researchers into the perplexing areas of
norm creation, systems of classification, and probably the area in which they
feel most uncomfortable – morality (Bauman 1995; Cohen 1979).

The growing interest in public tolerance

Despite the reluctance to engage directly with the notion of toleration a
number of criminologists have begun to address the question of public attitudes
(Hough and Roberts 1999; Mirrlees-Black et al. 1996). Policy-makers and
funders who sense it is necessary to take some account of public attitudes are
increasingly turning to public attitude surveys. These surveys, however, are
fraught with methodological problems and are able at best to provide a ‘snap-
shot’ of public opinion and at worst provide a distorted one-dimensional picture
of what members of the public actually think (Herbst 1998; Norris 1999). This
shift in interest has been largely a consequence of external developments rather
than arising from pressures within the discipline itself. Among the major devel-
opments which have taken place in recent years are the changing role of the
national state, globalisation, increased political apathy and cynicism, growing
public insecurity, the fragmentation of communities and the breakdown of regu-
latory institutions, as well as growing anxieties about the direction of political
leadership.
We are currently witnessing a significant change in the organisation and role
of the national state. This involves a process in which decision-making is simul-
taneously becoming more global and international and more local. As a
consequence, the sovereign state has spread the responsibility for crime control
and community safety onto a range of statuary, private and voluntary agencies
while sending out the message that the state alone is not, and cannot, be
responsible for preventing and controlling crime (Garland 1996; Jessop 1994).
There has also been a growing involvement of citizens in crime prevention and
crime control. This, however, is not just a process of the state and other agen-
cies pressurising local communities to take responsibility for their own security;
but also represents a new social movement in which local social networks
 Crime, community safety and toleration 105

attempt to actively defend their communities from the effects of globalisation.

The net result is that a significant proportion of citizens are involved, albeit in
some cases on a token level, in crime prevention and community safety
programmes through neighbourhood watch, tenants associations, block watch,
citizen patrols, local support groups and other community-based organisations.
An indication of this can be found in David Bayley’s (1994) findings, that in
New York organised self-protection activities on the part of citizens have mush-
roomed with over 151,000 people involved in crime prevention programmes in
1985. This involvement reflects a deeply felt need to supplement the usual
police protection in the context of growing concerns about public security.
Similar trends can be found in Britain. There are now over 150,000 neighbour-
hood watch schemes operating in England and Wales covering approximately
10 million households, although the involvement of these members may in
some cases be fairly token. There are a growing number of tenants organisa-
tions, community liaison groups, citizen patrols and the like who are involved
in crime control and improving public security (Hancock 2001). The growing
involvement of local community groups has, however, been met with ambiva-
lence by the state authorities, since the attempts to encourage ‘active citizens’
has been tempered by the spectre of vigilantism (Johnston 1996).
The devolving of responsibilities to local authorities following the publica-
tion of the Morgan Report in 1991 has given crime prevention and crime
control a more differentiated and local focus. One consequence of this strategy,
however, is that it has promoted interest in the experiences and attitudes of
local populations. It also creates new demands and expectations from local resi-
dents which need to be accommodated. There is the prospect that within this
emerging set of relations changes are likely to occur in the prioritisation of
crime and disorder and that the newly established community safety groups will
reflect local priorities and concerns as well as shifts in toleration.
Communities, however, have become more fragmented as a result of
economic restructuring. As the routines and habits of everyday life are under-
mined through the introduction of new forms of flexible accumulation the basic
trust in the security of the world has been thrown into disarray (see Young in
this volume). In these conditions the willingness of citizens to tolerate difficult
and deviant activities is likely to be curtailed, and creates a climate which
encourages forms of scapegoating, particularly in the form of racism (Cohen
1998). It is often as the conditions for the expression of intolerance grow that
official calls for greater tolerance arise.
The development of globalisation and the advent of the ‘information
society’ has at the same time increased the complexity of decision-making and
the plurality of values and interests which have to be taken into account such
that the actors shaping our political life are themselves diminishing in stature.
The lack of any clear profile amongst different political parties is associated
with growing political apathy and cynicism. There is a consequential legitima-
tion deficit and it becomes increasingly difficult to mobilise the population
behind governmental policies. At this juncture there is seen to be a need to
106 Lynn Hancock and Roger Matthews

reinvigorate flagging institutions and to increase confidence, particularly in the

criminal justice process which plays an increasingly important practical and
symbolic role in late modernity (Hirst 1997).
The combined effect of these related processes has been to increase interest
in public attitudes and changing forms of toleration. However, as yet the inves-
tigation and understanding of how these attitudes are formed and transformed
remains underdeveloped.

Why tolerate?
It is clear that whatever one’s beliefs are that one cannot express intolerance
towards everything one dislikes or disapproves of. Toleration, therefore, is
necessarily selective. The critical issues are how toleration is constructed,
changed and expressed. Strategic decisions are often made in relation to prag-
matic and moral considerations. The problem of analysis is that pragmatism and
moralism tend to be interconnected. The main pragmatic consideration, often
suggested by utilitarians, is that the benefits of toleration may outweigh the
harms. Toleration may also enable us to mitigate or avoid damaging social
conflict. The difficult question is, however, how do we make a distinction
between acts which should be prevented and punished and those that should be
permitted and tolerated. Harm itself is not a sufficient criterion of intolerance
since the notion of harm is open to wide interpretation and comes in a multi-
tude of different forms, while the decision about whether and how to respond
will be conditioned by a range of other moral and practical considerations.
Primary amongst these considerations is the relation between the activities
to be tolerated and the level of social support which is available. According to
Richard Taub and his colleagues (Taub et al. 1984) in their study of community
transition in Chicago, the exercise of tolerance is conditioned by the quality
and the quantity of local amenities. The factors, they suggest, that will influ-
ence the capacity of different individuals and groups to tolerate different
activities will include the organisation and density of housing, the seriousness of
other local problems and the range and quality of amenities. It is also suggested
that the level of tolerance is not just a function of the level of absolute depriva-
tion in an area but often a consequence of relative deprivation. That is,
intolerance is seen to arise from the gap between what people have got and
what they expect (Campbell 1991). The focus on relative deprivation may go
some way to explain the apparent paradox which has been reported that it is
those groups located near the top as well as those at the bottom of the social
scale who exhibit the greatest level of punitiveness (Brillon 1988).
In everyday discussions of toleration reference is often made to ‘thresholds’
and to ‘the limits of tolerance’, which suggests that there are critical points in
the exercise of tolerance at which a qualitative change occurs. In this point of
transition it would seem that activities which were once tolerated take on a
different significance. The implications of this observation are that it is not
only the number of incidents which is important but the point at which a
 Crime, community safety and toleration 107

threshold of toleration is crossed. Thus it may be the case that seven or eight
prostitutes working in a certain locality may be acceptable and although local
residents may not particularly approve of the activity they are willing to tolerate
it. However, at a certain point the visibility of, say, twelve women regularly
working on the streets transforms the nature of the activity such that it is seen
to ‘swamp’ or change the neighbourhood in a way that local residents are no
longer willing to tolerate, and they call on the authorities to do something
about it. This scenario appears by and large to have occurred in various parts of
Britain during the 1980s, and although the relevant authorities were slow to
respond to growing public pressure many residents groups took it upon them-
selves to deal with the problem (Matthews 1993). The other important
implications of this observation are that firstly we need to give greater attention
to the impact of crime and disorder in different areas and that, second, we need
to identify the thresholds of tolerance. In this way we might become less
concerned with the rise and fall of recorded offences – or with strategies which
simply aim to reduce crime and disorder or other problems by an arbitrary
amount – and instead develop a greater regional and local sensitivity to varia-
tions in the thresholds of toleration.

Liberalism and communitarianism

The articulation and defence of toleration has historically been associated with
the classic liberal theorists such as Locke and Hobbes and more recently the
writings of Rawls (1971) and Dworkin (1977). In essence, the standard liberal
defence of toleration is centred around the notion of respect for other persons
and the defence of individual liberty. This is normally expressed in terms of the
protection of rights. In this process the state is seen as a neutral arbiter between
competing moralities. Therefore, in principle the state should not prescribe any
particular moral or religious view but allow equal recognition to all citizens.
The emphasis on respect for other persons is based upon the principle which
maintains that citizens should be able to make choices about how they live and
be able to act as autonomous and rational agents. Thus for liberals there is a
private sphere of individual autonomy which the state should not interfere with
unless an individual inflicts harm on another individual. The protection of a
separate private sphere is seen to be a necessary condition for the maintenance
of individual freedom.
This familiar liberal position has been the object of a sustained critique by
communitarian thinkers who have taken issue with the individualistic premise of
the position as well as the preoccupation with rights and the associated concep-
tions of liberty, freedom and justice. This communitarian critique has gained
some ascendancy since the election of New Labour in 1997, and the govern-
ment’s thinking has been greatly influenced by the work of Amitai Etzioni (1993;
1997). This critique has direct implications for the status of the theory and prac-
tice of toleration since the distancing from liberalism suggests by implication a
rejection of, or at least a shift away from, the liberal defence of toleration.
108 Lynn Hancock and Roger Matthews

In contrast to liberalism the communitarian emphasises the primacy of the

community, claiming that the self is built up intersubjectively. Communities,
therefore, are more than a collection of autonomous individuals and they repre-
sent more than the sum of their parts. Communitarians have taken issue with
the liberal emphasis on individual rights. Some have argued that a rights-based
policy is ultimately unsustainable since claims about competing rights are them-
selves difficult to resolve, and that these policies find themselves in the awkward
position of deciding which rights to defend and which to prioritise over others,
since the exercise of one person’s rights often involves the infringement of the
rights of others. Some communitarians accept that there are certain basic rights
which should be defended such as the right to a fair trial, the right to be
presumed innocent until proven guilty, and so forth, but most would endorse
Etzioni’s contention that liberals tend to place too much emphasis on individual
rights and not enough on social obligations and duties. Communitarians also
take issue with the liberal conception of freedom as a state of non-interference
of the solitary individual. For the communitarian freedom is a social achieve-
ment through the guarantee of protection. It is not a state of chance but a
process in which each citizen is equally protected by law (Braithwaite and Pettit
1990). The emphasis on equality draws attention to the importance of treating
both the offender and the victim equally and in ways which are constructive but
not overly intrusive (Walzer 1983).
One of the few criminologists who has examined the notion of tolerance
from a communitarian perspective is John Braithwaite. In his influential study,
‘Crime, Shame and Reintegration’ (1989), he provides an explicit critique of
liberal notions of toleration and argues that liberal tolerance can be counterpro-
ductive. For example, when parents allow their children to freely use illicit
drugs, or engage in various forms of anti-social or undesirable behaviour without
receiving any sanction, it can lead to the escalation and intensification of these
problems. Braithwaite’s argument is reminiscent in certain respects of Herbert
Marcuse’s (1965) critique of ‘pure tolerance’, which he suggests can turn into
‘repressive tolerance’ when it does not address or remove social inequalities, but
instead implicitly endorses and perpetuates them.
In contrast to liberalism, Braithwaite argues that: ‘Low crime societies are
societies where people do not mind their own business, where tolerance of
deviance has definite limits, where communities prefer to handle their own
crime problems rather than hand them over to professionals’ (1989: 8)
Moralising control through informal networks, he argues, is likely to be more
effective than formal forms of ‘repressive’ social control which denies human
dignity, limits autonomy and serves to confirm deviance as a master status. The
aim for Braithwaite is to mobilise shaming strategies not simply in order to stig-
matise deviants but instead to reintegrate them whenever possible back into the
community. The important point here is that greater intolerance does not
necessarily lead to greater punitiveness.
In opposition to the atomistic vision of order which is evident in strands of
liberalism, it is argued by communitarians that the community structure itself
 Crime, community safety and toleration 109

shapes local crime rates by defining what is acceptable behaviour and by estab-
lishing normative structures which guide both the definitions of ‘problematic
situations’ and how they are categorised. A communitarian notion of toleration
is bound up with the defence of the community and the decision of whether to
exercise tolerance or not will be conditioned by how valuable the community
under threat is. Therefore, actions which may appear to be fairly innocuous can
have a cumulative impact on the community and consequently can become the
object of intolerance. However, in promoting intolerance the communitarian
has to give due consideration to the level of harm inflicted as well as the
capacity of the community to absorb or deflect these ‘anti-social’ activities as
well as considering the consequences of stigmatising or marginalising the
offender since this may cumulatively weaken the community. It is also necessary
for the communitarian to distinguish between interventions which are practi-
cally designed to improve community safety and those interventions which are
more symbolic and are designed to regenerate moral obligations between citi-
zens (Crisp 1992; Gutman 1985).
In general, the communitarian critique can be seen as providing a corrective
to certain liberal conceptions of the autonomous self-propelling individual,
although liberalism has not been without its defenders (Kymlicka 1989). As
Allen Buchanan (1989) has pointed out, this ongoing debate between liber-
alism and communitarianism has operated on a number of different levels and
that in trying to think through these important issues it is useful to distinguish
between radical, moderate and cautious communitarians. It is only in its radical
version that communitarianism rejects outright the significance of individual
rights and the liberal conception of freedom. In its more moderate version it
concedes the importance of certain rights, but denies the priority which the
liberal attributes to them. In its more cautious variant it urges those who value
community not to abandon the framework of individual rights, but rather to
appreciate the role which individual rights can play in protecting communities
under certain historical conditions. In opposition to the radical communitarian
he argues that the defence of individual rights of association, to pursue religious
beliefs and the right to political participation need to be defended and that:

Given the apparent diversity of the conditions of human flourishing, the

pronouncement that the best life for all (or even most humans) requires
participation in the most inclusive forms of political organisation is sheer
dogmatism in the absence of a well-defended, highly particularistic, and
absolutist theory of objective good.
(Buchanan 1989: 859)

In the absence of such a point of reference individual rights allow individuals

who are dissatisfied with current forms of community or who find existing
communities unduly restrictive or oppressive to develop alternatives. Closing
such a possibility raises for the liberal the spectre of totalitarianism.
Moreover, as Axel Honneth (1995) has suggested, while it may be the case
110 Lynn Hancock and Roger Matthews

that self-confidence and self-identity are built up intersubjectively, the estab-

lishment of legal rights is important in recognising the dignity, autonomy and
capacity of persons to act as morally responsible legal agents. He also argues
that crime itself can be seen as a struggle of interests or as a clash of identi-
ties. Crime involves not only a material loss or a physical violation but an
attack upon individual identity. If crime only involved a material or physical
deficit it would not be so socially important. It is precisely because it involves
issues of recognition and identity that it takes on such a degree of signifi-
cance, particularly in a period in which social recognition and identity
formation are becoming increasingly problematic (Fraser 1997).
A major problem with the communitarian position in terms of crime preven-
tion and community safety is that it is notoriously difficult to establish
community strategies to reduce crime in high crime areas and probably even
more difficult to sustain involvement in low crime areas. Further, social control
tends to be maintained in the more affluent suburban communities by social
avoidance, privatisation, low levels of interaction and through ‘moral mini-
malism’ (Hope 1995). This notion of moral minimalism is reminiscent of Iris
Young’s (1990) depiction of modern urban communities as being composed of
lightly engaged strangers. This state of affairs raises particular problems for
communitarianism since it would appear that there is a growing emphasis upon
difference and diversity within contemporary urban environments rather than
on close-knit communities. Where new collectivities arise there is little
evidence that they approximate to the notion of the ideal community. In fact,
the individualism/community opposition is becoming less tenable and masks
rather than elucidates the nature of modern urban groupings. Rather than
seeing the individualism/community opposition as exhaustive, Iris Young sees
them as mutually sustaining:

Too often contemporary discussion of these issues sets up an exhaustive

dichotomy between individualism and community. Community appears in
the oppositions individualism/community, separated self/shared self,
private/public. But like most such terms, individualism and community
have a common logic underlying their polarity, which makes it possible for
them to define each other negatively. Each entails a denial of difference
and a desire to bring multiplicity and heterogeneity into unity, though in
opposing ways. Liberal individualism denies difference by positing the self
as a solid, self-sufficient unity, not defined by anything or anyone other
than itself. Its formalistic ethic of rights also denies difference by bringing
all such separated individuals under a common measure of rights.
Proponents of community, on the other hand, deny difference by positing
fusion rather than separation as the social ideal. They perceive the social
subject as a relation of unity or mutually composed by identification and
symmetry among individuals within a totality. Communitarianism repre-
sents an urge to see persons in unity with one another in a shared whole.
(Young 1990: 229)
 Crime, community safety and toleration 111

It is the ideal of community, she argues, which by encouraging a sense of mutual

identification with some persons serves to problematise difference. The fear of
difference can promote racism and classism in society and fosters forms of exclu-
sion and marginalisation, while simultaneously suppressing diversity.
Rather than seeing a movement towards the ideal of ‘community’ we appear
to be witnessing the emergence of new groupings and forms of neo-tribalism
centred around temporary forms of solidarity and political organisation which
challenge rather than endorse the legitimacy of traditional moralities (Maffesoli
1996; Sennett 1991). In these ‘tribes’ membership is easily revocable and is not
tied to long-term obligations. Their existence is transient and always in flux. In
these transient urban networks the very notion of ‘community’ is becoming
increasingly problematic. In the context of increasing contingency, ambiva-
lence and fragmentation the search for ‘community’ appears more hopeless and
unrealistic. The identification of consensus becomes more elusive and the
ability to mobilise universal truths in order to sanction, humiliate or stigmatise
becomes increasingly difficult. The construction of order begins to look more
artificial and fragile. The protection of the ‘community’ as a geographically-
bounded space of local networks, which is increasingly invoked by
policy-makers as a justification for cracking down on various forms of anti-social
behaviour, appears to coincide with the transformation and dissolution of these
local networks.
Alongside the reorganisation and differentiation of communities the
communitarians face a related problem. This is the demise of regulatory and
socialising institutions. In much communitarian writing there is to be found a
backward looking and often romantic vision of these institutions – the family,
the school, the police, the prison and indeed the state. It is no accident,
however, that they have repeatedly been referred to by commentators from
across the political spectrum, as being in a state of ‘crisis’. The implications of
this assessment is that the pivotal traditional mechanisms through which social
order and discipline are normally maintained are in disarray.
It is in this context that the repeated references to the relation between
single parent households and delinquency should be seen. The suggestion that
these parents are often irresponsible and feckless combined with calls for
making parents take more responsibility for their children, overlooks the reality
that such families lack resources to impose ‘discipline’ on children who live in a
world in which there is great uncertainty about what constitutes an appropriate
form of discipline. Therefore, attempts to salvage the ‘cornflake’ family are
unlikely to work in the longer term and instead of perpetuating a backward-
looking view of the role of weak and outdated institutions, the critical political
question is how to establish the appropriate socialising and disciplinary mecha-
nisms in a post-Fordist and postmodern urban environment characterised by
new forms of individualism and collectivism.
The paradox of the dominant forms of political thinking on both sides of the
Atlantic is that they are guided by principles which are simultaneously wildly
utopian and nostalgic and appear increasingly out of touch with the changing
112 Lynn Hancock and Roger Matthews

realities of urban life, while espousing a mixture of opportunistic and pragmatic

policies couched in the language of managerialism.
There is a lingering problem about the scope, legitimacy and accountability
of state institutions which classic liberalism finds difficult to address because it
was designed to protect society from the state, rather than deal with the prob-
lems of liberty in a state and civil society dominated by large hierarchical
institutions (Hirst 1997). Thus the attempts of liberalism to shape and develop
state power and improve accountability have been largely ineffective. The
weakness of communitarianism, on the other hand, in this respect and its failure
to confront and reform institutions derives from its desire to urge a consensus
on a deeply divided society with plural and competing values.
By contrast to those ‘moral communitarians’ who see the solution to problems
of crime and community safety through the creation of informal control mecha-
nisms and particularly through shaming, Tim Hope (1995) has argued there are
two dimensions along which crime control strategies work in communities:

The first is a ‘horizontal’ dimension of social relations among individuals

and groups sharing a common residential space. This dimension refers to
the often complex expressions of affection, loyalty, reciprocity, or domi-
nance amongst residents, whether expressed through informal relationships
or organised activities. Second, there is a ‘vertical’ dimension of relations
that connect local institutions to sources of power and resources in the
wider civil society of the locality is acknowledged to be a part.
(Hope 1995: 24)

Although these two dimensions are related it is the case that:

While the principal mechanisms for maintaining local order may be

expressed primarily through the horizontal dimension, the strength of the
expression – and hence its effectiveness in controlling crime – derives, in
large part, from the vertical connections that residents of localities have to
extracommunal resources.
(Hope 1995: 24)

Although research has shown that community organising in some locations can
make a difference (Sampson et al. 1997), the ability of residential communities
to withstand the impact of different forms of crime and incivilities will depend
in large part on the available resources and their access to support networks
(Cullen 1994; Matthews 1992).
If it is the case that the cautious communitarian can establish some degree of
compromise with the liberals over rights, there still remains a major split
between those communitarians who prioritise the community and community
controls and those liberals who see crime control and community safety as an
individual responsibility within a market-driven framework. This tension is
located at the heart of current government thinking on ‘law and order’. In the
 Crime, community safety and toleration 113

post-welfarist society, the emphasis on individual self-help and ‘freedom’

consists in extracting oneself from the supposed restraints and debilitating
consequences of welfare. Government policy encourages the individualisation
and commodification of security, offering only minimum guarantees, above
which individuals are encouraged to take responsibility for themselves. This
neo-liberal notion of ‘freedom’ and responsibility, in turn, influences notions of
personal morality and the dependent subject of welfarism is replaced by the
juridical subject of neo-liberalism (Pratt 1996). It is in this context that we can
begin to make sense of the contradictory policy initiatives which on one side
establish agencies to reduce social exclusion, while at the same time implement
a reduction in welfare payments to marginalised populations: and how at one
moment the government is advocating minimum mandatory sentences for
burglary, drug dealing and violence, while at another it is generously funding
programmes of restorative justice and community programmes for disaffected
youth.
The tension between market-based policies and more socially-oriented poli-
cies creates further difficulties for ‘balancing’ communitarian approaches with
liberalism. Most communitarians have a commitment to distributive justice but
often defer to market mechanisms for the overall allocation of resources. This
has resulted in a growing divide between the rich and the poor as well as
growing regional disparities. As inequality increases, the basis of the communi-
tarian vision begins to collapse since it undermines the realisation of both social
and distributive justice, while creating new conflicts and antagonisms (Walzer
1983). The major difficulty for policy-makers is the formulation of realistic
criteria by which the balance between rights and duties could be struck. Despite
the proliferation of official pronouncements, reports and glossy publications
there are no identifiable criteria of how and where to intervene, when to express
intolerance, how to balance up social protection and individual rights and how
to decide on the appropriate response (see Braithwaite and Pettit 1990) .
Some recognition of these difficulties has been expressed by Anthony
Giddens (1998) in his elaboration of a ‘Third Way’, which constitutes an
attempt to move beyond neo-liberalism and old-style statism and collectivism.
Although he recognises that the ideal of ‘community’ is no longer a realisable
objective and that the modern world is characterised by diversity, increasing
‘ontological insecurity’ and rising crime, his policy programme involves a
mixture of public order and community policing as well as the development of
local multi-agency interventions incorporating a range of state and non-state
agencies, with a particular focus on the expansion of the voluntary sector. In
these proposals he seems to be unaware of the fact that this is precisely the
mixture of strategies which has been employed over the last decade or so under
both New Labour and previous conservative governments and that it is
precisely the effectiveness of this combination of strategies which is currently in
question (see Crawford in this volume).
Thus although it may be possible to develop a theoretical compromise by
adopting a more cautious form of communitarianism the danger is that in the
114 Lynn Hancock and Roger Matthews

absence of any clear principles or criteria for intervention, government policy is

seen as either oscillating uncontrollably between a free-market mechanism and
the promotion of idealised communities or alternatively falling between these
two objectives. Ultimately, however, it has been suggested that the difficulty of
working out such a conceptual and political compromise has been that much
more difficult because the social relations and social structures around which
these divisions have been historically articulated are themselves in a process of
transformation and reconstruction. Thus, it would appear that this long-
standing political antagonism between liberalism and communitarianism is
being overtaken by events and that there is a growing sense in which classical
liberalism is unable to defend individual liberty against the growing power of an
unchecked corporate and managerial elite; while old-style communitarianism
seeks to restore ailing institutions by trying to change people’s values in an
attempt to avoid engaging in major organisational and structural reforms.
Paradoxically, however, the current drive towards ‘modernisation’ encourages
greater diversity and flexibility, the fragmentation of traditional communities,
new forms of social exclusion, the problematisation of identities, the rise of neo-
tribalism and creates a growing lack of moral consensus in a world in which
ever greater percentages of the population are becoming ‘problems’ (Bauman
1995; Young 1999).

Conclusion
In general, a number of social and political commentators seem to agree that
toleration is an important element in the functioning of liberal democracies. But
debates continue about where to draw the line between tolerance and intoler-
ance and how such ‘moral’ decisions are linked to images of the ‘good society’.
Increasingly, criminologists have become aware that toleration plays an impor-
tant, if not pivotal, role in the construction of crime and in influencing crime
control policies. As we move into a situation in which the public are an increas-
ingly important point of reference in providing information, co-operation and
legitimacy within a restructured system of regulation a greater awareness of what
the public want and what they are willing to accept has gained greater promi-
nence. Alongside these organisational imperatives there have been growing
demands for security from different social groups and a greater willingness to
participate, at some level, in the development of community safety.
These developments have opened the door to greater community involve-
ment in decision-making and to the introduction of new forms of accountability.
The shifting responsibility for crime control and community safety to local
authorities has been identified as a ‘sea change’ or a ‘watershed’ in policy devel-
opment in Britain, introducing new forms of governance (Crawford 1997).
There is now a clear sense of the limits of the nation-state and that there is no
way back to the centralised forms of regulation which characterised previous
periods. The watch words are now diversity, flexibility, cost-effectiveness and
risk management.
 Crime, community safety and toleration 115

Within this changing framework the relation between tolerance, forms of

punishment, the distribution of justice and security become more problematic.
There is an issue at one level of how to balance tolerance and intolerance,
rights and obligations, as well as guilt and justice (Cohen 1979). There are no
clear principles which tell us how such a balance is to be struck. At a second,
and related, level there is an issue of how to identify, analyse and express forms
of toleration.
The pursuit of tolerance and intolerance is not mutually exclusive. There is
also nothing intrinsically negative or undesirable about intolerance. Intolerance
can be instructive and constructive, just as the application of tolerance can be
damaging and repressive in certain situations. In fact, to argue for tolerance in
the abstract can be vacuous and counterproductive. If it is the case, however,
that the aim of government policy is to argue for a different balance between
tolerance and intolerance or for an increase in specific forms of toleration then
a number of material, relational and social issues need to be addressed. On one
level material inequalities and deprivations need to be reduced. This is an issue
of distributive justice. At another level issues of recognition, solidarity and
identity require serious consideration. These issues are bound up with notions of
social justice. The spheres of distributional and social justice are directly and
indirectly linked and are conceptually connected through notions of ‘fairness’ in
the distribution of rewards and deserts. These notions, in turn, will influence
the expression of toleration.
It has also been suggested that seeing the business of crime control and commu-
nity safety through an examination of toleration could provide a very different
vantage point for analysing and evaluating policy developments. However, the
mechanisms for investigating and channelling public attitudes remain underdevel-
oped. The current preoccupation with the relatively arbitrary reductions in
officially recorded crime rates and the emphasis on developing policy through the
cumulative production of ‘evidence’ constructed through a series of uncoordi-
nated, patchy and idiosyncratic forms of research stands in contrast to the
suggestion that crime and crime control policies are profoundly influenced by
shifts in public tolerance, however restricted its expression might be.
The approach of the present British Labour government, we have argued,
involves an uneasy and unresolved mix of liberal and communitarian ideas on
toleration, rights and justice. These two philosophies were always in themselves
problematic, but it has been suggested that the changing problem of security
facing different social groups and the tranformation of ‘communities’ has
rendered these modes of thinking even more problematic. The current impasse
will require some serious rethinking. It is a problem which involves three
aspects, first, a change in the nature of risks and hazards which individuals and
groups face; second, a change in the social relations, social networks and associ-
ated identities; and third, it involves the changing nature of the socialising and
regulatory mechanisms which are in place to shape and control these develop-
ments. The interrelated nature of these problems makes the possibility of a
‘quick fix’ resolution increasingly untenable and calls instead for a fundamental
116 Lynn Hancock and Roger Matthews

re-examination of crime and control and this in turn, we have suggested, will
involve the mobilisation of certain basic organising and sensitising concepts
such as toleration.

References
Bauman, Z. (1995) Life in Fragments: Essays in Postmodern Morality, Oxford: Blackwell.
Bayley, D. (1994) Police for the Future, New York: Oxford University Press.
Box, S. and Hale, C. (1986) ‘Unemployment, Crime and Imprisonment: The Enduring
Problem of Prison Overcrowding’, in R. Matthews and J. Young (eds) Confronting
Crime, London: Sage.
Braithwaite, J. (1989) Crime, Shame and Reintegration, Cambridge: Cambridge University
Press.
—— and Pettit, P. (1990) Not Just Deserts: A Republican Theory of Criminal Justice,
Oxford: Clarendon.
Brillon, Y. (1988) ‘Punitiveness, Status and Ideology in Three Canadian Provinces’, in
N. Walker and M. Hough (eds) Public Attitudes to Sentencing, Aldershot: Gower.
Buchanan, A. (1989) ‘Assessing the Communitarian Critique of Liberalism’, Ethics, 99
(July): 852–82.
Burney, E. (1999) Crime and Banishment, Winchester: Waterside Press.
Campbell, A. (1991) The Sense of Well Being in America, New York: McGraw Hill.
Cohen, M. (1998) Culture of Intolerance, New Haven: Yale University Press.
Cohen, S. (1979) ‘Guilt Justice and Tolerance: Some Old Concepts for a New Crimi-
nology’, in D. Downes and P. Rock (eds) Deviant Interpretations, Oxford: Martin
Robertson.
Crawford, A. (1997) The Local Governance of Crime: Appeals to Community and Partner-
ships, Oxford: Clarendon.
—— (1998) Crime Prevention and Community Safety, London: Longman.
Crisp, R. (1992) ‘Communitarianism and Toleration’, in J. Horton and P. Nicholson
(eds) Toleration: Philosophy and Practice, Aldershot: Avebury.
Cullen, F. (1994) ‘Social Support as an Organising Concept for Criminology’, Justice
Quarterly, 11(4): 527–59.
Dennis, N. (1997) (ed.) Zero Tolerance: Policing in a Free Society, London: Institute of
Economic Affairs.
Doob, A. and Roberts, J. (1988) ‘Public Punitiveness and Public Knowledge of the Facts:
Some Canadian Surveys’, in N. Walker and M. Hough (eds) Public Attitudes to
Sentencing, Aldershot: Gower.
Downes, D. (1988) Contrasts in Tolerance: Post-war Penal Policy in the Netherlands and in
England and Wales, Oxford: Clarendon.
Dworkin, R. (1977) Taking Rights Seriously, Cambridge, Mass: Harvard University Press.
Elias, N. (1994) The Civilising Process, Oxford: Basil Blackwell.
Etzioni, A. (1993)The Spirit of Community, New York: Crown Publishers.
—— (1997) The New Golden Rule, London: Profile Books.
Foucault, M. (1979) History of Sexuality: Volume 1, Allen Lane: Penguin.
Franke, H. (1990) ‘Dutch Tolerance: Facts and Fables’, British Journal of Criminology,
30(1): 81–94.
—— (1992) ‘The Rise and Decline of Solitary Confinement’, British Journal of Crimi-
nology, 32(2): 125–43.
 Crime, community safety and toleration 117

Fraser, N. (1997) Justice Interruptus: Critical Reflections on the Post-Socialist Condition, New
York: Routledge.
Garland, D. (1990) Punishment and Modern Society, Chicago: University of Chicago
Press.
—— (1996) ‘The Limits of the Sovereign State: Strategies of Control in Contemporary
Society’, British Journal of Criminology, 36(4): 445–71.
Giddens, A. (1998) The Third Way, Cambridge: Polity Press.
Goffman, E. (1963) Stigma: Notes on the Management of a Spoiled Identity, Englewood
Cliffs, NJ: Prentice Hall.
Gutman, A. (1985) ‘Communitarian Critics of Liberalism’, Philosophy and Public Affairs,
14(3): 308–22.
Hancock, L. (2001) Community, Crime and Disorder: Responses to Change in Urban Neigh-
bourhoods, London: Palgrave.
Harrison, R. (1992) ‘Tolerating the Offensive’, in J. Horton and P. Nicholson (eds) Toler-
ation: Philosophy and Practice, Aldershot: Avebury.
Herbst, S. (1998) Reading Public Opinion, Chicago: University of Chicago Press.
Hirst, P. (1997) From Statism to Pluralism, London: UCL Press.
Home Office (1998) Annual Report, Cm. 3908, London: HMSO.
Honneth, A. (1995) The Struggle for Recognition, Cambridge: Polity Press.
Hope, T. (1995) ‘Community Crime Prevention’, in M. Tonry and D. Farrington (eds)
Building a Safer Society: Strategic Approaches to Crime Prevention, volume 19, Chicago:
University of Chicago Press.
Horton, J. and Nicholson, P. (1992) (eds) Toleration: Philosophy and Practice, Aldershot:
Avebury.
Hough, M. and Roberts, J. (1999) ‘Sentencing Trends in Britain: Public Knowledge and
Public Opinion’, Punishment and Society, 1(1): 11–27.
Hughes, G. (1998) Understanding Crime Prevention, Buckingham: Open University Press.
Jessop, B. (1994) ‘Post-Fordism and the State’, in A. Amin (ed.) Post- Fordism: A Reader,
Oxford: Blackwell.
Johnston, L. (1996) ‘What is Vigilantism?’, British Journal of Criminology, 36(2): 220–36.
Kymlicka, W. (1989) Liberalism, Community and Culture, Milton Keynes: Open Univer-
sity Press.
Langham, P. and Farrington, D. (1998) Crime and Justice in the United States and in
England and Wales, 1981–96, Bureau of Justice Statistics, Washington DC: US
Department of Justice.
Lea, J. (1992) ‘The Analysis of Crime’, in J. Young and R. Matthews (eds) Rethinking
Criminology: The Realist Debate, London: Sage.
McIntyre (1981) After Virtue, London: Duckworth.
Maffesoli, M. (1996) The Time of the Tribes: The Decline of Individualism in Mass Society,
London: Sage.
Marcuse, H. (1965) ‘Repressive Tolerance’, in R. P. Wolff, B. M. Moore Jr. and H.
Marcuse (eds) A Critique of Pure Tolerance, Boston: Beacon Press.
Matthews, R. (1992) ‘Replacing Broken Windows: Crime Incivilities and Urban
Change’, in R. Matthews and J. Young (eds) Issues in Realist Criminology, London:
Sage.
—— (1993) Kerb-Crawling, Prostitution and Multi-Agency Policing, Paper 43 Crime
Prevention Unit Series, London: Home Office.
—— and Young, J. (1992) ‘Reflections on Realism’, in J. Young and R. Matthews (eds)
Rethinking Criminology: The Realist Debate, London: Sage.
118 Lynn Hancock and Roger Matthews

Mendus, S. (1989) Toleration and the Limits of Liberalism, London: Macmillan.

Mirrlees-Black, C., Mayhew, P. and Percy, A. (1996) The 1996 British Crime Survey,
England and Wales, London: Home Office.
Moynihan, D. (1993) ‘Defining Deviance Down’, American Scholar, 62 (Winter): 17–30.
Muncie, J. (1999) ‘Institutional Intolerance: Youth Justice and the 1998 Crime and
Disorder Act’, Critical Social Policy, 19(2): 147–75.
Newey, G. (1999) Virtue, Reason and Toleration: The Place of Toleration in Ethical and
Political Theory, Edinburgh: Edinburgh University Press.
Norris, P. (1999) Critical Citizens: Global Support for Democratic Governance, New York:
Oxford University Press.
Offe, C. (1996) Modernity and the State: East, West, Cambridge: Polity Press.
Pahl, J. (1985) Private Violence and Public Policy, London: Routledge and Kegan Paul.
Pease, K. (1994) ‘Cross-national Imprisonment Rates’, in R. King and M. Maguire (eds)
Prisons in Context, Oxford: Clarendon.
Podolefsky, A. and Dubow, F. (1981) Strategies for Community Crime Prevention: Collective
Responses to Crime in Urban America, Springfield, Illinois: Charles C. Thomas.
Pratt, J. (1996) ‘Reflections on Recent Trends in the Punishment of Persistence’, Crime,
Law and Social Change, 25: 243–64.
—— (1998) ‘Towards the Decivilising of Punishment’, Social and Legal Studies, 7(4):
487–515.
Rawls, J. (1971) A Theory of Justice, Cambridge, Mass: Harvard University Press.
Sampson, R. J., Raudenbush, S. W. and Earles, F. (1997) ‘Neighborhoods and Violent
Crime: A Multilevel Study of Collective Efficacy’, Science 277 (August): 918–24.
Sandal, M. (1982) Liberalism and the Limits of Justice, Cambridge: Cambridge University
Press.
Saraga, E. (1996) ‘Dangerous Places: The Family as a Site of Crime’, in J. Muncie and E.
McLaughlin (eds) The Problem of Crime, London: Sage.
Schur, E. (1971) Labelling Deviant Behaviour, New York: Harper and Row.
Sennett, R. (1991) The Conscience of the Eye, New York: Alfred A. Knopf.
Skogan, W. G. (1988) ‘Community Organizations and Crime’, in M. Tonry and N.
Morris (eds) Crime and Justice a Review of Research, volume 10, Chicago: University of
Chicago Press.
Spierenburg, P. (1984) The Spectacle of Suffering, Cambridge: Cambridge University
Press.
Straw, J. (1999) Foreword to the ‘Government’s Crime Reduction Strategy’, London:
Home Office.
Sumner, C. (1990) Censure, Politics and Criminal Justice, Milton Keynes: Open Univer-
sity Press.
Taub, R., Taylor, D. and Dunham, J. (1984) Paths of Neighbourhood Change: Race and
Crime in Urban America, Chicago: University of Chicago Press.
Tonry, M. (1995) Malign Neglect: Race Crime and Punishment in America, New York:
Oxford University Press.
Turner, M. and Cullen, F., Sundt, J. and Applegate, B. (1997) ‘Public Support for
Community Based Sanctions’, The Prison Journal, 77(1): 6–26.
Van Krieken, R. (1989) ‘Violence, Self Discipline and Modernity: Beyond the Civilising
Process’, Sociological Review, 37: 193–218.
Vaughan, B. (2000) ‘The Civilising Process and the Janus Face of Modern Punishment’,
Theoretical Criminology, 4(l): 71–93.
 Crime, community safety and toleration 119

Walzer, M. (1983) Spheres of Justice: A Defence of Pluralism and Equality, New York: Basic
Books.
Wilkins, L. and Pease, K. (1987) ‘Public Demand for Punishment’, International Journal of
Sociology and Social Policy 7(3): 16–29.
Young, I. (1990) ‘The Ideal of Community and the Politics of Difference’, in L.
Nicholson (ed.) Feminism/Postmodernism, New York: Routledge.
Young, J. (1999) The Exclusive Society, London: Sage.
Young, W. and Brown, M. (1993) ‘Cross National Comparisons of Imprisonment’, in M.
Tonry (ed.) Crime and Justice: A Review of Research, volume 17, Chicago: University
of Chicago Press.
Yovel, Y. (1998) ‘Tolerance as Grace and as Rightful Recognition’, Social Research, 65(4):
897–921.
5 ‘Broken Windows’ and the
culture wars
A response to selected critiques
George L. Kelling

Introduction
Since the publication of my article, ‘Broken Windows: The Police and
Neighborhood Safety’, in the journal Atlantic Monthly in 1982, it has caused
considerable consternation in much sociological, legal and criminological litera-
ture.1 Its main ideas, however, are widely accepted by policy-makers in most cities
in the United States. These include most police and criminal justice agencies,
neighbourhood organisations and local political leaders – regardless of their
ethnicity, race or political orientation. Intellectually and politically, ‘Broken
Windows’ grew out of the centre right. Professor James Q. Wilson, a well-known
neo-conservative, co-authored the original ‘Broken Windows’ article. New York’s
Mayor Rudolph Giuliani, also centre right, and Commissioner William Bratton
gave ‘Broken Windows’ unprecedented publicity when it was implemented as
policy and practice in New York City. The Manhattan Institute, a thriving conser-
vative New York urban think-tank, is one of its strongest intellectual strongholds.
This chapter argues that the middle/left liberal criminological establishment has
been held captive for decades by the radical individualism of the far left. The
consequence has been that liberals, traditionally champions of cities, have aban-
doned them. The stridency of the far-left attack on ‘Broken Windows’, indeed its
nastiness, signifies its desperation. It also signifies the intellectual stinginess of the
near left, as it attempts to reposition itself in relation to the issue of disorder.

‘Broken Windows’
‘Broken Windows’ is a metaphor. It argues that just as a broken window left
untended is a sign that nobody cares and leads to more damage, so disorder left
untended also signifies that nobody cares and leads to fear of crime, more
serious crime and urban flight and decay. The original article, and subsequent
writings, have put forward the following points:

• incivilities, disorderly behaviour and conditions, and minor offences can

cumulatively be as significant to citizens as felonies like rape, robbery,
burglary and assault;
 ‘Broken Windows’: a response to critiques 121

• such problems increase the fear of crime;

• citizens act on these fears – they move, take protective measures and in
other ways shape their lives on the basis of their fears;
• citizens want something to be done about disorder;
• disorder and fear are sequentially linked to serious crime and urban decay;
• nevertheless, the effects of disorder on a neighbourhood vary considerably:
some neighbourhoods have the capacity to tolerate and absorb a significant
amount of disorder while others are highly vulnerable to lower levels of
disorder;
• intervening to restore order is a means of interrupting or reversing the down-
ward spiral into crime and urban decay in vulnerable communities, while
managing disorder can prevent this spiral from starting in more resilient
ones; and
• police order maintenance activities are, by their very nature, highly discre-
tionary.

Each of these points is to be found in the original ‘Broken Windows’ article.

To be sure, some are spelled out in more detail than others, but all are there.
Moreover, each is discussed in considerable detail in Fixing Broken Windows:
Restoring Order and Reducing Crime in Our Communities (1996), written with
Catherine Coles after I was involved in implementing police order maintenance
activities in the New York City subway system.
And, lest I be misunderstood, I do not suggest that ‘Broken Windows’ should
go unquestioned. On the contrary, the original formulation about the causal
links between disorder, fear and crime was put forward as a hypothesis and
invited questioning and research. To date, the research on these linkages has
been mixed. On the one hand, Skogan’s (1990) research is highly supportive,
on the other hand, Harcourt (1998), using Skogan’s data, draws different
conclusions. My own belief, based on my experiences and reading of the
research, is that the original formulation will withstand research, although it
will be refined (e.g., only specific categories and subcategories of crime are
linked to disorder). And finally, I am not unaware of the potential abuses that
can result from police order maintenance activities. I am keenly aware of them
and wrote about them, and their history, at length in Fixing Broken Windows. I
am so concerned about doing police order maintenance properly, that immedi-
ately after completing Fixing Broken Windows, I authored Broken Windows and
Police Discretion, which identifies a model for guiding police discretion in the
conduct of order maintenance (Kelling 1999).
For me, the sources of the ideas in ‘Broken Windows’, aside from some litera-
ture, include observing foot patrol officers, interviewing citizens, citizen surveys,
observing neighbourhoods and communities, meeting formally and informally
with citizen groups (especially in inner cities), and problem-solving activities
(disorder in the New York City subway and ‘squeegeeing’ on New York City
streets). These ideas took form within the context of community policing as it
122 George L. Kelling

evolved in the United States during the mid- and late 1980s and, for me at
least, are inseparable from this context.
Professor Wilson and I continue to be pleasantly surprised by the original
article’s ‘legs’. It has been reprinted often, is available on the Internet, has been
reported by Atlantic Monthly as one of its most popular articles, has been trans-
lated into several languages, has engendered countless ‘Broken Windows’ police
programmes, and has been applied to probation, education, health and other
professional endeavours. Most visibly, it has been associated with the historic
reductions in crime in New York City. Both Mayor ‘Rudy’ Giuliani and his first
Police Commissioner, William Bratton, are outspoken advocates of a ‘Broken
Windows’ approach to policing. Fixing Broken Windows is now being translated
into Polish, Portuguese and Spanish.
The original article also spawned immediate and often vitriolic attacks. The
National Institute of Justice (the research arm of the US Department of Justice)
issued a request for a proposal the same year that ‘Broken Windows’ was
published to test the article’s thesis that crime was linked to disorder. After the
research was designed, the US Attorney General quashed it as being too contro-
versial. Professor Samuel Walker published the first academic critique in 1984.
Others followed. Former Kansas City, Missouri and San Jose, California Police
Chief Joseph McNamara has been the most vitriolic practitioner opposing
‘Broken Windows’, primarily on radio and television talk shows and in news-
paper op-ed editorials (McNamara 1997). Among other allegations, ‘Broken
Windows’ supposedly gives rise to racism, ‘wars’ on the poor, police brutality
and criminalises the homeless. At least one author has labelled James Q. Wilson
and myself as racists – ‘aversive’ racists at that (Stewart 1998).2 These critiques
have not exactly been in lightweight journals. They include the Yale and
University of Michigan law journals, both representing prestigious law schools
in the US (Harcourt 1998; Stewart 1998).
On reflection, the academic and professional assaults are not surprising.
Certainly Professor Wilson foresaw them. By the early 1980s, his neo-conser-
vatism had already alienated many criminologists and he predicted that merely
by writing with him I would fall under academic suspicion.3 Professor Wilson’s
reputation notwithstanding, ‘Broken Windows’ is controversial in at least two
overlapping yet distinct respects. First, it directly challenges the reigning ‘root
causes’ of criminal justice ideology – not surprising given the traditional liber-
alism of sociology/criminology. Second, it runs head-on into the ‘culture war’
raging over the competing demands of individual rights and community interests.

Criminal justice
The reigning criminal justice ideology in the US is a direct outgrowth of
President Lyndon Johnson’s 1960s crime commission and its 1967 report, The
Challenge of Crime in a Free Society. A conceptual offshoot of the ‘War on
Poverty’, Challenge emphasises the tragic consequences of racism, poverty and
social injustice and links them to crime, social disorder and violence
 ‘Broken Windows’: a response to critiques 123

(President’s Commission, 1967). These broad social problems are the ‘root
causes’ of crime. Not surprisingly, their amelioration is fundamental to
preventing crime and violence. Police, prosecutors, corrections and criminal
courts comprise a criminal justice system whose purpose is to process offenders –
murderers, rapists, robbers, burglars and others who commit serious offences.
Minor offences and ‘victimless’ crimes should be actually or virtually decrimi-
nalised to end the ‘overreach’ of the criminal law.
The most obvious policy implications of the report are, of course, that basic
structural changes are required in American society to reduce crime: wealth
redistributed, racism ended and social injustices rectified. Beyond this, the
report has two major sets of recommendations. First, the relationship between
criminal justice agencies and minorities must be improved, partially by ensuring
that minorities are represented among police, prosecutors, corrections and
courts. Second, the criminal justice system should be improved through
enhanced planning, technology and modernised administrative processes
(recruitment, training, supervision, etc.).
It is impossible to exaggerate to non-Americans the extent to which these
ideas permeate criminology, criminal justice academic programmes and criminal
justice practice in the US. ‘Root causes’ became conventional wisdom, in
popular as well as professional culture. At its extreme, the premise is stretched
to a syllogism: crime is caused by poverty, racism and social injustice; police can
do nothing about these problems; ergo, police can do little, if anything, about
crime. The business of justice agencies is processing cases. To the extent that,
say, police can have an impact on crime, it is by being the front end of the
‘system’: arresting serious offenders and forwarding them for processing.
One can be concerned about the problems of poverty, racism and social
injustice and still be unhappy with this formulation. Indeed, one can find much
in The Challenge of Crime in a Free Society with which to agree: its calls for
lawfulness in the justice process; the urgency with which it calls for improving
the relationship between criminal justice agencies and minority communities;
and its mandate that minorities must be recruited, hired and promoted in crim-
inal justice. In each of these areas, the document was genuinely and
appropriately revolutionary (President’s Commission, 1967).
As a primer of crime control, however, the President’s report was a disaster.
The ‘root causes’ assumption has never stood up to rigorous analysis. Crime
varies too much by time, demography, economy, region and country to be
attributed to such structural characteristics. Moreover, the ‘system’ metaphor is
so entrenched and self-apparently ‘true’ that it thwarts careful scrutiny about
how agencies actually function. The reality that criminal justice agencies,
almost literally, never talk to each other and often work at cross-purposes is
obfuscated by the power of the metaphor. The idea that such agencies could
plan together was, at best, ahead of its time and remains optimistic today –
although there now are some wonderful examples of what can result when they
do (Kennedy 1997). Also, the report looked solely to government to solve the
crime problem by declaring a ‘war on crime’ – akin and complementary to the
124 George L. Kelling

United States’ other domestic war, the ‘war on poverty’. The private sector,
funding and sponsoring by far the largest crime control capacity in the United
States, is completely disregarded in the report. Likewise, citizens and commu-
nity organisations are relegated to consumers of professional crime control
agents: they are to be good observers, reporters and witnesses – and to increase
their financial support of criminal justice agencies. Finally, most of the ‘system
improvement’ agenda – improved recruiting, training, supervision, co-ordina-
tion and technology – has been on the table since the 1920s in the US.
Regardless, ‘root causes’ ideology largely held crime prevention hostage until
well into the 1990s in the US.

The culture wars

Professor Francis Fukuyama has labelled the social revolution that characterised
the US during the 1960s and 1970s as ‘The Great Disruption’ (Fukuyama
1999). Between 1960 and 1990, profound changes in values, beliefs and
philosophies swirled through, not just the US, but through every western
country with enormous social and cultural impact. Fundamental to these
changes was a shift in the balance between individual rights and community
interests. James Q. Wilson summarised this issue:

The competing demands of liberty and community constitute a funda-

mental cleavage that divides contemporary political philosophers and has
produced among the public at large the American culture war. The
defenders of liberty envisage a world of autonomous individuals who freely
choose their destinies and whose liberties are essential to personal develop-
ment and social democracy. The advocates of community rejoin that no
one is truly autonomous, that liberty can only exist in an environment of
reasonable order, and that personal development requires familial and
neighborhood support.
(Wilson 1999: xii)

This shift towards unfettered individualism affected every facet of personal and
public life: the family, church, education, sex roles, medicine, housing, mental
illness, law, commerce, drug policy, and so on. But nowhere are these competing
demands – between individual rights and community interests – more
discernible than in matters of criminal justice policy. Regarding ‘serious’ crime –
murder, rape, assault, robbery and burglary – criminologists/sociologists have
quarrels, however, a broad consensus exists that these crimes need to be taken
seriously. While issues like sentencing, the death penalty, the transfer of juve-
niles to adult court, and others are debated heatedly, at least agreement exists
that felonies should be taken seriously.
For minor offences – begging, prostitution, public drinking, graffiti, and so on
– the cleavage is yawning. The far left – including a good share of sociologists,
criminologists and civil rights lawyers and advocates – not only does not want
 ‘Broken Windows’: a response to critiques 125

anything done about such offences, it views perpetrators of minor offences as

victims of a corrupt/unjust society who are ‘enriching’ society with their
messages. Thus, begging is elevated to the status of a political message about the
inequitable distribution of wealth; graffiti is the ‘folk art’ of disenfranchised youth
who have no other means to express their beliefs; and ‘squeegeeing’ (the unso-
licited washing of car windows) is the ‘work’ of unemployed and homeless youth.
And, of course, schizophrenics are not ill: they are heralds of society’s pathology.
Liberal enlightenment in this world means that, short of violence – and even this
may be questionable – there are no outrages that are not ‘understandable’ and
deserving of toleration given society’s inequities and pathology. ‘Tolerating the
intolerable’ in America’s great cities is the acid test of one’s true commitment to
civil rights and social justice (Myerson 1998; Podhoretz 1999). Every ‘in your
face’ indignity is someone’s constitutional ‘right’ and must be endured.

Some background about ‘Broken Windows’

The ideas in ‘Broken Windows’ originated during an era when policing was
undergoing serious examination in the US – the depths of which may not be
easy for non-Americans to comprehend. By the late 1970s, the impact of
preventive patrol and rapid response to calls for service, the then-core capaci-
ties of police, seemed negligible (Kelling et al., 1974; Kansas City Police
Department 1977). The Supreme Court had moved to bring criminal investiga-
tion under control earlier and a 1975 study of detectives by the Rand
Corporation suggested that secretaries could have done most of their work
(Chaiken 1975). Nothing seemed to work, disorderly conditions were rife and
crime was skyrocketing.
As befuddled by these circumstances during the mid-1970s as most others, I
started to look outside of policing’s then-conventional wisdom about how crime
could be controlled. In policing itself, I wondered about the potential of foot
patrol and looked for a location to study it. While police held foot patrol in
disdain, citizens and politicians pushed hard for its implementation (The
Newark Foot Patrol Experiment 1981). Outside of policing, I became aware of
the inchoate movement among citizens to organise themselves into neighbour-
hood crime control groups and I wondered about their potentials. Based on my
exploratory work on foot patrol in Boston, Massachusetts and Newark, New
Jersey I also started to speculate about the possibility of new measures of police
performance – specifically, what I called ‘quality of life’ measures (Kelling
1978). In a 1978 article inspired by a rereading of Jane Jacobs’ The Death and
Life of Great American Cities and a rereading of Egon Bittner’s classic work
(1967) on order maintenance, I wrote for the first time about minor offences
and disorderly behaviour and conditions:

[O]rder maintenance in the normal work-a-day world … is what people

miss on the streets of United States cities (and this is the central theme of
this paper): a police presence which operates to insure that the norms,
126 George L. Kelling

rules, and regulations of civil contact between people are facilitated and,
when these social control functions break down, to assist restoration of
street equilibrium.
(Kelling 1978: 111)4

Most recently, in Fixing Broken Windows(1996), Catherine Coles and I traced

the evolution of ‘Broken Windows’ forward from the findings of foot patrol
research, through the Atlantic Monthly article, and finally to its implementation,
first, in the New York City subway system and, later, in New York City itself
under Mayor Rudolph Giuliani and Commissioner William Bratton (1996).
There was an important ‘wrinkle’, however, in the evolution and implemen-
tation of ‘Broken Windows’. It first appeared during the early 1990s attempts to
reduce disorder in the New York City subway and later became an integral part
of police order maintenance activities there. At the time of writing, it is at the
core of the political struggles in New York City. Liberals, incensed at Giuliani’s
reputation for reducing crime avidly attempted to neutralise it as an issue when
it appeared that he would be Hillary Clinton’s opponent in New York State’s
2000 senatorial race (not that they will be any less incensed now that he has
decided not to run).5 It has become clearer since the publication of Fixing
Broken Windows (1996) that this wrinkle deserves some systematic discussion.
On reflection, it was a credit to Bratton that when he was chief of the New
York City Transit Police Department he both closely monitored his tactical
implementations and, when progress was noted, immediately provided feedback
throughout the department.6 For him, systematic feedback was a means of
ensuring accountability, adjusting tactics to changing conditions, providing a
vision of the ‘business’ of the transit police and communicating with line
personnel about their achievements. Early in the Department’s attempts to
reduce fare beating in the subway, crime analysts noted that in some stations
many who were arrested for fare beating either were carrying illegal weapons,
mostly guns, or had warrants out for serious felonies.7 The significance of this
was not lost on either Bratton or his emerging tactical strategist Jack Maple.
Not all fare beaters were serious criminals or gun carriers, but a significant
percentage was.
Until this time, the primary means of confiscating weapons on US streets
had been through traffic stops and vehicle searches – some of dubious legiti-
macy. Such practices go back to the 1950s when police in cities like Chicago
routinely stopped cars for traffic offences and searched them for guns. At times
these searches were legitimate: officers had good reason to suspect that guns
were in the cars. At other times, the searches were a tacit quid pro quo: ‘you let
me search your car and I will not give you a ticket’. At yet other times, they
were illegal: police simply stopped and searched cars with youths, often
minority youths not committing any traffic offence. Such stops were a pretext
to give officers the opportunity to search cars. If guns were found, it was not
unusual for the person to be arrested and jailed, the gun confiscated and the
matter dropped. Often police would not bother to appear at the hearing the
 ‘Broken Windows’: a response to critiques 127

next morning and the matter would be dismissed. From the point of view of
most police the matter had been handled: a gun was confiscated and the person
spent the night in jail (Kelling 1999).
Bratton, however, saw the matter differently in the subway and, later, city-
wide:

An unanticipated by-product of the sweeps came when we checked the

identification and warrant status of everybody we were arresting. During
the early stages of the initiative, we found that one out of every seven
people arrested for fare evasion was wanted on an outstanding warrant for a
previous crime. One out of every twenty-one was carrying some type of
weapon, whether a box cutter, a knife, or a gun. As so often happens in
policing, we had focused on one problem to the exclusion of others. Now
we were beginning to understand the linkage between disorder and more
serious crimes.
(Bratton 1998: 154)

Discovering that such a high percentage of fare beaters was carrying weapons
was one of those ‘eureka moments’. And, in Jack Maple’s fertile street-smart
mind, arresting someone with a weapon provided additional possibilities. Did
the person have more guns? Where did he get the gun? Who else had guns and
was carrying them? As Bratton put it when referring to the impact this had on
the subway: ‘After a while, the bad guys wised up and began to leave their
weapons at home’ (Bratton 1998: 154).
The true potential of this finding, however, was only to become apparent
after Bratton became Commissioner in New York City. With over 2,000
murders a year, most of them committed with handguns, New York City was in
the midst of a bloodbath during the late 1980s and early 1990s. And New York
City wasn’t alone: name the American city and violence was spinning out of
control. Something had to be done about gun carrying, but just what basically
stymied police, policy analysts and researchers. Research had been largely
fruitless. Sherman, Shaw and Rogan, for example, published the ‘Kansas City
Gun Experiment’ in 1995. Funded by the National Institute of Justice, this
study did little more than refine and endorse conventional police practice: use
of traffic stops to search cars when officers felt their safety was threatened;
asking citizens to submit to voluntary searches of their cars; looking into
stopped vehicles for weapons; and other forms of consent searches including,
for example, glove compartments (Sherman et al., 1995). In the meantime,
criminological pundits were loudly proclaiming that we had not seen the worst
violence yet: a second ‘baby boom’ cohort would soon turn loose a new gener-
ation of predators.
But in this context, Bratton and Maple saw investigative opportunities in
police order maintenance activities – specifically about the issue of gun
carrying. Despite the epidemic of gun violence, Jack Maple notes the NYPD’s
lackadaisical attitude about confiscating guns:
128 George L. Kelling

Taking guns off the street didn’t seem to interest the [detective] squads at
all. During the previous year patrol made almost 9,000 arrests for crimes in
which a firearm was confiscated, but the entire Detective Bureau developed
only four confidential informants to help identify gun sellers and took, as a
result, fewer than fifty additional guns off the street.
(Maple 1999: 26)

But Maple was not just concerned about special or detective units. ‘All …
arrests made must be viewed as opportunities to do more debriefings’ (Maple
1999: 26). Based on the subway experience this principle, of course, included
arrests for minor offences. Someone arrested for public urination? Question him.
Does he have a weapon? Where did he get it? Does he know someone who
carries a weapon? Where did that person get the weapon? For Maple, quality of
life arrests were a potential treasure trove, not just for guns, but for other forms
of contraband as well.
And this, of course, brings us to a critical issue, both about the whole idea
of ‘Broken Windows’ and the current political and professional controversies in
New York City. An important serendipitous finding – that many fare beaters
were felons carrying weapons – added a new dimension to the ‘Broken
Windows’ idea. Probably I should not have been surprised. Street cops prob-
ably could have told me about this all along and they might have, however, it
just never registered. Nonetheless, police leaders expanded order maintenance.
Heretofore, police saw it as a means to reduce citizen fear of crime and prevent
crime by restoring order, empowering citizens and maintaining neighbour-
hood/community standards. Now, police understood it to be an investigatory
tool as well, not the least for getting weapons off the street.
In the context of the national furore over gun-related violence and 2,200
murders a year in New York City – most of them with handguns – this was a
big-time finding. It was highly dramatised in 1996 after a man brutally
assaulted three women near Central Park. He was identified when fingerprints
at the scene of the assault matched those taken earlier when he was arrested
for fare beating. Maintaining order gained new meaning for line cops: it was
not just a matter of dealing with obstreperous citizens and unpleasant condi-
tions; it was an opportunity to apprehend some serious offenders. Like traffic
enforcement earlier, where stops had potential beyond maintaining safe and
orderly automobile and pedestrian movement, order maintenance put police
into contact with serious trouble-makers and gave them opportunities to check
for weapons, drugs and other contraband. But like traffic stops as well, it
opened new possibilities of excesses in the name of contraband reduction,
especially guns.
In the New York City context of the early 1990s, it was a small step from
order maintenance activities as part of routine police activities to ‘stop and
frisk’ by special anti-crime units attempting to reduce gun carrying. Guns were a
major problem; disorderly behaviour, like traffic offences before them, provided
the legal basis to stop persons on the street. The focus, however, shifted from a
 ‘Broken Windows’: a response to critiques 129

wide range of order maintenance activities to a concentration on arrest and its

threat to allow officers to search and interrogate citizens, especially youths.
Also complicating the picture during the late 1990s was the emergence of
the phrase ‘zero tolerance’ and its use as a substitute term for ‘Broken Windows’.
I will not go into length about the phrase here. I never used it, and view the
phrase as a political sound-bite, not a viable set of policies. On the surface, the
phrase is antithetical to the highly discretionary activities that ‘Broken
Windows’ implies and which Coles and I made explicit in Fixing Broken
Windows. Unfortunately, zero tolerance as a phrase is so entrenched in some
countries that it is futile to attempt to do other than make people aware of its
problems. I have often suggested that both the far left and far right love the
term: for the far left it demonstrates that ‘Broken Windows’ is fascist; for the far
right it justifies ‘head knocking’ by police.
A final point to be made about the evolution of ‘Broken Windows’ during
the 1990s: I have little quarrel with, nor am bothered by the investigatory use of
‘Broken Windows’ as part of routine policing functioning. When people break
the law, they open themselves up to such questioning. It is the use of order
maintenance by special units for investigatory purposes that I find worrisome.
While one can develop justifications for their limited use, they tend to be
elitist, have little touch with neighbourhoods and communities, develop a life
and agenda of their own, demoralise beat officers, and tend to get into trouble.
‘Sweeps’ by special units was always inconsistent with ‘Broken Windows’ as I
have conceived of it. I have been explicit about this, especially in Fixing Broken
Windows. At best, in my judgement, special units should have limited life, be as
decentralised as possible, and should be brought in to support neighbourhood
patrol officers. But this is a complicated issue in need of much study.8

The response to ‘Broken Windows’

As I write this chapter, demonstrators are protesting about a series of police
shootings of civilians: most notably, the Diallo incident in New York City. In
this case an unarmed African immigrant was shot and killed in a high crime
area by four police officers who fired their weapons forty-one times and hit
Diallo nineteen times. A mixed race jury found the officers innocent of any
criminal act.9 At a demonstration at Rutgers, one of my faculty colleagues – not
a criminal justice colleague – waved Fixing Broken Windows in front of the audi-
ence claiming that it is New York Mayor Giuliani’s blueprint for policing in
NYC and, as such, responsible for the killing of Diallo. Likewise, at Harvard, a
leaflet announcing a protest there began with ‘Protest Harvard University
Research Fellow George Kelling: architect of racist “zero tolerance” police poli-
cies’ – again, laying Diallo’s death at my feet.
I mention this to make a point. My view of these demonstrators is consider-
ably different than my view of some of my ‘academic’ critics. The
demonstrators’ activities are patently political. I wish that speakers, especially
academics, would accurately reflect my points of view, however, I do not kid
130 George L. Kelling

myself. The far left’s animosity towards Giuliani is so rabid, and their anti-police
feelings so passionate, that one would hardly hope for a careful reading of the
text. Moreover, I had no right to be surprised. I had entered the political arena
by authoring an op-ed piece in the Wall Street Journal about the incident.
Nor do I expect much quarter from civil rights and advocacy lawyers in the
central arena in which ‘Broken Windows’ and order maintenance issues are
being fought: the courts. It is in the nature of the advocacy process, really an
extension of political processes, that briefs and writs will contain misrepresenta-
tions like ‘criminalising the poor’ and ‘police harassment of minorities’ in
litigation when discussing order maintenance and ‘Broken Windows’.
I had expected something else, however, from academic critics writing in,
say, prestigious academic journals like The American Sociological Review. Before I
say what I expected in academic circles, let me indicate what I had not
expected. I had not expected that people would agree with me or would not see
potentially negative consequences of ‘Broken Windows’ policies. My hope,
instead, was for ‘second order agreement’. By second order agreement, I mean
that people who disagree put forward opposing points of view accurately, if not
convincingly.10 Most academic critiques of ‘Broken Windows’, even the best,
have not been characterised by such intellectual forthrightness nor, I would
add, honesty.
I have responded to many of the more outrageous misrepresentations (‘aver-
sive racist’, etc.; see Kelling 1998). Moreover, I have ignored many criminological
ditherings (Foucault and techniques of punishment, and whether or not ‘Broken
Windows’ constitutes a legitimate ‘theory’). My interest here is in what is
happening in liberal/left criminology and why ‘Broken Windows’ sticks in their
craw as much as it does.
The fact is that liberals lost cities – politically and intellectually. In the post-
1960s era, liberals, including most American criminologists, simply walked away
from disorder, fear of crime and serious crime and their impact on neighbour-
hoods and communities. Early twentieth-century Progressives/liberals were the
great defenders of public spaces and neighbourhood interests. They mapped an
agenda of what was wrong with cities and took concrete steps to remedy the
wrongs. Now, political conservatives reign in most American cities.11
Hunter University Professor Fred Siegel has written about the transformation
of liberals:

Beginning roughly a quarter-century ago, the work of the Progressives and

their New Deal Successors was undone. In a great but now all too obviously
failed experiment, the Progressive ideals of integration and uplift that had
served the city so well were largely abandoned as repressive mechanisms of
middle-class social control. In the great wave of moral deregulation that
began in the mid-1960s, the poor and the insane were freed from the fetters
of middle-class mores. The ‘revolution’ achieved some important gains: it
freed people unjustly imprisoned in mental hospitals, made the police a bit
less highhanded, and gave welfare recipients the dignity afforded by due
 ‘Broken Windows’: a response to critiques 131

process rights. But like every revolution, it had its victims. In the new
climate in which drug use was redefined as a ‘victimless crime’ and purity of
legal procedure took precedence over the safety of the community, people
who had once been shackled by the inequities of the law found themselves
trapped by a pervasive lawlessness. Delivered into the tender mercies of the
streets, the most vulnerable among us have suffered immeasurably.
(Siegel 1992: 37)

Nowhere did liberals lose touch as badly as in crime control policy. The
liberal crime control agenda, stuck as it was in ‘root causes’, was completely
overwhelmed by conservatives, especially in sentencing policy. ‘Truth in
sentencing’ and ‘three strikes and you’re out’ carried the day in policy innova-
tions. The major political crime control battles raged over gun control (liberals
for, conservatives against), capital punishment (conservatives for, liberals
against) and drug legalisation (a funny mix on each side) – all largely symbolic
issues that said little or nothing about what could be done today to prevent
crime. In the meantime, neighbourhood residents could tell you that the kids
who were stealing hubcaps from cars would soon be stealing the cars if some-
thing wasn’t done pretty soon. They could tell you as well that if you made the
cars theft-proof, some of the youths would find a way to steal them anyway –
and they did. They simply hijacked them at gunpoint, increasing the danger.
Police and criminal justice agencies in Newark, New Jersey, for example,
became such a joke that kids would steal cars and, to provoke police who were
forbidden to chase them, bump into police cars to ‘air-bag’ them in the hope
that police would get angry enough to get into ‘hot pursuits’. It was a game:
even if they were arrested, nothing would happen. In a world without mean-
ingful sanctions for minor offences, some youths – not most, but a lot – spun
out of control and by the time they were entering their late teens were starting
to accumulate felony records. And then it was too late. ‘Three strikes you’re
out’ and ‘truth in sentencing’ ended the joke, but only after innumerable minor
offences and many other serious crimes. Now, liberals like to castigate conser-
vatives and ‘authoritarian elements’ for the tragically high imprisonment rate
in the US, but liberals should be looking to themselves as well to attribute
blame.
Regarding police, the liberal agenda paid little attention to their crime
control potentials – after all, since crime was linked to structural features of our
society what could police do except process offenders? Community policing, at
least in its early days in the 1980s, was widely viewed as ‘soft’ policing –
providing services and seeking community approval – to such an extent that it
was widely rejected by most practising police officers. Generally, criminological
liberals lauded community policing because it was ‘nice’ policing characterised
by great concern for establishing good relationships to communities and neigh-
bourhoods. Community policing might enhance criminal investigation as a
consequence of its access to information; however, the inherent aggressiveness
of community policing was largely lost on most criminologists. Traditional
132 George L. Kelling

policing was inherently passive and non-intrusive: patrol, wait for a crime to
happen, and then respond. The response might appear forceful – fast car, lights,
siren – however, its premises and execution were reactive. Community policing
was far more intrusive and aggressive than traditional policing. More crimes
came under the purview of police (minor offences as well as serious), their
source of authority was broadened to include civil as well as criminal authority,
and police got out of their cars to ‘re-police’ public spaces, streets and neigh-
bourhoods. Nonetheless, until very recently, the agenda of most criminologists
for police was limited to civilian review, corruption control and reduction of
abuse – all focusing on what police should not do rather than what they should
do. Even problem solving, an enormously powerful, and inherently aggressive,
tool first articulated by Herman Goldstein, took on an ideological edge as well.
For example, in New York City the resolution of ‘squeegeeing’ – extortion of
money from automobile drivers under the guise of washing car windows at busy
intersections – was not a problem-solving exercise according to some of its
advocates, because it did not deal with the ‘root causes’ of squeegeeing.12 And,
of course, all of this came to a head when Giuliani and Bratton mobilised the
NYPD around very specific crime control goals and succeeded beyond anyone’s
dreams. They not only reduced crime, they ‘called their shots’ and explicitly
challenged the premises of root cause criminology. Liberals fumed: this is not
community policing; police are ‘cooking the books’; the crime reductions are
the result of structural changes – demography, the economy, changing drug use
patterns; and finally, well, the NYPD might be reducing crime, but at the unac-
ceptable cost of high-handed ‘stop and frisk’ practices and the abuse of citizens,
especially minorities. Never mind that in New York City murders are down
1,600 a year and robberies 60,000; that civilian deaths from police shootings are
down from 41 in 1990 to 11 in 1999; or, that allegations of abuse of force are
down by over 1,000, during a time when 8,000 officers were added to the
NYPD.
In fact, liberals ceded crime control as an issue to conservatives. The radical
left notion that order maintenance is really racism, cultural imperialism, ‘crimi-
nalisation of the poor’, ignores the plain reality that liberals refuse to
acknowledge: poor and minority neighbourhoods are being ravaged by disor-
derly persons and conditions and the refusal of officials to take neighbourhood
priorities seriously.
It is in this context that I will briefly address two critiques of ‘Broken
Windows’. One, ‘Replacing “Broken Windows”: Crime, Incivilities and Urban
change’ by Roger Matthews (1992), is written from a British point of view.13 It
is well known in both Britain and the US and has shaped many commentaries
on ‘Broken Windows’.14 The second is ‘Systematic Social Observation of Public
Spaces: A New Look at Disorder in Urban Neighborhoods,’ by Robert J.
Sampson and Stephen W. Raudenbush. The most recent critique, it was
published in the prestigious American Sociological Review. Both are important
articles by established social scientists. I have no doubt that the latter article
will become as influential as the first. It already has been put forward in the
 ‘Broken Windows’: a response to critiques 133

National Institute of Justice Journal as a response to ‘Broken Windows’ (National

Institute of Justice Journal, April 2000).
For Matthews, ‘Broken Windows’ is the focal point of the debate between
new realists and radical realists with ‘Broken Windows’ an example of new
realism. Within this context, Matthews raises three issues: ‘The first involves
the degree to which crime and incivilities are inextricably linked. The second is
whether ‘Broken Windows’ – either metaphorically or literally – have the same
effect in different areas; and the third involves the accuracy of the develop-
mental sequence which is suggested’ (Matthews 1992: 22). Matthews properly
suggests that these issues are more than theoretical; they have ‘a direct bearing
on the formulation of policy’ (ibid.).
The discussion that Matthews presents during roughly the first half of the
article is a welcome one, sober and straightforward, based upon an examination
of both US and British research conducted since ‘Broken Windows’ first
appeared in 1982. Regarding Matthews’ first and third issues, Wilson and I were
well aware that we were hypothesising about the relationship among disorder,
fear, serious crime and urban decay. We believed strong sequential linkages
existed; however, we presented the article as exploratory in nature. The second
issue that Matthews raises – whether ‘Broken Windows’ have the same impact
in different neighbourhoods – is a non-issue. Wilson and I argued explicitly that
disorder is less likely ‘in places where people are confident they can regulate
public behaviour by informal controls’ (Wilson and Kelling 1982: 33).
Matthews’ discussion goes distinctly awry, however, once he gets into the
area of the ‘problems of policing disorder’ (1992: 34). He writes: ‘[Wilson and
Kelling] unequivocally endorse the use of extra-legal methods and the need to
‘kick-ass’ to keep people in order. Significantly, the asses of the people they
want kicked are winos, street prostitutes, panhandlers, and juveniles’ (1992:
35). First, extra-legal is not synonymous with illegal.15 Throughout all my writ-
ings, lectures and consultations, I have encouraged police only to invoke the
law (cite or arrest) as a last resort, and to use extra-legal means – counselling,
education, cajoling, persuasion, warning, etc. – as much as possible to solve
problems. But, more to the real point, the article simply does not include an
unequivocal endorsement to kick the asses, as Matthews writes, of winos, street
prostitutes and others. To be sure, Wilson and I describe an officer who himself
says, when describing the problems of dealing with gangs in public housing,
‘We kick ass’, yet any fair reading of this section of the paper would understand
that we are raising the same complex issues about which Matthews and others
are rightly worried. These include the endorsement by a troubled community
of illegal police means to help solve its problems, the lack of consensus in
many communities, the role of due process when responding to neighbourhood
problems, the possible over-reliance on police to keep order and other such
complex issues. ‘Unequivocally endorse’ is nothing less than a misrepresenta-
tion of our argument – and a gross one at that. To the contrary, we ‘hedge’
throughout the article, struggling with the complexity and subtleties of the
issues we are raising. For example, in response to police deployment issues, we
134 George L. Kelling

first call for experimentation. Then we add: ‘The second answer is also a hedge
– many aspects of order-maintenance in neighborhoods can probably best be
handled in ways that involve the police minimally, if at all’ (Wilson and
Kelling 1982: 36).
Notwithstanding the substance of ‘Broken Windows’, or its cautious
exploratory tone, Matthews’ critique degrades from an interesting discussion of
valid substantive and policy issues about which there is ample opportunity for
honest disagreement, into a diatribe that has little to do with ‘Broken
Windows’ theses, tone, or spirit. Examples abound. Matthews writes: ‘When
the public report these problems [rowdy youths, etc.] to the police, it is often
not because they want a heavy-handed, truncheon-wielding army of police offi-
cers descending on their neighborhood’ (1992: 37) Who suggested a
heavy-handed, truncheon-wielding army? This is absolutely out of the spirit of
what was written. Matthews goes on: ‘Wilson’s and Kelling’s contention is that
the police are the key to controlling disorder’ (ibid.: 38) In fact, as the quote
from ‘Broken Windows’ above demonstrates, we say the opposite. Nonetheless,
continuing in this line, Matthews avers:

Wilson and Kelling express little interest in developing mechanisms and

agencies which may empower the poor and the powerless, and they seem
even less interested in directing resources towards the disadvantaged and
marginalised. Instead, their main concern is to remove these undesirables
from respectable areas, and since they do not want to do very much about
their condition, they will, presumably, be deflected towards those poorer
areas which already have more than their fair share of social problems. In
their diatribe against decriminalisation and decarceration, they seem to be
suggesting that ‘if only we got really tough’ on these people and put them in
prison for longer, then urban decline in respectable areas could be
prevented.
(Matthews 1992: 41)

A diatribe against decriminalisation and decarceration? Put them in prison

longer? While some might extend the logic of ‘Broken Windows’ in this
fashion, it is by no means an inevitable logical extension. In fact, much in
‘Broken Windows’ suggests that we would find such an extension inappropriate.
Regardless, Matthews carries his logic further:

[T]hey advocate more ‘get tough’ policies, greater use of imprisonment, and
the extension of selective incapacitation. At the same time, they want, not
only more legal sanctions, but also endorse the use of extra-legal tactics by
the police and seem to suggest that the appropriate response to intimida-
tion and harassment on the street is more intimidation and harassment by
the police.
(Matthews 1992: 43)
 ‘Broken Windows’: a response to critiques 135

Imprisonment? Selective incapacitation? I can only assume that we are not

talking about ‘Broken Windows’, but about James Q. Wilson’s other writings:
what started as a discussion of the ‘Broken Windows’ idea has become a critique
of Wilson’s position on other criminal justice policy issues. The issue here, of
course, is how independently discrete an article should be viewed within the
context of a scholar’s total work. Matthews could argue that it is fair to critique
‘Broken Windows’ within the context of Wilson’s (or my) total position.
Clearly, when I write that ‘Broken Windows’ should be viewed within the
context of my writings about community policing generally, I am inviting anal-
ysis within this larger context. Yet, scholarship requires acknowledgement of
this fact and proper citation – both of which are lacking in Matthews’ paper.
Matthews’ justification of such a broad view argues that ‘Broken Windows’
comes complete with all the conceptual baggage he delineates above (1992:
43). This is mistaken; Wilson and I bring conceptual baggage. ‘Broken Windows’
says what it says and deserves to be treated as such. Beyond this, its premises
and implications are certainly fair game, in so far as they are accurately repre-
sented. It is not too much to ask that when scholars move from critiquing the
‘Broken Windows’ ideas to critiquing either Wilson’s or my other work they are
explicit about it.
Taking another cut at the same issue, it can be argued that one of the prob-
lems of ‘Broken Windows’ is that extremists of one ilk or another rely on it to
justify harsh police or other governmental action against ‘problem’ populations.
Matthews is rightly concerned about this possibility. Moreover, Matthews can
properly claim that he and others were attempting to fashion a centre social
democratic position during the late 1980s when his article probably was being
conceptualised. Clearly one can go from the basic arguments of ‘Broken
Windows’ – disorder matters, people want something done about it, it is linked
to serious crime and urban decay – to oppressive policies and regimes. But, this
is not inevitable, either logically or in reality: Wilson and I did not do so in the
article (in fact we worried about it in the article). The very policies that
Matthews finds so promising for dealing with troubled communities – multi-
agency approaches, the development of ‘intermediary agencies’ (park keepers),
improved building design, improved housing policies, and others – are entirely
congruent with, and can be deduced from, the original ‘Broken Windows’ piece
and were discussed in Fixing Broken Windows (1996) at considerable length. In
fact, his ‘centre’ left and my ‘center’ right have more in common with each
other than with each of their extremes. Failing to acknowledge this overlap
restrains both policy and scholarly development.
The second article, ‘Systematic Social Observations of Public Spaces: A New
Look at Disorder in Public Spaces,’ was a lead article in the November 1999
American Journal of Sociology – the place for any sociologist, criminological or
otherwise to be published, at least in the US. Aside from AJS reviewers, crimi-
nological heavy-hitters like Albert J. Reiss, Jr., John Laub and others are
mentioned as sources of inspiration and as reviewers. In contrast to Matthews’
article, which is a state-of-the-art ‘think piece’, Sampson and Raudenbush
136 George L. Kelling

report on research that is well-funded by the John D. and Catherine T.

MacArthur Foundation and the National Institute of Justice.
Sampson and Raudenbush’s (1999) basic argument goes something like this:
first, disorder is conceptualised as an integral aspect of crime; second, the same
factors that produce disorder produce more serious crime; third, ‘both crime and
disorder are rooted in neighbourhood structural characteristics such as concen-
trated disadvantage and neighbourhood social processes such as lowered
collective efficacy’ (Sampson and Raudenbush 1999: 2).
The publication of the article was announced by a press release that opens
with: ‘Does disorder lead to crime? The “Broken Windows” thesis holds that’
(Sampson and Raudenbush 1999: 1). Thus it is clear that for good or ill the
authors are ‘spinning off’ ‘Broken Windows’. Although I had heard rumours
that such an article was being published, I first became aware of it when a
reporter faxed me a copy of the press release on 14 December. Sure enough, the
reporter got the message and queried me about ‘Broken Windows’ being
disproved. Nonetheless, it is an important article.
‘Collective efficacy’ is the authors’ ‘big idea’. Collective efficacy is defined as
‘the linkage of cohesion and mutual trust with shared expectations for inter-
vening in support of neighbourhood social control’ (Sampson and Raudenbush
1999: 612). It is an apt and useful phrase that captures a point that advocates of
community policing and, later, community justice have been making since the
1980s: for disorder, fear and crime to be managed in neighbourhoods, the
inherent strengths of neighbourhoods must be active and/or mobilised. It prob-
ably is a good thing that criminologists/sociologists are recognising it too. (By
the way, it is a point that Wilson and I made in 1982 in the original ‘Broken
Windows’ article and Coles and I discussed in great detail in Fixing Broken
Windows in 1996, although this is not acknowledged by Sampson and
Raudenbush.)
A variety of issues could be raised about ‘A New Look’. I will raise three. The
first is methodological. The second is interpretative: how the authors present
some of their findings. The final is political: how Sampson and Raudenbush
represent ‘Broken Windows’.
The methodological issue concerns sampling – the underpinning of good
research and analysis (this does not mean that there are not other methodolog-
ical problems – there are). The observations that make up the main database of
the project, and are put forward as methodological advances, are measures of
disorder videotaped between 7 a.m. and 7 p.m. The limitations of such a time
sample are casually dismissed in a footnote: ‘Although it would be desirable to
assess disorder during the night-time hours, a pretest confirmed that this was
not feasible with current videotaping technology (or with the naked eye).’
Drunks leaving a bar and creating neighbourhood problems at closing time
cannot be seen with contemporary videotaping technology or the naked eye?
Drug dealers and prostitutes at 6 a.m. on a street corner near an expressway exit
soliciting the ‘on the way to work’ trade cannot be observed? Drunken youths
blaring a ‘boom box’ in a neighbourhood park at 1 a.m. are not observable? This
 ‘Broken Windows’: a response to critiques 137

sampling methodology appears to be determined by the ‘law of the instrument’

rather than the nature of the phenomenon to be studied.
Consequently, it is not surprising that the indicators of physical disorder
wildly diverge from disorderly behaviour (pages 617 and 618). Signs of physical
disorder are in the many thousands while disorderly behaviours, aside from
loitering, are in the 10s, 20s and 30s. But the deck is stacked: physical disorder
is a continuous event. Graffiti, save cleaning or painting it over, is always avail-
able for observation. Likewise, garbage and beer bottles, unless picked up, are
there. They exist in space. Disorderly behaviour, however, is highly ephemeral
and time-sensitive; it exists in time and space – occurring more or less often
depending on a wide range of situational variables. In the random method
chosen to look for disorder in ‘A New Look’, disorderly conditions will always
be observable, while disorderly behaviour may or may not be. The authors
claim to resolve this issue by using statistical techniques, however, looking at
the right times might have been more appropriate. Looking for disorderly
behaviour between 7 a.m. and 7 p.m. is like looking for lost keys under a lamp-
post – not because that’s where they were lost, but because that’s where the
light is good.
It is interesting to contrast Sampson and Raudenbush’s approach with the
sophisticated sampling and measurement methodology of Anthony Braga and
his colleagues (Braga 1999). While they also used videotape technologies to
measure disorderly conditions, they understood both that disorderly behaviour
was time sensitive and that measurement techniques could affect the
behaviour of disorderly persons, especially those breaking the law.
Consequently, rather than being driven by their technology, they used a
different method to measure disorderly behaviour. The idea that a SUV (sports
utility vehicle) cruising slowly down Chicago streets with two video cameras
pointed out the side windows, the core of Sampson and Raudenbush’s SSO
(social science observation), would go unnoticed by street people like prosti-
tutes and drug dealers and not influence their behaviour defies experience and
logic. To give just one example, drug dealers hire kids to watch for unusual
happenings in neighbourhoods (e.g., police attempting to record their drug
dealing behaviour).
Second, a variety of interpretative issues could be raised, but let me address
just one: the handling of robbery. The authors rely heavily on homicide data
‘arguably one of the best measures of violence’. Arguably indeed. Murder is a
reliable measure because it certainly is the most accurately reported to police
and recorded by police. But this does not mean that it is necessarily a more valid
measure, especially given its relative rarity. Nonetheless, Sampson and
Raudenbush conclude in their press release: ‘The results therefore support the
inference that public disorder and most predatory crimes share similar features
and a consequently explained by the same constructs at the neighbourhood
level, especially the concentration of disadvantage and lowered collective effi-
cacy’ (page 4). Ignored in the press release is the following from the article
itself:
138 George L. Kelling

The exception to the emerging conclusion that disorder is spuriously

related to predatory crime is robbery. … Areas with greater cues of
disorder appear to be more attractive targets for robbery offenders, perhaps
because disorder increases the potential pool of victims without full
recourse to police protection, such as those involved in drug trafficking
and prostitution.
(Sampson and Raudenbush 1999: 630)

In plain language, this means that robbery is related to disorder. The authors,
however, are dismissive of what might be one of their most interesting findings:
‘Only for robbery does the estimated effect of disorder remain large’ (page 629).
Only for robbery? Isn’t robbery a serious problem?16 Since the only other violent
crime analysed is homicide, the rarity of which almost precludes a statistical
link to disorder, this is a rather cavalier dismissal. And, by the way, where is
aggravated assault in this analysis? If one is to make claims about measures of
violence, ignoring aggravated assault is a glaring omission
To be fair, ‘A New Look’ raises an important issue that is obscured by the
need to put ‘disadvantage and collective efficacy’ in the best light and to
diminish ‘Broken Windows’. In respects, the ‘Broken Windows’ formulation –
disorder to fear to serious crime to urban decay – is oversimplified. Clearly,
whatever causal sequencing is ultimately developed, ‘serious crime’ is an over-
simplified aggregation of a complex range of offences, some of which are forms
of debt settlement, others predatory, while others are situational. For example,
I would be hard pressed to say that all categories of robbery are sequentially
linked to disorder. The estranged lover who demands back rent from his
former roommate at gunpoint is probably not much affected by disorderly
behaviour or conditions. Plenty of other examples could be given from each
legal category of offence. My guess is that the main contribution of Sampson
and Raudenbush here is that while they do not disaggregate the crime data
themselves, their preliminary results on homicide and robbery suggest that
future analyses require disaggregation to establish sequencing, precursors, or
causes.
The final issue has to do with how Sampson and Raudenbush portray
‘Broken Windows’. Throughout the article, the authors exaggerate differences
between their findings and ‘Broken Windows’, ignore similarities and limit their
presentation of ‘Broken Windows’ policies to their construct: ‘strong “Broken
Windows” ’– whatever that may be.
Apparently by ‘strong version’ they mean something like the following:
‘What we would claim, however, is that the current fascination in policy circles
(see Kelling and Coles 1996; Kelling 1998) on cleaning up disorder through law
enforcement techniques appears simplistic and largely misplaced, at least in
terms of directly fighting crime’ (1999: 638). Note the citations: they suggest
that Kelling and Coles either reflect or contribute to such a narrow approach.
This is a deliberate misrepresentation. Fixing Broken Windows argues for a subtle
and nuanced approach to dealing with neighbourhood problems. Both ‘Broken
 ‘Broken Windows’: a response to critiques 139

Windows’ and Fixing Broken Windows are extremely cautious about the poten-
tials of police to control crime on their own. Fixing Broken Windows, for
example, while maintaining that police had an impact on crime in New York
City during Giuliani’s and Bratton’s administration, further argued that it was
only understandable in the context of the efforts of business improvement
districts (BIDs), the transportation authority, neighbourhood groups, the evolu-
tion of a community court and other community efforts (Kelling and Coles
1996: Chapter 4). Fixing Broken Windows argues for services, collaborations,
involvement with citizens, problem solving, involvement of the faith commu-
nity and private business and provides examples of all. (I suppose this was
acknowledged to me in a note that Sampson attached to a copy of the article in
which he indicates, ‘The results are complex and we think the ‘‘softer’’ version
of “Broken Windows” is more compatible with the data’ [personal communica-
tion, undated.]) ‘Softer’ version? I think I know what this means: it is a contrast
to the ‘tough’ version of ‘Broken Windows’ – nothing more than Sampson and
Raudenbush’s construct in the first place. To be sure, neither Wilson nor Coles
and I back away from law enforcement, but Fixing Broken Windows is very
explicit in its rejection of highhanded police tactics. Anyone who has read it is
well aware of this.
Reading Sampson and Raudenbush carefully, it is clear that their article is
misleading; ‘Broken Windows’ is inexcusably misrepresented. Moreover, the
idea, professed in their article, that: ‘This article is part of a larger effort to build
a social science of ecological assessment’ (Sampson and Raudenbush 1999: 639)
is pretentious. As a matter of fact, police researchers have been struggling with
measurement problems in neighbourhoods for decades. Sampson and
Raudenbush simply failed to learn from them – as is evident when their
research is compared with that of Braga (1999) and his colleagues. Sampson
and Raudenbush’s acceptance of the limitations of their sampling design in the
name of ‘advanced systematic social observation’ reflects a stunning lack of
‘street smarts’.
It does not take much imagination to understand, however, what is going
on here. The article was written and disseminated to gain maximum atten-
tion: it disproves ‘Broken Windows’; it is a major methodological
breakthrough. The dawning has taken place and left-centre criminologists and
sociologists are aware that ignoring disorder and minor offences is simply
unsustainable. Their quandary appears to be how to say something that
protects root causes and yet does not acknowledge that conservatives were
correct, both when they argued that minor offences mattered and that, if
police and criminal justice agencies were to take citizens’ concerns seriously,
and do something realistic about them, serious crime could be reduced short of
solving all of society’s problems. This does not mean that police can do it
alone or that poverty, racism, social injustice and family breakdown (the
equivalent of ‘root causes’ for the political right) should not be addressed.
They should be in their own right. But crime prevention should not be held
hostage to solving these problems.
140 George L. Kelling

Conclusion
The last decade of the 1990s proved to be a pivotal one for criminology and
criminal justice. The liberal ‘rediscovery’ of disorder and its consequences in
neighbourhoods represents a significant shift in criminological thinking. I have
no quarrel with liberals who want to ‘spiff up’ the idea a bit to make it more
palatable. This is exactly what the liberal New Yorker journalist Malcolm
Gladwell is doing in The Tipping Point, a trade book now approaching the best-
seller list in the US (Gladwell 2000). He argues that, in many respects, liberals
were right all along adding, provocatively, that Mayor Giuliani is really a liberal
when it comes to crime. Conservatives, he maintains, view character as the
source of crime. Liberals, on the other hand, maintain that social context
explains crime. Their mistake is that they view context too broadly: poverty, et
al. What is important is the immediate context. While Gladwell specifically uses
‘Broken Windows’ as a prime example, context is a core concept in both situa-
tional crime prevention (Cornish and Clarke 1986) and routine activity
analysis (Felson and Cohen 1980) as well. This explains, for Gladwell, why
circumstances changed so dramatically in New York’s subway: a small change in
the context, eradicated graffiti and ended fare beating, achieved a tipping point
that altered how people, including those disposed to crime, behaved in the
subway. But Gladwell makes his own points positively; he has no need to
diminish the insights of others – even conservatives.
My quarrel instead is with ‘scholarly’ works that deliberately attempt to
‘spin-off’ of ‘Broken Windows’, distort it, and in so doing, denigrate it, and then
use the distortion to enhance their own work. The shame is that just as the
importance of Gladwell’s ‘tipping points’ is not diminished by his forthright
recounting of ‘Broken Windows’ in New York’s subway, Sampson and
Raudenbush’s ‘collective efficacy’ would not have been diminished by a simi-
larly straightforward presentation of ‘Broken Windows’. Their work is important
in its own right; they don’t have to devalue it by misrepresenting other points of
view. My own guess is that their views will converge very nicely with ‘Broken
Windows’ as research progresses. (Putting aside the problems of sampling and
homicide data, it would be fascinating to disaggregate robberies to see if types of
robberies are related to types or levels of disorder.)
If the issue of the accurate representation of ‘Broken Windows’ was a mere
academic quibble, I would have been loath to respond to either of the above
articles. History has a way of sorting these things out. But the debate is not
academic; it is about crime control now. Political tracts are fine in the current
policy debate, but they should not be confused with scholarly work. Finally, if
Wilson and I had ‘only’ been right about the link between disorder and robbery,
I still would be delighted.

I would like to thank Catherine M. Coles, William Sousa and Michael Wagers
for their comments on drafts of this chapter. I would also like to thank Marissa
Potchak for her research assistance.
 ‘Broken Windows’: a response to critiques 141

Notes
1 The full explication of ‘Broken Windows’ is found in the original article, cited
above, and in George L. Kelling and Catherine M. Coles (1996) Fixing Broken
Windows: Restoring Order and Reducing Crime in Our Communities, New York: The
Free Press, and in George L. Kelling, ‘Broken Windows’ and Police Discretion,
National Institute of Justice Research Report, NCJ178259, October 1999.
2 Gary Stewart, ‘Black Codes and Broken Windows: The Legacy of Racial
Hegemony in Anti-Gang Civil Injunctions’, The Yale Law Journal, vol. 107, no. 7,
May 1998, pp. 2249–79. An ‘aversive’ racist, according to the author, is a smart
racist. One can tell that someone is a racist when he or she doesn’t talk about
race.
3 In a somewhat amusing example of attempts to redeem me from this encum-
brance, one doctoral student (not from Rutgers) has indicated to me that he is
working on a paper in which he conducts a contextual analysis of the original
‘Broken Windows’. He has identified what parts of the paper are the ‘good’ parts –
and I authored them – and what are the ‘bad’ parts – and they are authored by
Wilson. However, I was not a stranger to such ‘suspicion’. The Kansas City
Preventive Patrol Experiment (Kelling et al. 1974) had already put me at odds
with most of the police establishment during the 1970s. Moreover, the sociolog-
ical/criminological establishment, for example Albert Reiss Jr., had attacked the
Kansas City study even before its publication (personal experience and communi-
cations).
4 While this paper is relatively obscure, it represents for me an initial foray into
what later crystallised as community policing: seeking local legitimacy; the local
configuration of problems; the need for decentralised decision-making; the links
between order maintenance and crime control; the role of citizens in crime
control, etc.
5 Giuliani withdrew from the race in late May 2000 as a result of the discovery that he
was suffering from prostate cancer.
6 For those who don’t know the history, Bratton was chief of the transit police from
April 1990 to 1992. He then returned to Boston. In 1994 he returned to NYC as
Commissioner under Giuliani. Later, during Bratton’s administration, the transit
police were absorbed by the NYPD.
7 Fare beating was an unusually serious problem. Up to 250,000 people a day were
not paying their fare. This not only created an extraordinary sense of the subway
being a lawless place out of control, but the loss of income threatened the entire
infrastructure of the subway. Transit officials estimated losses at over $100 million a
year.
8 The use of laws against disorderly behaviour has also been an important means of
controlling gang behaviour – right now, a dangerous mix of drugs, drug selling, ‘turf’
and guns. Such matters have appeared in the US Supreme Court on at least one
occasion (Chicago v. Morales) and are bound to occur in the future. While the
Supreme Court invalidated one Chicago law that gave the police broad discretionary
authority to order gangs to ‘move on’, the Court literally invited Chicago to redraft a
somewhat more limited ordinance.
9 For a detailed discussion of this event, and the reasons he believes that it was not an
example of police abuse, see, James Fyfe (2000) ‘Reflections on the Diallo Case’.
10 This phrase, for me at least, has its origins with Mark H. Moore, my colleague in the
Kennedy School of Government at Harvard University.
11 I write political conservatives rather than Republicans because mayors like Daley in
Chicago and John Norquist in Milwaukee are not Republicans, but are conservative
nonetheless. Norquist, for example, has taken an extremely strong stance against the
use of federal funds in cities – once a funding mainstay of the liberal urban alliance
(Norquist 1999).
142 George L. Kelling

12 I became aware of this when I attended a panel on problem-solving at a meeting of

the American Society of Criminology. When I approached a presenter after the
session to discuss with her some points about my squeegeeing problem-solving
process in New York City – which eradicated squeegeeing in three weeks once it was
implemented – it was summarily dismissed because I resorted to law enforcement to
solve the problem, rather than solve the problem that led to young men becoming
squeegee merchants.
13 The most recent British critique of ‘Broken Windows’ is Benjamin Bowling (1999)
‘The Rise and Fall of New York Murder: Zero Tolerance or Crack’s Decline’, British
Journal of Criminology, vol. 39, no. 4 (Autumn), pp. 531–54. Basically, the article
reviews the homicide declines in New York City, restates the arguments of the last
half decade about the cause of those declines and dismisses ‘Broken Windows’ in
terms so similar to Matthews (1992) that one wonders if Bowling has limited his
familiarity to ‘Broken Windows’ to Matthews’ account of it. What is inexcusable
about Bowling’s article is that he not only misrepresents ‘Broken Windows’ – for
example, ‘Their [Wilson and Kelling’s] main policy recommendation to the police is
to “kick ass” ’ (548) – he had Fixing Broken Windows available to him. But this is
another story.
14 In fact, I have listened to many speakers who rely on it and practically could have
told them the chapter and verse of ‘Replacing’ they were using as they critiqued
‘Broken Windows’.
15 This may well be an example of ‘two peoples separated by a common language’
(English and Americans). If it is, it makes the point of how careful scholars must be
to understand the nuances of languages across cultures, especially technical terms
like ‘extra-legal’.
16 Interestingly, the experience in New York’s subway mirrors Sampson and
Raudenbush’s finding (although not their interpretation of it): when order was
restored, robberies declined dramatically.

References
Bittner, E. (1967) ‘The Police on Skid Row: A Study of Peacekeeping’, American Socio-
logical Review, 32: 699–715.
Bowling, B. (1999) ‘The Rise and Fall of New York Murder: Zero Tolerance or Crack’s
Decline’, British Journal of Criminology, 39(4): 531–54.
Braga, A. (1999) ‘Problem-oriented Policing in Violent Crime Places: A Randomized
Controlled Experiment’, Criminology, 37(3): 541–80.
Bratton, W. (1998) Turnaround, New York: Random House.
Chaiken, J. (1975) The Criminal Investigation Process: Volume II. Survey of Municipal and
County Police Departments, The Rand Corporation, R-1777-DOJ.
Chicago v. Morales, 1998 (97–1121), 177 Ill. 2d 440 687 N.E. 2d53.
Cornish, D. B. and Clarke, R. V. (1986) Situational Prevention, Displacement of Crime and
Rational Choice Theory, London: HMSO, 1–16.
Felson, M. and Cohen, L. (1980) ‘Human Ecology and Crime: A Routine Activity
Approach’, Human Ecology, 8(4): 389–406.
Fukuyama, F. (1999) The Great Disruption: Human Nature and the Reconstitution of Social
Order, New York: Basic Books.
Fyfe, J. (2000) ‘Reflections on the Diallo Case’, Subject to Debate. A Newsletter of the
Police Executive Research Forum, 14: 4.
Gladwell, M. (2000) The Tipping Point: How Little Things Can Make a Big Difference,
Boston: Little, Brown, and Company.
 ‘Broken Windows’: a response to critiques 143

Harcourt, B. E. (1998) ‘Reflecting on the Subject: A Critique of the Social Influence

Conception of Deterrence, the Broken Windows Theory, and Order-Maintenance
Policing New York Style’, Michigan Law Review, 97(2): 290–389.
Jacobs, J. (1961) The Death and Life of Great American Cities, New York: Vintage.
Kansas City Police Department (1977) Response Time Analysis: Volume II – Part I Crime
Analysis, Kansas City, MO.
Kelling, G. L. (1978) ‘Police Field Services and Crime: The Presumed Effects of a
Capacity’, Crime and Delinquency – Hackensack, NJ, 24(2): 173–84.
—— (1998) ‘Crime Control, the Police, and Culture Wars: Broken Windows and
Cultural Pluralism’, in Perspectives on Crime and Justice: 1997–1998 Lecture Series,
volume II, US Department of Justice, Office of Justice Programs, National Institute of
Justice.
—— (1999) Broken Windows and Police Discretion, National Institute of Justice Research
Report, NCJ178259.
—— and Coles, C. M. (1996) Fixing Broken Windows: Restoring Order and Reducing
Crime in Our Communities, New York: The Free Press.
——, Pate, T., Dieckman, D. and Brown, C. E. (1974) The Kansas City Preventive Patrol
Experiment, Washington, DC: Police Foundation.
Kennedy, D. M. (1997) ‘Pulling Levers: Chronic Offenders, High-Crime Settings, and a
Theory of Prevention’, Valparaiso Law Review, 31: 449–84.
McNamara, J. D. (1997) ‘Brutality in the Name of Public Safety’, Los Angeles Times,
August: M 1 and 3.
Maple, J. (1999) The Crime Fighter: Putting the Bad Guys Out of Business, New York:
Doubleday.
Matthews, R. (1992) ‘Replacing “Broken Windows”: Crime, Incivilities and Urban
Change’, in R. Matthews and J. Young (eds) Issues in Realist Criminology, London:
Sage Publications, 19–50.
Myerson, H. (1998) ‘Why Liberalism Fled the City …’, The American Prospect,
March–April: 46–52.
The Newark Foot Patrol Experiment (1981) Washington, DC: Police Foundation.
Norquist, J. (1999) The Wealth of Cities: Revitalizing the Centers of American Life, New
York: Perseus Books.
Podhoretz, N. (1999) ‘My New York’, National Review, June: 35–41.
President’s Commission on Law Enforcement and Administration of Justice (1967) The
Challenge of Crime in a Free Society, Washington, DC: US Government Printing
Office.
Sampson, R. J. and Raudenbush, S. (undated) ‘A New Look at Crime and Disorder in
Urban Neighborhoods’, Press Release. p. 1.
—— (1999) ‘Systematic Social Observations of Public Spaces: A New Look at Disorder
in Urban Neighborhoods’, American Journal of Sociology, 105(3): 603–51.
Sherman, L. W., Shaw, J. W. and Rogan, D. F. (1995) The Kansas City Gun Experiment,
National Institute of Justice, Research in Brief, US Department of Justice, Wash-
ington DC.
Siegel, F. (1992) ‘Reclaiming Our Public Spaces’, The City Journal, Spring: 37.
Skogan, W. G. (1990) Disorder and Decline: Crime and the Spiral of Urban Decay in Amer-
ican Neighborhoods, New York: Free Press.
Stewart, G. (1998) ‘Black Codes and Broken Windows: The Legacy of Racial Hegemony
in Anti-Gang Civil Injunctions’, The Yale Law Journal, 107(7): 2249–79.
144 George L. Kelling

Walker, S. (1984) ‘ “Broken Windows” and Fractured History: The Use and Misuse of
History in Recent Patrol Analysis’, Justice Quarterly, 1(1): 75–90.
Wilson, J. Q. (1999) Foreword in George L. Kelling and Catherine M. Coles, Fixing
Broken Windows: Restoring Order and Reducing Crime in Our Communities, New York:
The Free Press.
—— and Kelling, G. L. (1982) ‘Broken Windows: The Police and Neighborhood
Safety’, Atlantic Monthly, March: 29–38.
6 Ethnic minorities and
community safety
Marian FitzGerald

Introduction
The 1999 Macpherson report into the murder of the black London teenager
Stephen Lawrence set the tone and the content of the British government’s
commitments to ethnic minority communities on issues of crime and commu-
nity safety. These commitments were unprecedented in that not only were they
explicit and high profile but they were also attached to a wide-ranging
programme of action. This chapter examines what has been achieved to date
with regard to two key issues addressed by the programme of action – stop and
search and racist incidents. I raise a number of questions which will need to be
resolved if the government is to meet these specific commitments in the
context of its commitments to community safety more widely.
In March 1999, the Home Office, the Lord Chancellor’s Department and the
Attorney General published a Strategic Plan for 1999 to 2002, along with a
Business Plan for its implementation in the first year. The plans represented a
‘new approach in working together’, and the Foreword from the three ministers
invokes a protean notion of ‘community’. One of the priorities they identified
was ‘improving the confidence of the community as a whole, but especially the
ethnic minority communities, in the criminal justice system’ (Home Office
1999a). That is, the priority is based on the notion of a single community, but
there is a further assumption that – within a minority section of this community
– there is also a multiplicity of communities.
The context in which the government made this commitment was strongly
influenced by the report of the inquiry into the death of the black teenager,
Stephen Lawrence. Shortly after coming to office in 1997, the Home Secretary
appointed the senior judge Sir William Macpherson of Cluny to ‘inquire into
the matters arising from the death of Stephen Lawrence on 22 April 1993 to
date, in order particularly to identify the lessons to be learned for the investiga-
tion and prosecution of racially motivated crimes’ (Macpherson 1999: 6).
Three members were appointed to assist Sir William – a black bishop, the
retired deputy chief constable of a provincial force and the chair of the Jewish
Council for Social Responsibility. Late in 1997 they began a series of public
hearings which continued until the summer of 1998. These were largely
146 Marian FitzGerald

concerned with the murder itself and its investigation. Part two of the inquiry
began in the autumn of 1998, when they turned their attention to the investi-
gation and prosecution of racially motivated crimes. Eschewing ‘a narrow
interpretation of our terms of reference’, they chose at this stage to broaden
their focus to what they perceived to be the central problem, ‘the lack of trust
which exists between the police and the minority ethnic communities’
(Macpherson 1999: 311).
The report was eventually published in February 1999; and its overall
conclusion was that ‘the investigation was marred by a combination of profes-
sional incompetence, institutional racism and a failure of leadership by senior
officers’. It defined the problem of ‘institutional racism’ as:

The collective failure of an organisation to provide an appropriate and

professional service to people because of their colour, culture or ethnic
origin. It can be seen or detected in processes, attitudes and behaviour
which amount to discrimination through unwitting prejudice, ignorance,
thoughtlessness, and racist stereotyping which disadvantage minority
ethnic people.
(Macpherson 1999: 321)

The problem was not peculiar to the Metropolitan Police Service (MPS), it
affected all police services; and other institutions and organisations could not be
complacent: ‘Collective failure is apparent in many of them, including the
Criminal Justice system’ (ibid.: 321). The report made seventy recommenda-
tions, all of which were accepted by the government, albeit a small number
were accepted only in principle or subject to ‘further consideration’; and a
month after the report appeared, the Home Secretary published his Action Plan
for implementing them (Home Office 1999b). The process was to be overseen
by a steering group personally chaired by the Home Secretary. The success of
the Action Plan was to be measured ‘using existing indicators in 1999/2000’,
but ‘more comprehensive indicators’ would be put in place from 2000 onwards
(Home Office 1999b: 3).
Even though the main body of the report was taken up with the first part of
the inquiry and part two was covered in just six of its 300-plus pages, most of its
recommendations related to part two. They were addressed to a range of agen-
cies in addition to the police, including government departments and local
authorities. Those addressed to the police included a number which focused on
stop and search and racist incidents. Police performance in both areas featured
strongly in the report’s description of the ‘central problem’ of ethnic minorities’
lack of trust in the police. For many years they have also served as a dual focus
of research into issues concerning ethnic minorities in Britain as victims and as
suspects of crime.
The data available on both stop and search and racist incidents are of partic-
ular symbolic importance in showing the extent to which the government is
meeting its stated commitment to ethnic minorities. Setting the figures in
 Ethnic minorities and community safety 147

context, though, and examining the trends they show highlights a number of
largely unacknowledged tensions between these commitments and other
government priorities.
At the time of writing, published statistics for all forces were available only
up to the end of the financial year 1998–99. This means that the last two years
in these series effectively reflect, first, the situation inherited by the govern-
ment on coming to office; and second, the year of the Macpherson Inquiry.
The figures were published in the now annual Home Office publications on
‘race and the criminal justice system’. These were originally instigated under
s95 of the 1991 Criminal Justice Act and are commonly referred to as ‘s95
publications’.
In addition to these national figures, the Performance and Information
Bureau of the MPS provided me with statistics ad hoc on both stop and search
and racist incidents for the year 1999–2000. These shed further light on devel-
opments in London in the year following the Macpherson report. While
London may not be typical of other areas, it is the force most directly affected
by the inquiry; and, as this chapter highlights, it has a major influence on the
national statistics for both searches and racist incidents.

Stop and search

Statistical sources
Although police stops have long been a major source of tension between black
people and the police, it was not until 1993 that any obligation was placed on
forces systematically to monitor the issue. The 1984 Police and Criminal
Evidence Act (PACE) had already required all police forces to provide the
Home Office with raw figures on searches undertaken following a stop under s1
of the Act; and London had consistently accounted for nearly half of the
national total. Since concerns about the searching of black people in the capital
predated PACE (not least as an important factor in triggering the Brixton riots
of 1981), the MPS had already kept an ethnic breakdown of the figures for some
years. Numerous studies, however, supported widespread anecdotal evidence
that recording across all forty-three police forces in England and Wales was very
unreliable (Brown 1997).
At first, the police inspectorate (HMIC) simply asked forces to monitor
which of their records related to white people and which to a generic category
of ethnic minorities – even though the existing research suggested that this
would mask the extent of searches on black people relative to whites. Three
years later, though, the Home Office asked for the non-white category to be
broken down into ‘black’, ‘Asian’ or ‘other’ minorities (FitzGerald and Sibbitt
1997). The first published Home Office figures covered the financial year 1996
to 1997, and they compared the numbers of recorded searches with the 1991
Census population figures for each police force. The 1997 figures confirmed
that, on this yardstick, black people were significantly more likely to be
148 Marian FitzGerald

searched by the police than white people but there was no comparable over-
representation of Asians. This pattern was repeated again the following year –
that is, the year which ended just as the Macpherson Inquiry got underway.

Recent trends
Figures for the year during which the Macpherson Inquiry was conducted,
however, showed a fall of 15 per cent in the number of black people searched in
England and Wales, with a more modest drop (6 per cent) for Asians. Read
superficially, the figures might imply that the inquiry was already proving effec-
tive in reducing the level of searches on black people relative to whites. In fact,
the national figures give a misleading picture of trends for two inter-related
reasons.
The first is that the different minority populations are not distributed evenly.
Nearly two-thirds of all black people (62 per cent) live in London and the next
largest black community is in the West Midlands which accounts for a further 9
per cent. The Asian population is more dispersed; but again the main areas of
settlement are London and the West Midlands, which account respectively for
35 and 16 per cent of the population. By contrast, only 17 per cent of the white
population lives in these two conurbations put together. What happens in these
two force areas, therefore, will significantly affect any presumed ‘national’
picture for both the black and the Asian group; and it may distort their national
showing relative to the white population. In effect, national trends for black
people are largely determined by what happens in London.
The second reason why the national figures are misleading is that there is
wide variation in the recorded use of the power by different forces. The Home
Office ‘s95’ figures (Home Office 1999b) gave a national figure of twenty-two
PACE searches per thousand population in 1997 and 1998; but the figures for
individual forces ranged from four to just under a hundred. For London the
number was fifty-one per thousand; but in the West Midlands it was only nine-
teen. In the year of the Macpherson Inquiry, the national average remained the
same; but the London figure fell to forty-five per thousand and that in the West
Midlands to eleven. Yet there was almost no change in the proportion of
searches on black people in either area. In London the proportion of black
people searched fell by one percentage point (from 26 to 25 per cent); but in
the West Midlands it rose from 14 to 15 per cent, with a higher rise for Asians
(from 17 to 20 per cent).
The apparent fall in the proportion of searches on black people during the
year of the inquiry, therefore, did not represent any marked drop in searches on
black people per se. It was largely due to the fall in searches on all groups in
London. Meanwhile, the rise in the proportion of black people searched in the
West Midlands had no effective impact on the national figures, despite the rela-
tively large number of black people in the area. This was because the fall
occurred in a force which contributed relatively little to the national total in
the first place.
 Ethnic minorities and community safety 149

Developments in London
On closer inspection, the figures for London show that this fall during the year
of the Macpherson Inquiry was especially precipitate in the early part of 1999 –
the point at which the report was awaited. At the same time there was a signifi-
cant increase in recorded crime in London; and sections of the media were
sympathetic to claims by some police officers that the two phenomena were
associated.
Anticipating the wider criticisms the inquiry was likely to level at them, the
MPS had begun to try to seize the initiative. Its new approach to tackling racist
incidents is described in the next section; but it had also recognised stop and
search as a key issue on which action was needed. At the beginning of 1998,
therefore, it designated five pilot sites (later increased to seven) to try out a new
approach to stop and search. The initiative largely consisted of intensified
scrutiny of two things: officers’ performance within and across the pilot sites;
and statistical data. Its stated aims were to ensure that the power was used both
more effectively and fairly. Progress was to be closely monitored throughout and
the results were to be evaluated at the end of the first year.
In the course of this monitoring and evaluation it became apparent that the
data already held by the MPS were a potentially rich source of new insights into
the pattern of searches and the factors which drove these. Research into the
impact of the pilots required more detailed analyses of these data than had ever
previously been undertaken, and it generated further information ad hoc to
complement these analyses (FitzGerald 1999a; 1999b).

Effectiveness
The interim evaluation report in the summer of 1999 showed a marked
improvement in the arrest rate from searches – which has conventionally been
used as the key measure of success. However, it also showed that this success
should not be taken entirely at face value. Although the MPS was known to
record a higher proportion of its searches than other forces (FitzGerald and
Hale, forthcoming), it became apparent that many searches in London were
still going unrecorded. Where searches resulted in an arrest, officers might
simply not bother to complete a record for the search in addition to doing the
paperwork for the arrest. Once the search figures were put under the twin spot-
lights of the pilots and the Macpherson Inquiry, this provided a major incentive
to rectify these omissions. Thus the apparent improvement in the arrest rate
gave the impression that the power was being targeted more effectively; yet
much of it was probably due simply to improved statistical housekeeping.
The focus on the improved arrest rate distracted attention from a develop-
ment which was arguably far more important in terms of effectiveness. It was
taking place in the context of a major fall in searches overall which had set in as
the Macpherson Inquiry started. The fall in searches became more precipitate as
the inquiry moved into phase two and dropped more sharply still in anticipation
of the report. While the burst of statistical housekeeping referred to above meant
150 Marian FitzGerald

Figure 6.1 Arrests and searches: London 1997–1999

Source: MPS Performance Information Bureau (PIB)

that the number of arrests attributed to searches held up for a while, by the last
quarter of 1998 they too started to plummet. While the improved arrest rate was
being represented as a success, in practice searches were now contributing far less
to arrests overall (see Figure 6.1).
It cannot be assumed, however, that the fall in arrests from searches had any
serious consequences for community safety. The largest proportion of arrests
from searches was for drugs; and data from the pilot sites confirmed that by far
the majority of these drugs arrests were for personal possession of small amounts
of cannabis. Cannabis use is not believed to be causally associated with other
types of offending; nor does it appear to be a matter of public concern – except
inasmuch as the fact that it is illegal may serve to legitimate potentially discrim-
inatory police searches (Young 1994; Runciman 2000). The second largest
category of searches was for stolen property; and a fall here also seemed unlikely
to have any significant impact on the crime figures. Burglary is by far the largest
category of property crime, but only a small minority of burglaries are cleared up
through arrests of any sort and only a small minority of these would have been
the result of a search (see FitzGerald and Sibbitt 1997: 70).
On the other hand, searches for offensive weapons and firearms or for going
equipped to steal or burgle together accounted for under 15 per cent of search
arrests early in 1998. Their relative insignificance in the context of searches
overall meant that their potential importance in the context of community
safety was largely overlooked. Yet searches for offensive weapons and firearms or
for going equipped to steal or burgle consistently produced the majority of all
 Ethnic minorities and community safety 151

arrests for these preparatory offences. So, while the decrease in searches gener-
ally might have little influence on levels of crime overall, the fall in arrests in
these particular categories might nonetheless have a disproportionate impact.
The claims of a link between the rise in crime and the fall in searches
overall, seemed almost certain to be spurious; although they were as difficult to
disprove as to prove using the administrative statistics available. Academic
analysis, however, did find a completely counter-intuitive statistical relationship
between the sudden fall in searches and the marked rise in crime which
followed in the early months of 1999 (Penzer 1999). The explanations for this
finding are likely to be complex. It cannot be taken to imply that the fall in
searches per se caused a rise in crime or even that the rise in crime was directly
linked to the fall in searches for offensive weapons, firearms and going
equipped. However, other research has found an association between a sudden
fall in the use of searches and a rise in crime (NACRO 1997). The findings
from the MPS study suggest that it is worth undertaking further work to explore
the reasons for this pattern; and this might usefully focus on the preventive
effect of the power – both direct and indirect.

The issue of disproportionality

In addition to confirming existing concerns about drugs searches, the MPS
study systematically confirmed other well-rehearsed complaints about the way
the power was used. It also revealed new ones, though, and it began to set the
question of ‘disproportionality’ in a different light. The final report (FitzGerald
1999b) showed that there was more support in principle for the power as a
means to improve community safety than might have been expected. The
support came from all ethnic groups, including from young people who had
themselves been searched. However, there was widespread dissatisfaction with
the gratuitously abrasive way it was applied by many police officers.
The report also highlighted considerable variability in the extent to which
the targeting of searches was directly related to patterns of crime and it showed
that searches seemed to be used on occasion beyond what was legitimate in
terms of PACE. Of particular interest in the context of this chapter, however,
was the way in which officers themselves often rationalised this by invoking
considerations of community safety. Thus searches were seen as an effective way
of gaining intelligence on people who were ‘known’ to the police; and the
power was also used to break up and move on groups of youths whose very pres-
ence constituted a nuisance or a perceived threat to others. Yet these purported
benefits to some sections of the community were having a negative effect on
those sections of the community who were on the receiving end of the searches.
In the case of people already known to the police, they constitute a form of
harassment. At the same time, the use of the power as a means of social control
was arguably criminalising people who were not previously known to the police
but who were arrested for possession of cannabis where searches were used as a
means of social control on groups of young people.
152 Marian FitzGerald

Figure 6.2 Searches, population and suspects: London 1998–1999

 Ethnic minorities and community safety 153

With regard to the ethnic patterns in the search figures, the study exposed
numerous complexities in interpreting local statistics. It established that
comparisons with the local population in general were irrelevant and potentially
misleading for many reasons, not least because of the non-resident population
represented in the search figures. In some areas people from outside the area
accounted for the majority of those people searched; but there were further
ethnic variations within this. The two main influences on the ethnic patterns in
the search figures, though, appeared to be age and the extent to which different
groups featured among the suspects described in crime reports. This is illustrated
by Figure 6.2 which compares the pattern of searches on different ethnic groups
in London with: (a) the population at large; (b) the population of secondary
school age; and (c) the descriptions of suspects of crime.
Figure 6.2a shows the basis on which ‘disproportionality’ has usually been
calculated. It confirms the extent to which, measured in this way, black people
are over-represented in the search figures in relation to their presence in the
population at large. Yet Figure 6.2b shows that this ‘over-representation’ reduces
considerably when age is taken into account since there are important varia-
tions in the age structures of the different ethnic groups. This is especially
relevant in the context of community safety. For the young and the old may be
thought of as different communities of interest whose conflicting needs may
sometimes be difficult to reconcile. In many areas white people will constitute a
much higher than average proportion of the old population while ethnic
minorities constitute a much higher than average proportion of the youth popu-
lation than is apparent from figures for the population as a whole. In over half
of the thirty-two London boroughs, ethnic minorities account for 45 per cent or
more of the school age population; and in some they constitute a majority of
the young people in the area (FitzGerald 2000). This not only highlights the
importance of controlling for age with some precision when making inter-
ethnic comparisons. It also flags up the danger of racialising what are essentially
age-related differences by the uncritical use of crude ethnic statistics.
In fact, the closest match for the ethnic composition of the search figures in
Figure 6.2 is with descriptions given to the police of suspects of crime (6.2c).
Importantly, the study also showed that black people were no more likely than
white people to be stopped by police officers acting entirely on their own discre-
tion. It established that black people who were searched were as likely as white
people to have previous criminal records and they were arrested at the same rate
following searches. Taken together, the evidence strongly suggested that – far
from disappearing – disproportionality would actually get worse if officers were
allowed only to search people in connection with crime reports. The only
obvious way of achieving a reduction in the ethnic differential between black
and white people in the search statistics, therefore, might be to weight the
police response to victims of crime according to their description of the suspect.
Yet it is hard to imagine how such an approach could be acceptably presented in
the context of local community safety strategies, regardless of any implications
for their successful implementation.
154 Marian FitzGerald

On the other hand, the research identified significant issues with regard to
searches on Asians which had been overlooked in a debate framed by tradi-
tional assumptions about (dis)proportionality. As Figure 6.2a also shows, Asians
were not searched disproportionately when compared with their presence in the
population at large or even in relation to their presence in the school age popu-
lation (6.2b). However, they were searched more than would be expected from
descriptions of suspects in crime reports (6.2c). Moreover, this disparity was
much starker in the few areas with relatively large Asian populations. At the
same time, the arrest rates from searches for Asians at the beginning of the
study were much lower than for white or black people; and those searched were
much younger than the average for all groups. Proportionately many fewer of
the Asians had criminal records; they were more likely to be searched in groups;
those groups tend on average to be larger; and a higher proportion of searches
on Asians were for drugs.
In other words, the Asians – and in particular the poorer, younger groups
within this supposed community – were at special risk from the use of the power
as a mechanism of social control. Their age structure meant that the cohort at
the peak age for offending was now much larger than in the past, so increasingly
large numbers of these young people had begun to experience the confronta-
tional way in which the power was often applied. In so far as it was starting to
criminalise them for possession of cannabis, the study argued that searches
might yet prove the equivalent of the ‘Sus’ laws which did so much damage to
police relations with the parallel generation of black youth twenty years ago
(Demuth 1978).
For black people the implications of the research were different. They chal-
lenged the orthodox view that the ‘over-representation’ of black people in the
figures relative to their presence in the population at large could serve as a
proxy measure for police discrimination. But this did not imply that there was
no particular cause for concern in relation to black people. Rather, black people
were disproportionately suffering from the abuse of the power to harass people
‘known’ to the police. It also meant that the overbearing and aggressive manner
of officers (whether or not they were using the power legitimately) was having a
disproportionate impact on black people – including black people who were
innocent of any crime. The bottom line, however, was that the extent to which
the power is used would make little difference to ‘disproportionality’ as tradi-
tionally conceived. Whether the power was used more or whether it is cut back
to the minimum, the black-white ratio within it was unlikely significantly to
change.

Implications
Taken together, these findings posed a need for new thinking both about the
effectiveness of the power and about whether it was being used fairly. Searches
in general were not being used effectively to tackle crime. Although they had
fallen dramatically, the fall had not occurred as the result of any explicit inter-
 Ethnic minorities and community safety 155

vention – still less one which was coherently designed to target the power better
or to address the problems associated with it. Yet, from the point of view of
community safety, there was more risk associated with this sudden fall than
might have been expected and some types of search (albeit a minority) might
yet prove indispensable for the prevention of crime unless equally effective and
more acceptable alternatives could be found.
At the same time, the disproportionate impact of the power on black people
and the potential for damage to police relations with future generations of
young Asians added particular urgency to the need to tackle generic problems
with the way the search power was used. However, it was essential to recognise
that these were problems which affected the community as a whole and to
tackle them as such. Any attempt at remedial action which conceived of the
problems as if they only affected the ethnic minority communities, would be
treating symptoms rather than causes. The danger was that these problems were
simply being scaled down until the Macpherson effect dissipated while crime
would remain at the level to which it had risen during the time when the fall in
searches was at its sharpest. Meanwhile, trends in recorded racial incidents also
give cause for concern. They too call into question conventional interpretations
based on simple headline statistics and suggest that these data also need to be
seen in a wider context.

Racist incidents

Statistical sources
The issue of racial violence and harassment had long been a problem for
minority communities in certain areas, but the issue reached the national policy
agenda in the late 1970s and early 1980s. In 1985 the Association of Chief
Police Officers responded to sustained criticism of the police response by
adopting an all-embracing definition of a ‘racial incident’. This attempted to
overcome the frequent tensions between ‘victims’ and officers’ perceptions of
racial motivation by requiring forces henceforth to record all incidents reported
to them in which it appears to the reporting or investigating officer that the
complaint involves an element of racial motivation, or which includes an alle-
gation of racial motivation made by any person. That is, the definition
deliberately embraced non-criminal forms of harassment and it did not require
racism to be the primary motivating factor. Hard evidence was not required,
only a perception or ‘an allegation’ and officers were not to have the final deci-
sion about whether the incident should be recorded as racially motivated.
These figures, too, were soon requested by HMIC – albeit forces were only
asked for a raw number. There was no suggestion that it might also be relevant
to break the figures down by offence type or by the ethnicity of the victims and
perpetrators. Publication of force totals was ad hoc at first, but the figures were
among the few sources of relevant statistics available in 1991, so they appeared
in the s95 publications from the outset. Meanwhile, the Home Office British
156 Marian FitzGerald

Crime Survey (BCS) had introduced an ethnic oversample in 1988; and in the
same year it asked a question about racial harassment (Mayhew et al. 1989).
Victims were asked if they believed what had happened to them was racially
motivated. However, the question was only asked of black and Asian respon-
dents and, while the survey covered threats as well as crimes, it did not cover
the low level nuisance behaviour which was typical of much harassment.
A study carried out by the Home Office in 1997 showed that few police
forces collected detailed figures on the racial incidents they recorded (Maynard
and Read 1997). Where these were available, though, they broadly confirmed
the consistent pattern of victims’ experiences described by the British Crime
Survey and earlier studies (FitzGerald and Ellis 1989). Unsurprisingly, the
largest single category of racist incident (38 per cent) was ‘verbal abuse’, which
is the most common currency of antagonism – whether between individuals or
groups. Along with damage to property, verbal abuse accounted for well over
half of all incidents across the fourteen forces covered by the study, with miscel-
laneous ‘assaults’ making up a further 21 per cent. Only 2 per cent of racial
incidents were classified as ‘serious crimes’.
In the same year, the Crown Prosecution Service published the first report of
its ‘Racial Incidents Monitoring Scheme’. This included a breakdown of the
type of charges put by the police to the Crown Prosecution Service (CPS) in
the previous year which the CPS viewed as racially motivated. It too showed a
preponderance of less serious offences, with 49 per cent classified as public order
and a further 14 per cent as criminal damage. Just over a quarter (27 per cent)
were assaults (CPS 1997).
Part one of the Macpherson Inquiry was concerned with the most serious of
all racially motivated offences: murder. But in part two it specifically criticised
the police for failing to realise: ‘the impact of less serious, non-crime incidents
upon the minority communities’(Macpherson 1999: 313). It recommended a
simplified version of the ACPO definition and it insisted on changing the term
‘racial’ to ‘racist’. The new definition adopted by the Home Secretary and by all
police forces was: ‘an incident which is perceived to be racist by the victim or
any other person’.

Recent trends
Despite its limitations, the BCS had become increasingly important as a bench-
mark against which to measure trends in the police figures. For it not only
reflected the underlying levels of harassment experienced by victims, it also
asked them whether they had reported incidents to the police. The police
figures steadily began to rise through the early 1990s (see Figure 6.3). Although
the British Crime Survey showed no change in the underlying rate of racist
victimisation, it did suggest that there was a significant shortfall in the numbers
of incidents being recorded by the police relative to those reported to them
(Aye-Maung and Mirrlees-Black 1994; FitzGerald and Hale 1996a; Percy 1998).
Since there was no evidence of any significant increase in levels of reporting
 Ethnic minorities and community safety 157

Figure 6.3 Recorded racial incidents in England and Wales, 1991–1999

Source: Home Office s95 reports

either, the gradual rise in the police figures up to 1996 mainly represented a
closing of this recording gap.
As with searches, the Macpherson Inquiry appeared already to have had a
significant impact on the racial incident figures before it even reported.
Following year-on-year rises of less than 10 per cent since the first s95 report,
the records for the financial year 1998 to 1999 showed a 66 per cent increase
(see Figure 6.3).

Developments in London
Meanwhile, the MPS had been proactive on the issue of racial incidents as well
as with regard to stop and search. The Race and Violent Crimes Task Force, set
up at New Scotland Yard in August 1998, rapidly began to establish a high
profile in the media and elsewhere. Under the charismatic leadership of Deputy
Assistant Commissioner John Grieve, it publicly took a robust – even aggressive
– approach to tackling racism. It mounted poster campaigns, opened the MPS
to its critics by setting up a lay advisers group to assist it and took road shows
out to public meetings across London. Within the force, the unit also spawned a
wide range of initiatives to raise standards and to improve performance and co-
ordination, including increasing the intelligence available for dealing with what
it increasingly referred to as ‘hate crime’.
The term ‘hate crime’ subsumed both racial and homophobic incidents. In
recent years, both of these had increasingly fallen within the remit of
‘Vulnerable Persons Units’ which had developed somewhat ad hoc primarily to
deal with domestic violence. Seizing the opportunities for restructuring in the
move to borough-based policing, the Race and Violence Crimes Task Force had
a specialist unit set up in each area. The core business of the units was their
responsibility for racial and homophobic incidents as well as domestic violence
(although there were some local additions to this), but the term ‘vulnerable
158 Marian FitzGerald

persons’ was dropped. Despite the fact that their remit was limited to specified
groups of victims, the new units were named ‘Community Safety Units’.
While the total number of recorded incidents in England and Wales rose
nationally by 66 per cent in 1998–1999 (i.e., the year of the Macpherson
Inquiry), they increased by 188 per cent in London. This meant that the MPS
now accounted for nearly half of all recorded racial incidents; and if London is
removed from the total, the percentage increase in 1998–1999 would have been
50 per cent instead of 66 per cent. Figures available from the MPS at the time
of writing showed a further, even larger increase for 1999–2000. The number of
racial incidents recorded in London rose a further 211 per cent in the year
following publication of the Macpherson report.
However, the more typical forms of victim experience appear consistently to
have accounted for a much smaller proportion of incidents in London than
those recorded elsewhere. The 1997 Home Office study, for example (op. cit.),
showed that only 12 per cent of incidents were classified by the MPS as verbal
abuse compared with an average of 38 per cent for all the forces in its sample.
The CPS monitoring report showed a similar pattern. In 1998–1999, 50 per
cent of charges were for public order offences in non-London forces compared
to 28 per cent for assault; but the MPS figure for both was 37 per cent. This
relative absence of incidents of low level harassment recorded by the MPS was
identified as a problem by the police inspectorate in a report on community and
race relations conducted in 1999. It fully acknowledged the pioneering work of
the Race and Violent Crimes Task Force, but it specifically noted that ‘the MPS
was failing to capture those incidents referred to by some as low level harass-

100

80

60

40

20

Figure 6.4 Trends in types of racial incidents recorded by the MPS, 1997–2000
Source: MPS Performance Information Bureau (PIB)
 Ethnic minorities and community safety 159

80
70
60
50
40
30
20
10
0

Figure 6.5 Victims and suspects of racist incidents by ethnic group, MPS 1997–2000
Source: MPS Performance Information Bureau (PIB)

ment, but which in fact do so much damage to the quality of life of victims’
(Home Office 2000: 48).
The London figures during and after the Macpherson Inquiry show huge
increases but they show little evidence to suggest that the concerns of
Macpherson and the inspectorate with regard to ‘low level harassment’ are
being addressed. The recorded increases were primarily in relation to crime
rather than non-crime, and there was a further change in the balance of types of
crime within this. Notifiable offences came to account for a much higher
proportion of the crime incidents, and a progressively higher proportion of these
were violent offences (see Figure 6.4).
Meanwhile, the MPS figures also present a more complex picture of the
ethnicity of victims and of suspects than has been apparent from the terms in
which the debate around racial incidents has usually been conducted. The
debate has tended to assume that the victims of racism were primarily – if not
exclusively – members of visible ethnic minorities and that the perpetrators
were white. Yet both official definitions embrace all incidents irrespective of the
ethnicity of the victim or the perpetrator, and the London figures had always
shown a higher number of white victims than might have been supposed.
Following the publication of the Macpherson report, the proportion of
victims of racist incidents recorded as white actually increased in the MPS. In
1999–2000, the proportion of white victims of racist incidents in London nearly
equalled that of Asians, with a slightly smaller proportion of victims recorded as
black. Meanwhile, although whites were still the main perpetrators (see Figure
160 Marian FitzGerald

6.5), a higher proportion of suspects were recorded as black than in previous

years.

Implications
The statistics published by the Home Office tell us little about the real scale or
nature of the problem of racist harassment and even less about how effectively
the problem is being tackled. However, the BCS and the CPS reports provide
some check, and the detailed information held by the MPS is capable of shed-
ding further light on the situation in London.
With regard to the scale of the problem, the sudden increase in the crude s95
figures in 1998–1999 could have been due to any or all of three developments:
(a) an increase in reporting; (b) an increase in the reported incidents which
were actually recorded by the police; and/or (c) an increase in the underlying
problem. Unfortunately, the BCS did not over-sample ethnic minorities in
1998, so there is no authoritative means of telling the extent to which each of
these factors was implicated. Some tentative inferences can, however, be drawn.
It seems likely that reporting will have increased during a sustained period of
sympathetic media coverage. High-profile statements of commitment by senior
police officers may have encouraged more victims to report in the expectation
that they would be taken seriously. Also, the Macpherson Inquiry had empha-
sised the extent to which racism could be implicit as well as explicit and this
may have encouraged more victims who were reporting crimes additionally to
identify any racist element to these. If so, this would also help to explain the
increase in the MPS figures; for one would expect the serious crimes which were
such an important component of this increase to have been reported in any
event.
Additionally, though, the increase in the figures may include a backlash
element. The 1999 study of searches (FitzGerald 1999a) produced incidental
material which suggested that publicity surrounding the Macpherson Inquiry
was indeed creating resentment among some white people. They believed that
ethnic minorities were receiving privileged treatment; and several officers I
interviewed spontaneously reported being taunted by white youths claiming
that the police were picking on them because they no longer dared to search
black people. One officer described a similar backlash among white victims of
crime as follows:

You know, people come up to you and they say ‘I’ve just had this bloke
smash the windows of my car in. What are you going to do about it?’ And
you say ‘I’ll take a report and that’s the best I can do.’ And they’re turning
round now and they’re saying ‘I suppose, if I was a black person you’d have
detectives down here looking at my car’, etc.

Anecdotally, it had always been claimed that some white victims of crime
where the perpetrators were black have insisted de facto that these incidents
 Ethnic minorities and community safety 161

were racially motivated. It seems entirely possible that such reporting may have
increased in the context of the inquiry; and this would help to explain the
increase in the proportion of white victims and of black suspects in the MPS
figures.
Recording also seems likely to have increased as a result of the inquiry which
is likely to have catalysed the pace at which the police had already begun to
close the recording gap identified by the BCS. Alongside this, though, a back-
lash element may also have been at work. Thus the officer quoted above had his
own views on white victims’ perceptions that black victims would receive
special attention. He had added, ‘They’re probably right’.
The HMIC report had also noted:

The view was widespread [among police officers] that the rationale behind
the definition of racist incidents and the prescriptive approach to their
recording was to provide a preferential service to minority ethnic commu-
nities thereby prejudicing the policing needs of the rest of society.
(HMIC 2000: 44)

This suggests that some officers may themselves have actively connived in the
white backlash by recording the type of perverse reporting referred to above.
Both reporting and recording in this instance would, of course, fall strictly
within the letter of the definition – even though they ran directly counter to its
spirit.
The white backlash to the Macpherson Inquiry, however, will not only be
reflected in the increase in the incidents reported to and recorded by the police.
It seems almost inevitable that it also produced a rise in the underlying level of
racial violence and harassment. The findings of existing studies of the actual
and potential perpetrators of racial violence and harassment suggest that many
come from marginalised groups within the white population. As justification for
their activities, they draw heavily on perceptions of preferential treatment for
minorities such as those cited above (Sibbitt 1997; Hewitt 1992 and 1996).
While the increase in the underlying level of racial violence and harassment
may have occurred for reasons beyond the control of the police, it adds urgency
to the need to show that the problem is being tackled effectively once it has
occurred. However, the evidence of improvement here is, at best, limited. The
most obvious measure of effectiveness comes from the CPS reports and from
MPS data regarding judicial disposals, although these relate only to the
minority of cases which are considered for prosecution. They show some
improvement, but it is by no means commensurate with the scale of increase in
the numbers of incidents recorded or their greater seriousness.
In 1998–1999 – the year which saw a 66 per cent overall increase in
recorded incidents and a 188 per cent increase in London – there was also an
increase in the number of actual charges involving racial motivation put to the
CPS by the police. However, for England and Wales as a whole, the rise was
only 10 per cent; and if the figures for London are removed it was only 4 per
162 Marian FitzGerald

cent. Meanwhile in London, 12 per cent of recorded incidents resulted in judi-

cial disposals (that is a charge or caution) in 1998–1999. This was exactly the
same proportion as in the previous year; and in the following year there was an
increase of just 1 per cent. Yet nearly two-thirds of all incidents were by now
classified as violent notifiable offences compared to only 15 per cent in
1997–1998 (see Figure 6.4 above).
The Macpherson Inquiry and the government’s response to its report signifi-
cantly raised awareness of the problems of racial violence and harassment,
along with expectations that the problem would be tackled with more vigour
than in the past. Evidence of its impact to date, though, is mixed. The increase
in the numbers recorded is encouraging in some degree, but is not matched by
evidence of improved effectiveness – certainly in terms of dealing with the
most common forms of the problem. These ‘low level’ incidents have a perva-
sive impact on the quality of life of victims and potential victims and this is
compounded by heightened levels of fear associated with the problem
(FitzGerald and Hale 1996b). For the present it seems possible that racist inci-
dents in general may have increased. At the same time the heightened focus of
the last two years on the relatively rare instances of extreme racial violence
will have done nothing to allay the constant sense of threat felt by people
whose level of fear of crime is already much higher than average (FitzGerald
and Hale 1996b).

Discussion and conclusion

The Macpherson report was important as a symbol of the government’s commit-
ment to race equality in general. It also helped define that commitment in
relation to the criminal justice system with regard to ethnic minorities both as
suspects and as victims of crime. Yet there is little evidence that the expectations
it raised on the part of ethnic minorities have been fulfilled; and at the same
time tensions may arise between the priorities established by the report and the
local priorities which have to be met under the Crime and Disorder Act (1998).
The report’s central finding of ‘institutional racism’ does little to resolve these
tensions. For, once its cathartic effect has dissipated, it lacks analytical content.
It has no legal force and, above all, it offers no practical basis for action. The
problem as identified simply is the problem as described. The optimistic view is
that this tautology will be enough to galvanise organisations to discern the
causes of the problem for themselves and tackle them effectively. More
pessimistically, though, the label of institutional racism may have three effects.
First, people who really are the cause of the problem may co-opt the idea
that its nature is institutional rather than individual in order to deny any
personal responsibility. Second, other individuals may feel personally accused
because, collectively, they are the organisation, and the verdict that they are
guilty of ‘unconscious’ racism is almost impossible to challenge. This may
generate a sense of impotence at best and resentment at worst. Third, now the
label is official, it is difficult to imagine it officially being lifted. But it is equally
 Ethnic minorities and community safety 163

difficult to imagine how members of ethnic minorities can have confidence in a

criminal justice system which is permanently marked in this way.
Inasmuch as the Macpherson report has any or all of these unforeseen
consequences, it may further compound the potential tensions between the
government’s objectives on different fronts. To avoid this, additional thinking is
needed which takes full account of the complex and fluid character of individuals’
sense of community attachment. In particular, it needs to recognise that people do
not define themselves exclusively in a single dimension – whether of ethnicity,
gender, age or any of the other intersecting strands described by Yinger (1986).
This means that individuals may be differentiated from each other on one dimen-
sion, but they have far more in common on other dimensions than may appear to
divide them.
Achieving the government’s objective of increasing the confidence of ethnic
minorities in the criminal justice system does not require additional statistics so
much as a more rigorous interrogation of those which are already available. In
particular, it needs to take full account of the limitations of statistics. These
include not only the crudeness of the ethnic categories used, which fail, for
example, to capture especially vulnerable groups such as refugees and asylum
seekers (many of whom will be subsumed within the omnibus ‘white’ category).
Critical attention also needs to be paid to the ways in which administrative
changes in recording practices may give a misleading impression of trends.
Analyses of these data, in turn, need to take account of the fullest possible
range of other relevant factors before assumptions are made about the reasons
for apparent ethnic differences shown by the headline figures. Above all,
though, the figures need to be interpreted in the light of understandings from
qualitative research – especially when they are used to inform developments in
policy and practice.
At the same time, systematic consideration needs to be given to the impact
of criminal justice policy at its widest on different groups. Challenges to the
abolition of the right to elect Crown Court trial have drawn attention to the
potential implications for black defendants (Bridges 2000). However, less
publicity has been given to other developments whose disproportionately
adverse effect on ethnic minority suspects and offenders do not appear to have
been taken into consideration at all. For example, the right of prospective
employers to check on job applicants through the new Criminal Records
Agency seems likely further to disadvantage young black men in the labour
market.
Coupled with this, though, it is essential to make the link between aspects of
criminal justice policy and practice whose disproportionate impact on certain
groups highlights wider areas of concern. For example, the majority of the 3,000
miscarriages of justice predicted by Bridges (2000) as a result of the abolition of
the right to trial by jury will be white. Moreover, the victims (of whatever
ethnic origin) may be affected differentially depending on where they live – or
even which court they appear before within that area. Thus the Home Office
Criminal Statistics suggests that there are considerable variations by area in the
164 Marian FitzGerald

rate at which magistrates themselves refer triable-either-way cases to the Crown

Court; and a recent Home Office study highlighted the persistence of uneven-
ness in sentencing practice in different areas (Flood-Page and Mackie 1998).
Indeed, an important finding of Hood’s study of sentencing of ethnic minorities
in the Crown Court related to the differences in outcomes for defendants in
general between different Crown Court centres within the West Midlands. It
was only in some that he found further differences in the sentencing of the
main ethnic group (Hood 1992).
As long as the criminal justice system remains capable of producing different
outcomes for similarly placed individuals irrespective of ethnicity, it cannot be
expected to treat any given group with consistent fairness. It is doubtful, there-
fore, that the government can achieve its particular aims with regard to ethnic
minorities independently of its broader objective of ‘improving the confidence
of the community as a whole in the criminal justice system’. Ironically, if other
sections of the community believe that the government is trying to address the
unfair treatment received by minorities while neglecting their own grievances,
this may further undermine their own confidence in the system, as well as
having damaging consequences for minorities themselves.
This last point is highly relevant to the issues raised here about local commu-
nity safety strategies. Especially in urban areas (where the larger part of the
ethnic minority populations tend to live) these strategies address a notional
community which is defined largely by administrative and political boundaries.
In effect these local ‘communities’ comprise numbers of different communities
of interest who simply happen to share (but may also contest) the same space,
and this may pose particular difficulties in developing and implementing local
community safety strategies. Yet, in fact, many different sections of local popu-
lations share the same concerns, but they may be unaware of what they have in
common and they lack any framework in which they might work together on
problems which affect them all. A study undertaken for the Home Office a
number of years ago provided practical examples of how this could be achieved
in a range of different local contexts. In particular, it showed what makes the
difference between projects which effectively address common interests by tran-
scending perceptions of ethnic difference and those which fail where the
problems facing them become racialised (Jeffers et al. 1996).
Many of the concerns shared by local people regardless of ethnicity relate to
issues of crime and disorder. Looked at in this way, the requirements of the
Crime and Disorder Act (1998) could actually present opportunities for
breaking down suspicion and antagonism based on perceived ethnic difference.
Exploiting these opportunities would not be easy; nor would it eliminate the
dilemmas and challenges identified by this chapter. It might, however, serve to
minimise some of the tensions which have arisen from the perception that
meeting the needs of ethnic minorities is inimical to meeting the needs of the
majority population. As such, it might help offset the damage this perception
may do to any prospect of achieving the ideal implied by the notion of ‘the
community as whole’.
 Ethnic minorities and community safety 165

References
Aye Maung, N. and Mirrlees-Black, C. (1994) Racially Motivated Crime: a British Crime
Survey Analysis, London: Home Office.
Bridges, L. (2000) ‘Limiting the Right to Jury Trial – Half Truths and False Assump-
tions’, Statewatch, 10(1): 17–19.
Brown, D. (1997) PACE Ten Years On: A Review of the Research, London: Home Office.
Crown Prosecution Service (1997) CPS Racist Incident Monitoring Scheme: Report on Year
Ending 31 March 1997, London: Crown Prosecution Service.
Demuth, C. (1978) Sus, a Report on the Vagrancy Act 1824, London: Runnymede Trust.
Euteneuer, R. (1999) Education in London: Key Facts, London: London Research Centre.
FitzGerald, M. (1999a) Searches in London: Interim Evaluation of Year One of Programme
of Action, London: Metropolitan Police. Online. Available: http://www.met.police.uk/
mps/htm
—— (1999b) Searches in London: Final Report, London: Metropolitan Police. Online.
Available: http//www.met.police.uk/mps/htm
—— (2000) London Borough Profiles: Social and Crime Characteristics. London, London:
London School of Economics and South Bank University. Online. Available:
http//www.policingforlondon.org (September 2000).
FitzGerald, M. and Ellis, T. (1989) ‘Racial Harassment: The Evidence’, in C. Kemp (ed.)
Current Issues in Criminological Research, Bristol: Bristol and Bath Centre for Criminal
Justice.
FitzGerald, M. and Hale C. (1996a) Ethnic Minorities: Victimisation and Racial Harass-
ment: Findings From the 1988 and 1992 British Crime Surveys, London: Home Office.
—— (1996b) ‘Ethnic Minorities and Fear of Crime’, unpublished paper presented to
ESRC workshop on Fear of Crime. Manchester, 18 April.
—— (forthcoming) ‘Stops, Searches and the Limitations of Ethnic Data’, Howard
Journal.
FitzGerald, M. and Sibbitt, R. (1997) Ethnic Monitoring in Police Forces: A Beginning,
London: Home Office.
Flatley, J. (1999) Social Care in London: Trends in Social Services’ Activity 1993–97,
London: London Research Centre.
Flood-Page, C. and Mackie, A. (1998) Sentencing Practice: An Examination of Decisions in
Magistrates Courts and the Crown Court in the mid-1990’s, London: Home Office.
Her Majesty’s Inspectorate of Constabulary (2000) Policing in London: ‘Winning Consent’
– a Review of Murder Investigation and Community and Race Relations Issues in the
Metropolitan Police Service, London: Home Office.
Hewitt, R. (1992) Sagaland: Youth Culture, Racism and Education: A Report on Research
Carried out in Thamesmead, London: University of London and London Borough of
Greenwich.
—— (1996) Routes of Racism: The Social Basis of Racist Action, Staffordshire: Trentham
Books.
Home Office (1999a) Criminal Justice System: Strategic Plan 1999–2002 and Business Plan
1999–2000, London: Home Office, Lord Chancellor’s Department and Attorney
General’s Office.
—— (1999b) Statistics on Race and the Criminal Justice System: A Home Office Publication
Under section 95 of the Criminal Justice Act 1991, London: Home Office.
—— (2000) Statistics on Race and the Criminal Justice System, London: Research, Devel-
opment and Statistics Directorate, Home Office.
Hood, R. (1992) Race and Sentencing, Oxford: Oxford University Press.
166 Marian FitzGerald

Jeffers, S., Hoggett, P. and Harrison, L. (1996) ‘Race, Ethnicity and Community in three
Localities’, New Community, 22(1): 111–26.
Macpherson, W. (1999) The Stephen Lawrence Inquiry, London: The Stationery Office.
Mayhew, P. Dowds, L. and Elliott, D. (1989) The 1988 British Crime Survey, London:
Home Office.
Maynard, W. and Reed, T. (1997) Policing Racially Motivated Incidents, London: Home
Office.
Metropolitan Police Performance Information Bureau: all statistics cited provided ad hoc
to the author on request.
NACRO (1997) The Tottenham Experiment, London: National Association for the Care
and Resettlement of Offenders.
Penzer, J. (1999) Reported Crime and PACE Stop and Search Activity: An Investigation of
the Possible Relationship, London: London School of Economics. Online. Available:
http://www.lse.ac.uk/Depts/statistics/ (26 October 1999).
Percy (1998) Ethnicity and Victimisation: Findings From the 1996 British Crime Survey,
London: Home Office.
Runciman, R. (2000) Drugs and the Law: Report of the Independent Inquiry into the Misuse
of Drugs Act 1971, London: The Police Foundation.
Sibbitt, R. (1997) The Perpetrators of Racist Harassment and Racist Violence, London:
Home Office.
Yinger, M. (1986) ‘Intersecting Strands in the Theorisation of Race and Ethnic Rela-
tions’, in J. Rex and D. Mason (eds) Theories of Race and Ethnic Relations, Cambridge:
Cambridge University Press.
Young, J. (1994) Policing the Streets: Stops and Search in North London, London: Centre for
Criminology, Middlesex University.
7 The new correctionalism
Young people, Youth Justice
and New Labour
John Pitts

Youth justice and community safety

An efficient and effective youth justice system was, and remains, a central plank
of New Labour’s community safety strategy. Hence, Section 37 of its Crime and
Disorder Act (1998) states that: ‘It shall be the principal aim of the Youth
Justice system to prevent offending by children and young persons’ (Home
Office 1998a: 2). The Youth Justice Task Force, established by British Prime
Minister Tony Blair immediately after the 1997 General Election, recom-
mended that this goal could best be achieved via the swift administration of
justice; confronting young offenders with the consequences of their offending;
punishment proportionate to the seriousness and persistence of their offending;
the reparation of victims; the enforcement of parental responsibility; and the
provision of relevant help in areas deemed to be associated with offending, such
as drug abuse and literacy.
This vision of an effective and efficient youth justice system stood in marked
contrast with the account proffered by the Audit Commission in its report,
Misspent Youth, published the year before which stated that:

The current system for dealing with youth crime is inefficient and expensive,
while little is being done to deal effectively with juvenile nuisance. The
present arrangements are failing the young people – who are not being
guided away from offending towards constructive activities. They are also
failing victims – those who suffer from young people’s inconsiderate
behaviour, and from vandalism and loss of property from thefts and
burglaries. And they lead to waste in a variety of forms, including lost time,
as public servants process the same young offenders through the courts time
and again, lost rents, as people refuse to live in high crime areas; lost busi-
ness, as people steer clear of troubled areas; and waste of young people’s
potential.
(Audit Commission 1996: 96)

Written by a Treasury economist and a social policy specialist, Misspent Youth

posed both a critique of the inefficiency and ineffectiveness of the existing
168 John Pitts

system and, through its specification of goals, administrative procedures,

criminological theories and ‘evidence-based’ rehabilitative practices a blue-
print for a new one. The report’s emphasis upon the more effective
integration of system elements, adherence to a limited range of shared objec-
tives, the establishment of clear performance ‘targets’ and cost effectiveness,
to be achieved via the adoption of evidence-based ‘rehabilitative’ techniques,
had a strong resonance for New Labour and its Youth Justice Task Force,
imbued as they were with the belief that prudent, popular and accountable
‘good governance’ was ‘the only possible programme for contemporary
governments’ (Mair 2000). The Crime and Disorder Act (1998) was the
legislative instrument by which these objectives were to be achieved.

The administrative structure of the new youth justice

The 1998 Act required chief executives of local authorities to bring into being
Youth Offending Teams (YOTs) staffed by personnel seconded from the police,
the probation service, education, social services, the health service and, in
certain instances, the youth service or other relevant voluntary sector agencies
if this were deemed appropriate. YOTs were required to produce a Youth
Offending Plan specifying how the team would organise and discharge its func-
tions, and how it would liaise with other statutory and voluntary bodies.
To oversee the development of the YOTs, the Act established the Youth
Justice Board of England and Wales (YJB). The YJB has assumed management
responsibility for what is now termed the Secure Estate. The Secure Estate
comprises local authority Secure Units, the Secure Training Centres for 12 to 14
year olds (introduced by the Conservative government in 1993), the two
specialist Youth Treatment Centres operated by the Department of Health and
Young Offender Institutions and Young Offender wings, for youngsters aged 15
to 18, which are operated by the Prison Service. The role of the YJB is to specify
and monitor standards of efficiency and effectiveness and provide advice, guid-
ance and training for local authorities and institutions. The YJB also holds
substantial budgets for which YOTs and institutions can bid in order to develop
innovative working methods, such as mentoring schemes, bail support projects
or outreach work with youngsters ‘at risk’, and specialist staff training. Beyond
its developmental role, however, the YJB has reserve powers to take control of
YOTs and institutions if they are not being run in accordance with the require-
ments of the 1998 Act or the standards promulgated by the YJB.
The government hopes that, in this way, similarly high standards of profes-
sional practice and provision will be achieved throughout England and Wales.
These changes represent a significant move towards a national system of juve-
nile justice in England and Wales. They also represent an unprecedented
centralisation of control over the system. However, if successful, they promise to
inject a greater degree of equity between regions into a system which, as much
previous research has noted, is characterised by significant variations in provi-
 The new correctionalism 169

sion, quality of service and sentencing outcomes (ILYJS 1995; Audit

Commission 1996).

The youth justice provisions of the Crime and Disorder Act

(1998)
At the core of the deliberations of both Misspent Youth and the Youth Justice
Task Force was an assumption that effective youth crime prevention could be
achieved by drawing children, young people and their parents into the purview
of the youth justice system at an early stage in the child’s ‘criminal career’. This
marked a radical departure from the dominant orthodoxy of the 1980s and
early 1990s, characterised by Elliott Currie (1985) as ‘progressive minimalism’,
which held that whenever and wherever possible young offenders should be
diverted away from the justice system in order to minimise stigma and re-inte-
grate them into mainstream services and institutions. New Labour and its
professional advisers rejected progressive minimalism out of hand. As Mark
Perfect, joint author of Misspent Youth and subsequently secretary to the YJB,
observed:

The present system replicates the inconsistent parenting which most young
offenders have received, making it necessary to replace it with a fast, effi-
cient system with a progressive, comprehensible sentencing tariff, which
offers the consistency and predictability which replicate ‘good parenting’.
(Perfect 1998)

The antipathy to ‘progressive minimalism’ found expression in the govern-

ment’s support for a robust interventionism. However, whereas Labour Party
interventionism in the field of youth crime had traditionally focused upon the
amelioration of the social and psychological needs and problems of young
offenders, a ‘tougher’ New Labour government focused upon the deeds of young
offenders, and their impact upon victims, rather than their needs. The youth
justice provisions of the Crime and Disorder Act (1998) had four key elements:
pre-emptive interventions, pre-court preventive interventions, non-custodial
preventive disposals, and incarceration (see Figure 7.1).

Pre-emptive interventions
A remarkable feature of the 1998 Act is that it allows pre-emptive formal inter-
vention with children and young people below the age of criminal responsibility
(10 years), including those who have committed no offence but are deemed
likely to do so, or who are deemed by the police and the local authority to
constitute a nuisance in their neighbourhoods. These measures will, necessarily,
draw more younger children and their parents, via Parenting Orders, into the
purview of the justice system. These pre-emptive interventions are civil
170 John Pitts

Figure 7.1 The Crime and Disorder Act (1998)

measures but, in the cases of the Anti-Social Behaviour Order and the
Parenting Order, breaches may be dealt with by criminal sanctions.

Pre-court ‘preventive’ interventions

The Act places strict limitations upon police discretion by replacing the more
flexible police caution with the Reprimand and the Final Warning. These
measures are to be backed-up by programmes of restorative justice,
cognitive/behavioural change and parental education. The abandonment of
police discretion to administer a series of repeat cautions if circumstances are
seen to merit it, means that youngsters who are unresponsive to the new
measures are likely to enter the Youth Court at a younger age and an earlier
stage in their offending careers.

Non-custodial preventive disposals

The 1998 Act effectively abandoned the Conditional Discharge, a disposal
previously imposed in 28 per cent of cases. Conditional Discharges required
young offenders to desist from future offending as a condition of discharge. The
Youth Justice and Criminal Evidence Act (1999) introduced a Referral Order to
fill the gap left by the Conditional Discharge. When implemented, a young
 The new correctionalism 171

person subject to a Referral Order, and having been sentenced for his or her
offence in the Youth Court, will be referred to a Youth Offender Panel. The
constitution of these panels has yet to be finalised but it is likely that they will
mirror the Scottish Children’s Hearings in which there is both a professional
and a lay presence. These panels will then decide on the type of programme the
youngster should pursue. It is likely that the emphasis will be on victim-offender
reparation and other ‘restorative’ interventions (Haines 2000). Whatever the
merits of the Referral Order, the Danish experience suggests that it is likely to
generate a higher volume of re-appearances in court, as a result of breaches of
the Order, than the Conditional Discharge it replaces (Mehlbye et al. 1998).
Whereas a breach of the Conditional Discharge required the young person to
re-offend, to be detected, apprehended, charged, prosecuted and found guilty, a
breach of the Referral Order requires only that the youngster fail to attend a
session of the programme.
A failure to successfully complete a Referral Order will probably result in the
imposition of an Action Plan Order, a Reparation Order or a Supervision Order
by the Youth Court. The Act specifies that youngsters would not normally
receive more than two community penalties so the imposition of such orders for
a young person who has already completed or breached a Referral Order would,
in effect, constitute a ‘last chance’.
The assumption that the bulk of young offenders will, as a result of participa-
tion in the ‘evidence-based’ programmes of victim-offender mediation, mentoring
and cognitive-behavioural change which underpin these orders, be made to desist
from offending is, as I argue below, rooted in a particularly optimistic reading of
that evidence (Pawson and Tilley 1997; Matthews and Pitts 1998).

Incarceration
In consequence, children and young people who breach an Action Plan,
Reparation or Supervision Order, or re-offend following their completion, are
likely to receive the new, partially indeterminate, Detention and Training
Order (DTO) or have a ‘residence requirement’ inserted in their Supervision
Order. This threatens to increase significantly the numbers of children and
young people in custody. Indeed, in September 2000 the numbers of youngsters
sentenced to custody was up 11.5 per cent on the preceding year. The ‘fast-
tracking’ of persistent young offenders, who are now more likely to make a
separate court appearance for each offence, or offending episode, exacerbates
this tendency. This process will gain further momentum in 2001 with the intro-
duction of the ‘three strikes’ sentencing strategy for young burglars, introduced
in Michael Howard’s Crime Sentences Act (1994). At this point, it is likely
that the numbers of 12 to 14 year olds sentenced to Detention and Training
Orders will increase substantially, quickly filling the 200 available places in the
new Secure Training Centres and placing even greater pressure upon Prison
Department young offender provision. And here lies one of the paradoxes of
New Labour’s new youth justice strategy.
172 John Pitts

The secure estate we’re in

In its final report in 1998, the Youth Justice Task Force concluded that:

The different types of facilities for young people that constitute the secure
estate are in need of major reform. Current arrangements are both inconsis-
tent and unsatisfactory. Young offender institutions are too large. Bullying
and abuse of one young offender by another occurs too often while the
education offered is often poor.
(Youth Justice Task Force 1998: 3)

Inasmuch as the YJB may strive to develop a rational, effective and humane
youth justice system and raise standards of provision and practice within the
secure estate, its efforts are being systematically undermined by the logic of a
system which, by default rather than design, is drawing in larger numbers of
less problematic children at a younger age, and pushing them through the
system more quickly. This systemic tendency is compounded by unequivocal
signals sent to the Youth Courts by the Prime Minister, the Home Secretary
and senior Home Office ministers that they should take a tough line on youth
crime.
This tension between the political and professional spheres is nowhere more
evident than in the case of Secure Training Centres (STCs). These were
proposed by the Conservative administration in 1993, and were resisted by the
then Shadow Home Secretary Tony Blair, who argued, correctly, that the
evidence pointed to the folly of bringing together the most difficult and
disturbed children in the country under one roof (Hoghughi 1983). However,
STCs were subsequently embraced by Shadow Home Secretary Jack Straw in an
attempt to beef-up New Labour’s ‘law and order’ profile in the run-up to the
1997 General Election. Medway STC opened in the spring of 1998. The YJB
required that staff were vetted in accordance with the recommendations of the
Warner report, Choosing with Care (1996). Accountability was ensured by the
recruitment of a resident Home Office monitor to ensure ‘contract compliance’,
and regular inspections by the Social Services Inspectorate. The rights of
inmates were to be vouchsafed by creating easy access to a variety of ‘watchdog’
organisations: Childline, The Samaritans, and the Voice of the Child in Care,
this latter organisation also providing ‘independent visitors’.
Thus, Medway STC represented the ‘cutting-edge’ in ‘institutional trans-
parency’. Yet in the Centre’s first seven months of operation, more than 100
assaults on staff were recorded along with £100,000 worth of damage. Over this
same period, thirty-five of the original 100 staff resigned. In June 1998, the
London newspaper, the Evening Standard, reported that:

More than 30 police wearing riot equipment had to use dogs to quell the
disturbance at the Medway Secure Training Centre. Three staff were
injured and windows smashed as 15 children, many wielding snooker cues
 The new correctionalism 173

and balls, ran riot last night. It was an hour and a half before the incident
was brought under control.
(Evening Standard, 26 June 1998: 1)

This was, more or less, a re-run of the scenes which had accompanied the
launch of the similarly hyped DHSS Youth Treatment Centre in Brentwood,
Essex, in the mid-1970s. Like Medway, Brentwood had also attempted to bring
some of the most difficult and disturbed children in the country together under
one roof.
The Social Services Inspectorate submitted a damning report on Medway
STC in January 1999, recommending that it be shut down and that the other
four STCs under construction, at an estimated cost of £30 million, should be
put on hold. Government minister Paul Boateng, recently transferred from the
Department of Health, where he launched the Quality Protects initiative to
safeguard children in residential care, refused to be drawn on the issue of
closure, referring to the Medway trainees as ‘little monsters’.
Political imperatives displace professional considerations and ethical
concerns while the problems identified by the Youth Justice Task Force, far from
being ameliorated, become more acute. Similar tensions afflict the Youth
Offending Teams.

What is a YOT and what is it not?

As noted, the 1998 Act required chief executives of local authorities to bring
into being Youth Offending Teams (YOTs). The constitution of these teams
echoes of the Multi-Agency Diversion Panels, comprising representatives from
the police, social services, education, the youth service and the voluntary sector,
which developed more or less spontaneously in the 1980s. Misspent Youth
(Audit Commission 1996) had commended the exemplary Northamptonshire
Diversion Scheme as a suggestive model for ‘service delivery’ within the new
system. However, Multi-Agency Diversion Panels were developed in the heyday
of ‘progressive minimalism’ and, as their name suggests, were designed to divert
youngsters out of the justice system. Whereas in 1980, 71,000 boys and girls
aged 14 to 16 were sentenced by the Juvenile Courts in England and Wales, by
1987 this figure had dropped to 37,300, a reduction of over 52 per cent. The
most effective Diversion Panels strove to develop a non-stigmatising shadow
tariff constituted from informal, but robust, contractually-based interventions in
the spheres of education, family relationships, use of leisure, vocational training
and drug abuse. This strategy had a remarkable impact upon police practice.
Between 1980 and 1987 the cautioning rate for girls aged 14 to 16 rose from 58
to 82 per cent. For boys the figures were 34 and 58 per cent respectively. This
rate was inflated in part by the development of ‘Caution Plus’, the practice of
issuing repeat cautions if the young person agreed to comply with requirements
specified by Diversion Panels (Pitts 1988).
For the first time, since the creation of a separate system of juvenile justice in
174 John Pitts

the UK in 1908, the YOTs assemble in one place, on a statutory and fully
funded basis, the ‘minimum sufficient network’ (Skynner 1974) of agencies,
professionals and expertise necessary to make a ‘holistic’ response to the
multiply disadvantaged young people who find themselves at the hard end of
the youth justice system. Moreover, by dint of its multi-agency, multi-disci-
plinary structure, the YOT may also serve as a catalyst for positive change in the
local authorities, the police, probation and voluntary sector services from which
YOT personnel are seconded.
However, responding effectively to the troubled youngsters who find their
way to the YOTs may take a long time and require the development of sensitive
referral protocols with a range of agencies, plus regular follow-up contact. In a
study of 270 young people involved with an inner London YOT (Jones 2001),
180 people (45 per cent) had been placed on the Local Authority Child
Protection Register at least once. Most had spent some time in local authority
care and the majority had a parent with recurrent mental health problems.
But the sheer volume of cases generated as a result of the propensity of the
new system to draw in new populations and speed their progress through it (the
Association for Youth Justice estimates an increase of 150 per cent per year,
based on pre-1998 levels of throughput) means that in many YOTs professional
staff, irrespective of their area of expertise, are simply concerned with the super-
vision of Court Orders.
These Orders are increasingly underpinned by offence-focused programmes
devised, and in some cases quite literally scripted, by the Home Office
Offending Behaviour Programme Unit or the Community Justice National
Training Organisation (Pitts 2000a and b). These programmes aim, quite
explicitly, to eliminate the exercise of judgement or discretion on the part of
the professionals administering them in order to avoid compromising
‘programme integrity’. This means that efforts to develop reflexive, holistic,
responses to youngsters in trouble, in which specialists come together to devise
programmes which will not only challenge young people’s behaviour, but also
address the underlying problems associated with their offending, is jeopardised,
and in some cases abandoned. Moreover, preliminary results from the forth-
coming YJB evaluation of ‘pilot’ YOTs suggest that professionals may be
spending less time in face-to-face contact with young offenders than in 1996,
when Misspent Youth bemoaned the paucity of professional contact with young
offenders.
On 25 September 2000, Jack Straw announced new measures, involving the
use of electronic tagging and voice verification systems, to keep Britain’s ‘2,500
most persistent tearaways’ under twenty-four hour surveillance (Guardian, 25
Sept. 2000: 6). He had earlier announced a cash injection to create several
thousand more probation officers. While at a political level this is an obvious
attempt to ‘steal a march’ on New Labour’s right-wing critics, it may also indi-
cate that the government is concerned that correctionalism has not fulfilled its
early promise, and is now seeking more direct, certain and cost-effective
measures with which to control young offenders.
 The new correctionalism 175

The politics of the new youth justice

As noted at the beginning of this chapter, there is a clear tension between the
professional and political dimensions of New Labour’s youth justice strategy. At
a political level, New Labour’s strategy on youth crime mirrors the ‘toughening’
of youth justice which has been evident in the US, as well as mainland Europe,
in recent years.
New Labour’s strategy was designed to seize the mantle of ‘law and order’ by
synthesising the authoritarian populism of its most powerful opponents and
most vociferous critics while show-casing a new sort of politics in which
economy efficiency and effectiveness were the watchwords. For Tony Blair, this
was a strategy which also held the promise of bridging some of the political divi-
sions in his own Party by promising to be both ‘tough’ enough ‘on crime’ for the
‘modernisers’ and ‘tough’ enough on the ‘causes of crime’ for Old Labour’s
remaining ‘social engineers’. This marked a shift from what Johnstone (2000)
has described as the elitist policy-making, which characterised the development
of the minimalist youth justice policies of the 1980s, to forms of policy-making
‘based on a growing bureaucracy of administrators and experts as well as a media
driven relationship to individual voters’ (Simon 2000).
The Crime and Disorder Act (1998) marked a repudiation of both 1960s
‘welfarism’ (Pitts 1988) and 1980s ‘progressive minimalism’ (Currie 1985) on
the grounds that both approaches represented soft options which failed to
confront the moral dimension of youth crime and disregarded the ‘rights’ of
victims. However, this strategy diverged from developments in the US, in its
unwillingness to adopt explicitly ‘cruel’ measures such as the ‘boot camps’,
‘house of pain’ regimes and ‘chain gangs’ which have re-emerged in several US
youth justice systems in recent times (Simon 2001). ‘Toughness’ for New
Labour may mean the extension of custodial confinement and the community
supervision of a broader range of young offenders and troublesome children, but
its stated purpose is to provide a forum in which ‘evidence-based’ programmes
can be administered in a setting bounded by administrative protections against
violations of children’s rights. This is not to suggest, of course, that, ‘on the
ground’, institutions, regimes and professional practices in the English youth
justice system may not sometimes be negligent, brutal or inimical to the best
interests of young offenders in a variety of ways (Crimmens and Pitts 2000).

Dejuvenilisation
The onus on children’s rights in the new youth justice is, however, in tension
with the attempt to erode the distinctiveness of the juvenile jurisdiction and to
ascribe greater levels of culpability to juveniles. As Jonathan Simon (2001) has
argued, the erosion of the juvenile jurisdiction in the US has been powered by
the belief amongst ‘opinion-formers’ and policy-makers, hostile to both
‘welfarism’ in youth justice and the welfare state in general, that the rehabilita-
tive responses developed in the 1960s and 1970s had failed and that only the
re-assertion of patently retributive responses would restore moral hegemony and
176 John Pitts

reduce crime. In the UK, pressures towards ‘dejuvenilisation’ were intensified in

the wake of the killing in 1993 of 2-year-old James Bulger by two 10-year-old
boys, Jon Venables and Robert Thompson. Since then, there has been a steady
movement away from strategies of informalism and normalisation and the aban-
donment of attempts to divert less serious young offenders from prosecution and
custody. As I have already noted, this has been accompanied by the steady
growth of custodial disposals for 15 to 18 year olds, and the introduction of new
types of secure and custodial penalties and institutions for youngsters aged
between 12 and 14.
Popular anxieties triggered by the Bulger case gave rise to a debate in the
media, the police and, eventually, Parliament, concerning the existence of a
‘new breed’ of under-age criminals whom the law ‘could not touch’ (Hagell and
Newburn 1994; Pitts 2000b). One of the political fruits of this debate was the
Home Secretary Jack Straw’s decision, shortly after the 1997 election, to
abandon the principle of doli incapax, a defining feature of the juvenile jurisdic-
tion in the UK from its inception. The problem with doli incapax, Straw
maintained, was that it pre-supposed that children aged between 10 and 13
were incapable of differentiating between right and wrong. In fact, the principle
of doli incapax means no such thing. Rather, it holds that in its dealings with
children below the age of 14 who have broken the law, the courts must proceed
from the assumption that, by dint of their immaturity, and even though they
may have known that they were doing wrong, the child does not have criminal
intent and does not fully understand the consequences of their actions for
themselves or their victim(s). As a result, the onus is upon the prosecution to
demonstrate criminal intent. In the case of 14 to 17 year olds, by contrast, the
onus is on the defence to demonstrate the absence of criminal intent. This was
a symbolic intervention; an unequivocal indication that considerations of
responsibility, culpability and retribution were henceforth to be a legitimate
concern of the youth justice system and its agents. The other principle over-
turned by the Home Secretary, that in order to minimise the stigma of
involvement in the youth justice system, the identity of defendants should not
be revealed to the press or the public, was supplanted by discretionary powers
for Youth Courts to ‘name and shame’ young offenders if they felt it to be
warranted. Moreover, the Crime and Disorder Act (1998) allowed curfews,
enforced by electronic tagging, to be extended to children and young people of
10 years and over, even though this measure would not be supervised by a
member of the new YOTs. However, the lowering to 12, or to 10 if the Home
Secretary deems it politic, of the age at which youngsters can be sentenced to
custodial confinement in a Secure Training Centre constitutes the radical edge
of the government’s dejuvenilising thrust.
The re-assertion of the responsibilities of young offenders which dejuvenilisa-
tion represents, and the suspicion that welfarism, and by extension the welfare
state per se, might in fact foster youthful misbehaviour finds expression in New
Labour’s conflation of the problem of youth crime with the problem of ‘the family’.
 The new correctionalism 177

Failing families
In the 1980s the ‘family’, and in particular the criminogenic, lower-class,
‘welfare-dependent’ family, became a key focus for neo-liberal politicians and a
newly emergent neo-liberal intelligentsia. Their lament, in essence, was that
the rules of civilised social life were no longer observed, that absolute moral
values and their associated norms and practices had been trashed and aban-
doned by irresponsible ‘intellectuals’ on the one hand, and lower-class
malcontents on the other hand. Moreover, a culture, which had once placed a
premium on duty and responsibility, had now been supplanted by a culture of
complaint, dependency and entitlement. Thus, they rejected the traditional
ethos of the welfare state, which emphasised the rights of poor families, focusing
instead upon the economic burden placed by welfare recipients upon the hard-
working majority. In its dealings with families, they maintained, the welfare
state has been counterproductive because it:

failed to reinforce the work ethic; the goal of self sufficiency, self support
and self-initiative; the importance of intact families; the fiscal responsi-
bility of the parents to the child; and the notion of reciprocity, the idea
that recipients have a social obligation to perform in return for receiving
assistance.
(Karger and Stoesz 1990: 65, cited in Glennester and Midgley 1991)

In the US, Charles Murray (1994) had argued that the US government
should scrap its Aid to Families with Dependent Children programme (AFDC),
Medicaid, Food Stamps and Unemployment Insurance, in order to ‘drive young
women back into marriage’. Imported to do a similar job in the UK by the influ-
ential right-wing political journalist Andrew Neil and the Sunday Times
newspaper, Murray argued that:

the civilising process cannot occur in communities where the two parent
family is not the norm, and this will turn out to be as true in England as
America. The real problem with the ‘alternative’ of unmarried parenthood
is that it offers no ethical alternative for socialising little boys. For males,
the ethical code of the two-parent family is the only game in town.
(Murray 1994: 26)

Journalist Melanie Phillips (1994) was one of the first erstwhile left-leaning
celebrities to embrace what Murray described as the New Victorianism in the
UK, and in an article for the Observer newspaper she gleefully monitored its
colonisation of metropolitan opinion in the 1990s:

Peter Lilley and Sue Slipman, doughty defender of single parents, are
suddenly singing in close harmony. In a speech last week, the Social
Security Secretary blamed low pay and unemployment for turning young
men into unmarriageable prospects. Miss Slipman talks about how the
178 John Pitts

collapse of the male role through lack of work is manufacturing the kind of
yob no self-respecting woman would choose as a spouse. Suddenly a govern-
ment widely assumed to be hostile to single mothers seems to agree that the
real problem is men. … The question being asked is how can we civilise
young men when family structures, employment and moral authority are so
weakened?
(Phillips 1994: 27)

In a somewhat similar vein, but coming this time from the ‘post-Marxist left’,
Beatrix Campbell (1995) argued that unemployment and the ‘collapse of the
male role’ had forced men back into spaces which, by day, have traditionally
been occupied by women. However, she states that far from:

Co-operating in the creation of a democratic domesticity … what they

admired and serviced was the criminalised brotherhood; what they harassed
and hurt was community politics. It was an entirely and explicitly gendered
formation.
(Campbell 1995: 249)

Thus, for Campbell, there is ‘an underclass’ but it is a gendered underclass made
up exclusively of men. What united these critiques was their conflation of
moral and fiscal concerns which found expression in a desire for the state to
control the bad behaviour of young men and their feckless fathers while,
Slipman and Campbell excepted, withdrawing support from their welfare-
dependent mothers.
These attempts to locate the origins of crime in the lower-class family ranged
from the blatantly rhetorical (Phillips 1994), through the pseudo-scientific
(Dennis 1997; Murray 1994), to accounts which inferred an aetiology from
empirically established correlates. Crime and the Family (1993) by Utting et al.,
drew upon data generated by West and Farrington’s (1973) Cambridge study of
‘delinquent development’ to argue that ‘the tangled roots of delinquency lie, to
a considerable extent, inside the family’. This timely publication paralleled, and
gave considerable impetus to, New Labour’s growing interest in the lower-class
family and in crime. Thus, by 1995, New Labour’s emergent ‘family policy’ was
virtually synonymous with its ‘criminal justice policy’. Indeed, the advent of the
Parenting Order in the 1998 Act must be seen against the backdrop of New
Labour’s high-profile commitment to the rejuvenation of the family as an ‘effec-
tive caring and controlling social unit’, and the pre-eminence it ascribed to
what Jack Straw described as family forms characterised by ‘two participating
parents’. As Tony Blair declared in his speech to the 1995 Labour Party confer-
ence: ‘a young country that wants to be a strong country cannot be morally
neutral about the family.’
Clearly, families play an important role in determining whether, or to what
extent, children and young people become involved in a broad range of socially
deviant and illegal behaviours. However, the causal primacy ascribed to the
 The new correctionalism 179

‘criminogenic’ lower-class family in New Labour’s youth justice strategy suffers

from what Elliott Currie (1985) has termed the ‘fallacy of autonomy’. This is
because it denies or ignores the relationship between socio-economic stress,
neighbourhood poverty and the biographies of young offenders (Braithwaite
1979 and 1981; Field 1990; Hope 1994; Wikstrom and Loeber 1997; Sampson et
al. 1997; Pitts and Hope 1998; Hagedorn 1998; Young 1999), the peculiar,
mutually-reinforcing, negative contingencies set in train by socially deviant acts
perpetrated by lower-class children and young people (Hagan 1993; Sampson
and Laub 1993), and the role of state agencies in the construction and amplifi-
cation of their ‘deviant careers’ (Muncie 1999; Goldson 1999 and 2000).
The appeal of this perspective for New Labour is fairly clear. In a time when
politicians are unwilling to countenance robust, ‘demand side’, social and
economic intervention to counter social problems, and are eager to demonstrate
that they are ‘tough on the crime’, an analysis which identifies poor child-
rearing practices and weak parental control as the fundamental problem, and a
strategy which targets families and their capacity to inculcate self-control in
unruly and disruptive children which, moreover, resonates with the moral sensi-
bilities of New Labour’s new constituencies in ‘Middle England’, is a political
godsend.

Professional perfidy
New Labour’s onslaught upon the lower-class family was paralleled by an
assault upon the public professionals responsible for keeping the lower-class
family in order. This critique owed much to the neo-liberal account of the role
of the welfare state in fostering a ‘culture of dependency’ and New Labour’s
neo-Thatcherite suspicion of local government and the political orientation of
members of the ‘social professions’. In post-Thatcherite Britain, anti-profes-
sional ire was characteristically directed against teachers and social workers,
although the police are not always exempt, as Campbell’s account of their
collusive involvement in the 1991 youth riots in England testifies (Campbell
1995).
Teachers were castigated as defenders of entrenched, outmoded and almost
invariably ‘permissive’, professional doctrines to which they apparently clung in
the face of ‘widely accepted contrary scientific evidence’, and ‘informed
opinion’. Social workers, for their part, were admonished for blundering into
normal families on the basis of ‘child protection’ hearsay while failing to protect
and control the children of the poor, even as they colluded with the fecklessness
and criminality of their parents. While the problem of youth crime was
presented as, first and foremost, a product of the moral failure of families, this
failure was, it appeared, nurtured and amplified by the stupidity, cupidity, collu-
sion, incompetence or sheer bloody-mindedness of the professionals who were
supposed to contain and educate them. These themes are articulated succinctly
in The Blair Revolution by Peter Mandelson and Roger Liddle (1996). The
authors write:
180 John Pitts

Schools require a new, much tougher, set of disciplinary sanctions to deal

with unruly and uncooperative pupils – such as compulsory homework on
school premises, weekend and Saturday night detention, and the banning
of favourite leisure pursuits such as football matches.
This greater emphasis on discipline should be matched in the local
community. The police, schools and local authority services must work
together closely to crack down on vandalism and other antisocial
behaviour. Excessive tolerance of low-level sub-criminal behaviour by
unruly young people undermines general respect for the rule of law, ruins
the environment and makes a misery of the lives of many innocent people
– and provides a breeding ground for more serious crime.
(Mandelson and Liddle 1996: 17)

New Labour’s solution to the unpredictability of state professionals has been

threefold. It has privatised state services, inserting detailed conditions for
contract compliance and penalty clauses for deviation. In the new Secure
Treatment Centres, for example, the Home Office monitors, mentioned above,
are on hand twenty-four hours a day to ensure that staff are managing
proscribed behaviours among the children in the stipulated manner. It has
supported the de-professionalisation of the professional task by allowing recruit-
ment of far more easily managed, unqualified workers to undertake face-to-face
work with offenders in both youth justice and the Probation Service. It has also
taken to video monitoring correctional programme ‘delivery’, most notably in
Prison Department establishments and the Probation Service. These videotapes
are then sent to monitors in the Home Office who determine whether the level
of compliance with the scripted interventions means that programme providers
may be accredited to run these programmes. At present there is discussion in
the Home Office concerning whether lectures and seminars on the new univer-
sity-based Community Justice degrees, which now train Probation Officers,
should be similarly monitored and accredited. In these ways, the government
aims to ensure that its policies go ‘all the way down’ and are not subverted by
the exercise of professional judgement or discretion.

The metamorphosis of rehabilitation

The idea that youth justice systems might rehabilitate children and young
people fell from favour in the 1970s in the face of research which appeared to
show that official intervention with young offenders tended to worsen the prob-
lems they were designed to solve (Martinson 1974; Preston 1980; Pitts 1992).
However, the ‘progressive minimalism’ to which these findings gave impetus fell
from favour in the early 1990s in the face of a protracted economic recession,
rising crime rates and the murder of James Bulger, which pulled youth crime
back to the centre of the political stage.
As more young people were drawn into the system, the ‘nothing works’
orthodoxy of the 1970s and 1980s gave way to a new rhetoric of ‘what works’
 The new correctionalism 181

(Pitts 1992). This ‘rehabilitation of rehabilitation’ was palpably political in

intent. As Christopher Nutall, described by the Independent newspaper as the
‘hard-headed’ director of research and statistics at the Home Office, told a
meeting of chief probation officers: ‘ “Nothing Works’ should be killed; not just
because it’s not right but because it has had a terrible effect. Let’s not talk about
it any more. Let’s talk about what does work. (Independent, 27 Oct. 1992).
The ‘what works’ movement emanating from the Home Office set about
exhuming aetiological theories and rehabilitative techniques, some of which
had lain abandoned since the 1950s (Pitts 1992). Whereas the rehabilitative
techniques of the 1960s had aimed to ameliorate emotional and social depriva-
tion, these newly disinterred rehabilitations aimed to restructure the modes of
thought, the values, the attitudes and the behaviour of young offenders, the
control strategies of their parents, and the classroom regimes presided over by
their teachers. While presenting themselves as new forms of rehabilitation,
these techniques were straightforwardly ‘correctional’ in both intent and
content.

The promise of correction

The ‘what works’ correctionalist phenomenon was important to New Labour’s
project in two main ways. First, after two decades in which criminological
theory had appeared to be incapable of making any coherent contribution to
criminal justice policy, the policy implications of the new correctionalism were
self-evident and straightforward (Vanstone 2000). This was of particular impor-
tance to New Labour which, through the YJB, was endeavouring to control its
policies ‘all the way down’ to the point of implementation.
It is an axiom of the new correctionalism that, if ‘properly applied’,
evidence-based techniques will arrest the development of ‘offending careers’,
thereby averting subsequent custodial sentences and maximising the economy,
efficiency and effectiveness of the youth justice system. Much has therefore
been made of the fact that these correctional measures – cognitive skills
training (also known as reasoning and rehabilitation), restorative justice and
mentoring – are ‘those which research points to as the most thorough and best
designed programmes showing the most “promising” results’ (Farrington 1996).
Indeed, this phrase has become a kind of official mantra amongst Home Office
ministers, within the YJB, amongst senior personnel in the welfare, justice and
community safety industries and the denizens of some university senior common
rooms. However, their apparent confidence in these measures appears to be
rooted in a partial reading of the ‘evidence’ and an over-estimation of their
‘promise’.
When the word ‘promising’ is used to describe such correctional ventures, it
has a very specific meaning; one bestowed upon it by the National Institute of
Justice (NIJ), a division of the US Department of Justice, in elaborating
national criteria for evaluating ‘correctional programs’. By promising, the NIJ
means:
182 John Pitts

programmes for which the level of certainty from available evidence is too
low to support generalisable conclusions, but for which there is some empir-
ical basis for predicting that further research could support such
conclusions.
(Sherman et al. 1998: 10)

However, should further research support such conclusions, the NIJ cautions,
these findings would only be valid if applied in ‘similar settings in other places
and times’. This raises important questions concerning whether, and to what
extent, the youth justice system of England and Wales might represent such an
analogous social context.

Cognitive skills training

Among the promising programmes cited by the NIJ are cognitive skills training
programmes. Those currently operating in the UK have their origins in the
Reasoning and Rehabilitation programmes developed by cognitive psychologist
Robert Ross and his associates, based on research done in Canadian jails. In the
1990s, the apparent success of these programmes led to their widespread adop-
tion by prison and probation services in England and Wales and latterly by
YOTs. The programmes ‘attempt to teach delinquents the cognitive (thinking
skills) in which they are deficient’, in the expectation that this would lead to a
decrease in their offending (Farrington 1996: 22).

The research evidence

The most frequently cited example of the effectiveness of cognitive skills
training is the correctional programme designed by Robert Ross and undertaken
in a Canadian jail in the mid-1990s. This programme was evaluated by Ross’
colleagues Porporino and Robinson (1995). The programme was, as Pawson and
Tilley (1997) observe, ‘probably the most ambitious and comprehensive rehabil-
itation programme ever attempted by this or any other prison service. By the
end of March 1993, 2,500 offenders had been assessed as suitable cases for treat-
ment and 1,400 had completed the programme.’ Porporino and Robinson’s
evaluation reveals that prisoners who volunteered for the programme had a far
lower reconviction rate, in the twelve months following release, than prisoners
who did not volunteer for the programme – around 20 per cent for the volun-
teers against 50 per cent and greater for non-volunteers. However, in addition
to the group who volunteered and completed the programme and those who did
not volunteer for the programme, there was a third group who volunteered for
the programme, were randomly allocated to a waiting list, and never actually
undertook the programme. Whereas 19.7 per cent of programme ‘completers’
were reconvicted in the twelve months following release, 23 per cent of the
control group were reconvicted. When other contingencies and inconsistencies
 The new correctionalism 183

are taken into account the measurable impact of the cognitive skills training
programme evaporates (Pawson and Tilley 1997; Matthews and Pitts 1998).
Meta-analyses of evaluations of cognitive behavioural programmes suggest
that, for the programmes to be effective, certain key elements must be present.
In the case of high-risk offenders, programmes should occupy between 40 and
70 per cent of the offenders time (Gendreau 1996) and should last for at least
twenty-three weeks (Lipsey and Wilson 1998). In addition to matching the
duration and intensity of programmes to the ‘risk’, ‘criminogenic needs’ and
‘learning styles’ of offenders, programme providers must attempt to match the
programme to the offender’s personality characteristics. ‘High anxiety’
offenders, for example, do not respond well to confrontation, ‘egocentric’
offenders show a poor appreciation of the victim’s perspective, while offenders
with both ‘below average’ and ‘above average’ intellectual ability tend to be,
respectively, perplexed or under-stimulated by cognitive skills programmes
(Fabiano et al. 1991). Meta-analyses also show that programmes are successful
only if providers relate to juveniles in ‘interpersonally sensitive’ and ‘construc-
tive’ ways and are trained and supervised appropriately. Moreover, in the case of
more serious young offenders, ‘relapse prevention programmes’ need to be
employed in the community to monitor and anticipate problematic situations
and train offenders to rehearse alternative behaviours (Vanstone 2000).
Alongside this, programme staff need to be active in linking young people into
other services which meet their needs. However, even when all these criteria
are achieved the average recidivism rate remains relatively high at around 50
per cent (Andrews et al. 1990). Similarly inconclusive results are reported by
Maurice Vanstone (2000), although he reports positive, if not statistically
significant, rehabilitative outcomes in some UK programmes for motivated
adult offenders.
The community penalties administered by the YOTs are unlikely to offer
either the levels of intensity or the consistency of application necessary to effect
change with the persistent or serious young offenders upon whom they will be
imposed. Meanwhile, if these interventions mix low- and high-risk offenders in,
for example, a drugs awareness or literacy element of the programme, they may
trigger an ‘interaction effect’ (Clear and Hardeyman 1990) in which offending
amongst low-risk groups, prematurely drawn into the system by the new
measures introduced by the 1998 Act, may escalate.

Theoretical problems
Beyond the frailty of the evidence about the impact of cognitive skills training
are other concerns about the flimsy theoretical rationale upon which cognitive
skills training is built. At its core is an assumption that offenders can be
taught new cognitive skills which will enable them to anticipate the conse-
quences of their behaviour for themselves and others, and so make
‘appropriate’ moral choices. It is assumed that if they can think ‘straight’ they
will go ‘straight’. In so reasoning, the purveyors of cognitive skills training, like
184 John Pitts

the eighteenth-century Classicists, conflate rational/logical thinking with

moral/law-abiding thinking and, in so doing, commit a categorical error since
they fail to distinguish between the cognitive ‘machinery’ which enables
logical thought and the ethical choices made possible by a capacity for logical
thought. The assumption that people engage in crime because they lack a
capacity for logical thought is, at least, tendentious since in reality people
often resort to crime, violence or deception because, in the circumstances in
which they find themselves, it ‘works’ for them (Sutherland 1934). Thus their
involvement in crime, while morally reprehensible, is nonetheless rational.

Restorative justice
The fundamental precepts of restorative justice hold that when an offence is
committed, both the immediate victims and the communities to which they
belong have been harmed and that the relationship which has been fractured by
that offence is in need of restoration. In this perspective, the primary victims
are those most directly affected by the offence, but others, such as family
members of victims and offenders, witnesses and other members of the affected
community, are regarded as having been victimised by the offence. The restora-
tive justice process encourages victims to participate in defining the obligations
of offenders, who are then offered opportunities to understand the harm they
have caused to victims and the community and to devise a way of making resti-
tution. In this process, obligations to victims take priority over sanctions or
obligations to the state. In its original form, restorative justice emerged as a
local, colloquial and informal alternative to formal systems of retributive justice
administered by state agencies (Christie 1977; Braithwaite and Daly 1994).
Latterly in most Western youth justice systems, as in the UK Crime and
Disorder Act (1998), restorative elements have been incorporated into an
avowedly retributive youth justice system.

The research evidence

The Northamptonshire Youth Diversion Scheme (the Northants Scheme) in
the UK, appears to have had a remarkable degree of success in terms of both
victim/offender satisfaction and reconviction. Victims and offenders reach
agreement about appropriate remedies in over 60 per cent of cases, and both
groups report high levels of satisfaction in terms of the restorative process and
its outcomes. These findings echo those of Elejabarrieta et al. (1993) who found
that, of the 40.2 per cent of cases dealt with in this way in the Catalonian
youth justice system in 1993, 68.5 per cent resulted in agreement between
victim and offender about an appropriate remedy. Of these cases, 78.2 per cent
were rated successful by both victims and offenders although, as was also the
case in Northants, there was a tendency for victim satisfaction to fade with
time. In the bulk of these cases the remedy involved ‘dialogue’ between the
victim and the offender. It proved hardest to reach agreement in cases with
 The new correctionalism 185

initial ‘not guilty’ pleas and those involving peer violence. It proved easiest to
reach agreement in cases of 12 and 13 year olds involved in ‘criminal damage’.
It is important to note that when success rates for restorative justice
programmes are cited, they usually refer to victim-offender satisfaction rather
than offender reconviction.
Nonetheless, the Northants Scheme boasts of a 23 per cent reconviction rate
over a two-year follow-up period for the youngsters passing through its
programmes. However, although this is a cause for celebration, it is important to
note that the Northants Scheme is voluntary, operating outside and as an alter-
native to formal involvement in the youth justice system. Moreover, because
the scheme has only initial administrative contact with the most serious and
persistent offenders and is open to any child or young person who is appre-
hended for an offence, many of whom are low risk and unlikely to reoffend, it is
difficult to identify whether, or to what extent, the scheme, or natural ‘evapora-
tion’, accounts for its low reconviction rate.
The Northants Scheme offers a comprehensive range of rehabilitative,
educational, vocational and ‘use of leisure’ programmes alongside the restorative
elements; but hereby hangs a key methodological problem. Because restorative
justice is just one component of a more complex and extensive programme of
intervention, it is difficult, if not impossible, to isolate its effects.
In their meta-analysis of ten US ‘restitutive’ intervention programmes,
Lipsey and Wilson (1998) found little effect on the recidivism rates of ‘non-
institutionalised’ but persistent juvenile offenders. Van Hooris (1985) found
that restorative justice routinely failed to change the attitudes of ‘low-maturity
persistent offenders’. A study undertaken by Gendreau (1996), found even less
promising results in terms of recidivism, but divined an ‘interaction effect’
which resulted in an escalation of offending rates amongst low-risk offenders.
This was largely because referral into the programme put them at heightened
risk of non-compliance and, as a result, vulnerable to the penal sanctions
which could then be imposed upon them. On the other hand, these
programmes were deemed to be insufficiently intensive to meet the needs of
higher-risk offenders.

Theoretical problems
At the level of theory, there is an important question about the mechanisms of
change which are triggered by restorative interventions. We do not know
whether, or to what extent, young offenders can generalise their identification
with the particular victim they have harmed to an empathy with other poten-
tial victims. The Kirkholt study (Forrester et al. 1990) suggests that ‘burglars’
divide the world into legitimate and illegitimate targets, raising the possibility,
suggested by reconviction studies, that even restorative justice programmes
rated ‘successful’ by both victims and offenders may serve only to delegitimise a
specific victim, rather than all potential victims, in the minds of young
offenders.
186 John Pitts

Mentoring
New Labour has wholeheartedly embraced the idea of mentoring. A central
aim of its Quality Protects, Leaving Residential Care and Connexions
programmes, as well as its youth justice strategy, requires that each young
person should have a mentor. At present, the YJB funds approximately fifty
mentoring schemes in YOTs.
The term ‘mentor’ denotes a loyal, wise and trusted teacher, guide and friend.
In recent years in the US, mentoring, in the guise of ‘buddying’, ‘big broth-
ering’, ‘positive peer influence programs’ or ‘companionship therapy’, has been
utilised extensively with young people experiencing problems of academic
attainment, personal conduct and crime and violence. Mentoring was embraced
by African-American activists in the 1980s as a means of transcending the
barriers to social mobility occasioned by the profound socio-economic disadvan-
tages experienced by about one-third of black Americans, and the
‘institutionalised racism’ encountered by them all. This approach to mentoring
proceeds from the assumption that the mentor and the mentee share similar
aspirations, values and goals. What distinguishes the mentee is his or her lack of
access to opportunities for the realisation of these aspirations, values and goals.
Thus, mentoring becomes a tool for emancipation, a means whereby the
mentee is helped to navigate hostile social terrain.
However, as mentoring has spread, it has been adopted by a broad range of
educational, employment, social welfare and youth justice organisations with
different goals and different priorities. In the process the focus of mentoring has
shifted from advice, guidance and advocacy to a concern to ‘correct’ those
shortcomings in the mentee which prevent him or her from entering the social,
economic or moral mainstream. While ‘emancipatory’ mentoring is borne of a
critique of an inequitable social order which keeps lower-class young people
down, and out, ‘correctional mentoring’ finds the origins of such ‘social exclu-
sion’ within the mentee.
Government minister Margaret Hodge, for example, writing in The Times, notes
that: ‘Mentors can help to break the vicious circle. They are invaluable in
helping boys who lack a father figure at home. It could be simply by making sure
that they get up in time for school (The Times, 10 April 1998). In most
mentoring schemes, ‘role modelling’ is identified as the key medium of change.
The usual rationales for role modelling are redolent of the traditional
approaches to the teaching of history which assert that the young need heroes
with whom to identify and emulate. However, what a ‘role model’ actually is,
and the mechanisms whereby its positive or negative attributes are transmitted
to, and introjected by, the mentee is the subject of continuing controversy.

The research evidence

Unsurprisingly, perhaps, the evaluations of mentoring cited by the Youth Justice
Board are remarkably upbeat, deriving as they do mainly from studies under-
taken by members of the US-based National Mentoring Association. The
 The new correctionalism 187

Association cites a number of studies which suggest positive outcomes for

mentoring. A comparison of 500 children in the Big Brothers and Big Sisters of
America scheme and a control group of non-participants, revealed that partici-
pants were 46 per cent less likely to be using illegal drugs, 27 per cent less likely
to begin using alcohol, and 52 per cent less likely to truant. The programme
also claimed to reduce violence by one-third. The Quantum Opportunities
Program found that its mentees were more likely to graduate from high school,
have fewer arrests, become involved in community service and, interestingly,
have fewer children (Pitts 2000a). The benefits of mentoring cited most
frequently in such evaluations are reduced offending, reduced school exclusion,
improved school performance, the achievement of paid employment, enhanced
self-esteem and confidence and improved relationships with peers and adults.

Methodological problems
However, these claims have tended to be most frequent in evaluations under-
taken by the agencies which provide mentoring programmes and they must
therefore be regarded with a degree of scepticism. Such evaluations may tend,
for understandable reasons, to ‘accentuate the positive’ and overlook, or under-
play, areas where no change occurs or where problems actually worsen. As with
certain evaluations of cognitive skills training programmes, some mentoring
schemes select their ‘control groups’ from young people who decline the offer to
join the scheme. The danger here is that they may then be comparing a group
of youngsters who are motivated to change with a control group which is not.
Other American research, on the effectiveness of mentoring with young
offenders, is far less encouraging. An analysis of the ten ‘best-evaluated’
mentoring programmes, carried out for the US Department of Justice in the
early 1990s (Brewer et al. 1995), suggests that ‘non-contingent supportive
mentoring relationships’ (i.e., programmes which did not utilise a cognitive-
skills training approach involving the systematic manipulation of rewards and
sanctions – but see above for a critique of cognitive-skills training), failed to
achieve their goals in terms of academic attainment, ‘dropping out’, behavioural
change or employment. This analysis also points to ‘interaction effects’, citing
evaluations which revealed significantly increased delinquency for young people
with no previous offences but significantly decreased recidivism for youngsters
with prior offences (Goodman 1972; Fo and O’Donnell 1974).
Hellen Colley (2000) points to a key problem at the heart of mentoring
programmes developed in furtherance of particular government policies:

It [the study] revealed significant gaps between the goals and assumptions
of the scheme, and the desires and needs of the mentees. Many mentees
and mentors felt unduly constrained by the tight focus. … Some mentees
developed resistance strategies to avoid scheme requirements, whilst others
were dismissed from the scheme. Some mentors accepted the young
person’s agenda whilst others tried to impose the official view. Either way
188 John Pitts

subversion resulted. Consequently, some mentors became highly critical of

the young people they were trying to help, reinforcing negative stereotypes
of them as feckless and inadequate. A combination of these factors risked
making things worse for some of the very people the scheme was supposed
to help.
(Colley 2000: 11)

Inadmissible evidence
Although the Home Office and the Youth Justice Board make great play of the
fact that the correctional techniques which underpin the community-based and
custodial penalties in the youth justice provision of the 1998 Act are ‘evidence-
based’, the evidence is, at least, equivocal. Clearly it is much harder to specify,
institutionalise and regulate a restricted repertoire of professional practices, as
the Home Office, the Youth Justice Board and the agencies and organisations
associated with them aim to do, if the evidence upon which they are based is
shown to be ambiguous. As a result, these powerful ‘claim-making agencies’
tend to cite only the evidence which indicates that the theories and practices
they have championed ‘work’. Yet, as Pawson and Tilley (1997) argue, whether
or not an intervention ‘works’, in the sense that it promotes the desired
behavioural change in a young offender, may have little to do with the correc-
tional technique applied to them. Like Colley (2000), they challenge the
assumption at the heart of the new correctionalism, that programmes somehow
work ‘on’ individuals; arguing instead that programmes can only work ‘through’
individuals. They write:

Social programs involve a continual round of interactions and opportuni-

ties and decisions. Regardless of whether they are born of inspiration or
ignorance, the subject’s choice at each of these junctures will frame the
extent and nature of change. What we are describing here is not just the
moment when the subject signs up to enter a program but the entire
learning process. The act of volunteering merely marks a moment in a
whole evolving pattern of choice. Potential subjects will consider a
program (or not), co-operate closely (or not), stay the course (or not), learn
lessons (or not), apply the lessons (or not). Each one of these decisions will
be internally complex and take its meaning according to the chooser’s
circumstances.
(Pawson and Tilley 1997: 38)

Conclusion
The judicial and administrative arrangements brought into being by the Crime
and Disorder Act (1998) threaten to draw greater numbers of younger, less
problematic, children and young people into the youth justice system, to subject
them to modes of correctional intervention which have not been shown to be
 The new correctionalism 189

particularly effective and, in consequence, to propel many more of them into

the custodial institutions from which, Sir David Ramsbotham the Chief
Inspector of Prisons believes, anybody under 18 should be immediately removed
(Home Office 1998b).
However, the effectiveness, or otherwise, of these correctional techniques or
institutional regimes does not appear to be the main criterion by which New
Labour’s youth justice system will stand or fall. While it is of course true that
these techniques have, at some times, in some places and with some people,
‘worked’, their incorporation into policy is, as I have argued above, largely
determined by pre-scientific, ideological and managerial imperatives. And this
is because the primary target of New Labour’s youth justice strategy is not the
criminal behaviour of a handful of young offenders, but the voting habits of a
far larger and much older constituency. New Labour has placed youth crime at
the core of its social and criminal justice policies, alongside, and inextricably
linked with, its assault upon ‘welfare mothers’, ‘failing families’, laggard educa-
tion authorities, self-serving public professionals and ‘neighbours from hell’. In
the process, the claims of justice for juveniles have been subordinated to the
political allure of correctionalism. As a result, it seems unlikely that the new
youth justice will contribute significantly to community safety.

References
Andrews, D., Zinger, R., Hoge, J., Bonta, P., Gendreau, P. and Cullen, F. (1990) ‘Does
Correctional Treatment Work? Clinically Relevant and Psychologically Informed
Meta-Analysis’, Criminology, 28: 369–404.
Audit Commission (1996) Misspent Youth: Young People and Crime, London: The Audit
Commission.
Braithwaite, J. (1979) Inequality, Crime and Social Policy, London: Routledge and Kegan
Paul.
—— (1981) ‘The Myth of Social Class and Criminality Reconsidered’, American Socio-
logical Review, 46: 36–57.
—— (1989) Reintegrative Shaming, Cambridge: Cambridge University Press.
—— and Daly, K. (1994) ‘Masculinities, Violence, and Communitarian Control’, in T.
Stanko and T. Newburn (eds) Just Boys Doing Business, London: Routledge.
Brewer, D., Hawkins, J., Catalano, B. and Neckerman, H. (1995) ‘Preventing Serious,
Violent and Chronic Juvenile Offending: A Review of Evaluations of Selected Strate-
gies in Childhood, Adolescence and the Community’, in J. Howell, B. Krisburg, J.
Hawkins and J. Wilson, Serious, Violent and Chronic Juvenile Offenders, London: Sage
Publications.
Campbell, B. (1995) Goliath, Britain’s Dangerous Places, London: Methuen.
Christie, N. (1977) ‘Conflicts as Property’, British Journal of Criminology, 17: 1–15.
Clear, H. and Hardeyman, C. (1990) ‘The New Intensive Supervision Movement’,
Crime and Delinquency, 36: 42–60.
Colley, H. (2000) What Do We Think We’re Doing? How Mentors’ Beliefs About Disaffec-
tion Influence the Process of Mentoring Disaffected Young People, Paper presented to
BERA Post-16 Special Interest Group, Milton Keynes.
Crimmens, D. and Pitts, J. (2000) Positive Residential Practice, Learning the Lessons of the
1990s, Lyme Regis: Russell House Publishing.
190 John Pitts

Currie, E. (1985) Confronting Crime, New York: Pantheon Books.

Dennis, N. (1997) Zero Tolerance: Policing in a Free Society, London: Institute for
Economic Affairs.
Department of Health and Social Security (1969) Children and Young Persons Act,
London: HMSO.
Elejabarrieta, F. et al. (1993) Els Programes de Mediacio: Que Pensen i Com Els Viuen Les
Parts Implicades, Barcelona: Universitata Autonoma de Barcelona.
Fabiano, E., Porporino, F. and Robinson, D. (1991) ‘Canada’s Cognitive Skills Program
Corrects Faulty Thinking’, Corrections Today, 53 (August): 102–8.
Farrington, D. (1996) Understanding and Preventing Youth Crime, York: Joseph Rowntree
Foundation.
Field, S. (1990) Trends in Crime and Their Interpretation, Home Office Research Study
No. 119, London: HMSO.
Fo, W. and O’Donnell, C. (1974) ‘The Buddy System: Relationship and Contingency
Conditioning in a Community Intervention Program for Youth with Non-Professionals
as Behaviour Change Agents’, Journal of Consulting and Clinical Psychology, 42: 163–96.
Forrester, D., Frenz, S., O’Connell, M. and Pease, K. (1990) The Kirkholt Burglary Preven-
tion Project: Phase II, London: Home Office.
Gendreau, P. (1996) ‘The Principles of Effective Intervention with Offenders’, in A.
Harland (ed.) Choosing Correctional Options That Work, Thousand Oaks: Sage Publi-
cations.
Glennester, H. and Midgley, J. (1991) The Radical Right and the Welfare State: An Interna-
tional Assessment, London: Harvester Wheatsheaf.
Goldson, B. (1999) ‘Youth (In)justice: Contemporary Developments in Policy and Prac-
tice’, in B. Goldson (ed.) Youth Justice: Contemporary Policy and Practice, London:
Ashgate.
—— (2000) (ed.) The New Youth Justice, Lyme Regis: Russell House Publishing.
Goodman, T. (1972) Companionship Therapy: Studies in Structured Intimacy, New York:
Jossey Bass.
Graham, J. and Bowling, B. (1995) Young People and Crime, London: Home Office.
Hagan, J. (1993) ‘The Social Embeddedness of Crime and Unemployment’, Criminology,
31: 455–91.
Hagedorn, J. (1998) People and Folks: Gangs, Crime and the Underclass in a Rustbelt City,
Chicago: Lakeview Press.
Hagell, J. and Newburn, T. (1994) Persistent Young Offenders, London: Policy Studies
Institute.
Haines, K. (2000) ‘Referral Orders and Youth Offender Panels: Restorative Approaches
and the New Youth Justice’, in B. Goldson (ed.) The New Youth Justice, Lyme Regis:
Russell House Publishing.
Hoghughi, M. (1983) The Delinquent, London: Burnett Books.
Home Office (1982) Criminal Justice Act, London: Home Office.
—— (1997a) Crime (Sentences) Act, London: Home Office.
—— (1997b) No More Excuses, London: Home Office.
—— (1998a) Crime and Disorder Act, London: Home Office.
—— (1998b) The Chief Inspector of Prisons Report on Werrington Young Offenders Institu-
tion, London: Home Office.
Hope, T. (1994) Communities Crime and Inequality in England and Wales, Paper presented
to the Cropwood Round Table Conference Preventing Crime and Disorder,
Cambridge, 14–15 September.
 The new correctionalism 191

ILYJS (1995) Statement of Principles and Practice, London: National Association for the
Care and Resettlement of Offenders.
Johnstone, G. (2000) ‘Penal Policy Making: Elitist, Populist or Participatory’, Punishment
and Society, 2(2): 161–80.
Jones, K. (2001) ‘An Analysis of the Deviant Careers of Young People Placed Upon the
Child Protection Register of One London Borough’, unpublished M.Phil. thesis,
University of Luton.
Karger, H. J. (1991) ‘The Radical Right and Welfare Reform in the United States’, in H.
Glennester and J. Midgley (eds) The Radical Right and the Welfare State: An Interna-
tional Assessment, London: Harvester Wheatsheaf.
Lea, J. and Young, J. (1988) What is to Be Done About Law and Order?, Harmondsworth:
Penguin.
Lipsey, M. and Wilson, D. (1998) ‘Effective Intervention for Serious Juvenile Offenders:
A Synthesis of Research’, in R. Loeber and D. Farrington (eds) Serious and Violent
Juvenile Offenders: Risk Factors and Successful Interventions, Thousand Oaks: Sage
Publications.
Loveday, B. and Marlow, A. (2000) Policing After MacPherson, Lyme Regis: Russell House
Publishing.
Mair, P. (2000) ‘Partyless Democracy: Solving the Paradox of New Labour’, New Left
Review, March/April: 21–36.
Mandelson, P. and Liddle, R. (1996) The Blair Revolution, London: Faber.
Marquand, D. (1999) ‘Progressive or Populist? The Blair Paradox’, in D. Marquand (ed.)
The Progressive Dilemma: From Lloyd George to Blair, 2nd edition, London: Verso,
225–46.
Martinson, R. (1974) ‘What Works? – Questions About Prison Reform’, The Public
Interest, Spring: 22–54.
Matthews, R. and Pitts, J. (1998) ‘Rehabilitation, Recidivism and Realism: Evaluating
Violence Reduction Programs in Prison’, The Prison Journal, 78(4): 390–405.
Mehlbye, J., Sommer, B. and Walgrave, L. (1998) Confronting Youth in Europe: Juvenile
Crime and Juvenile Justice, Denmark: AKF.
Mika, H. and Zehr, H. (1997) Fundamental Concepts of Restorative Justice, Harrisonburg,
VA: Mennonite Central Committee USA.
Muncie, J. (1999) Youth and Crime, London: Sage Publications.
Murray, C. (1994) Underclass: The Crisis Deepens, London: Institute of Economic
Affairs.
NACRO Young Offenders Committee (1997) 3Rs: Responsibility, Restoration and Reinte-
gration, London: NACRO.
Newburn, T. (1995) Crime and Criminal Justice Policy, London: Longman.
Pawson, R. and Tilley, N. (1997) Realistic Evaluation, London: Sage Publications.
Perfect, M. (1998) unpublished speech on the 1998 Criminal Justice Act delivered to
the Crime and Disorder Conference, Anglia Polytechnic University, New Hall,
Cambridge, September.
Phillips, M. (1994) ‘Is the Male Redundant Now?’, Observer, 26 May: 27.
Pitts, J. (1988) The Politics of Juvenile Crime, London: Sage Publications.
—— (1992) ‘The End of an Era’, Howard Journal, 31(2): 133–49.
—— (1997) ‘Youth Crime, Social Change and Crime Control in Britain and France in
the 1980s and 1990s’, in H. Jones (ed.) Towards a Classless Society, London: Routledge.
—— (2000a) Youth Justice, Research Matters, Community Care/Reed Business Informa-
tion, April–October: 56–60.
192 John Pitts

—— (2000b) ‘The New Youth Justice and the Politics of Electoral Anxiety’, in B.
Goldson (ed.) The New Youth Justice, Lyme Regis: Russell House Publishing.
—— and Hope, T. (1998) ‘The Local Politics of Inclusion: The State and Community
Safety’, Social Policy and Administration, 31(5): 37–58.
Porporino, F. and Robinson, D. (1995) ‘An Evaluation of the Reasoning and Rehabilita-
tion Program with Canadian Federal Offenders’, in R. Ross and R. Ross (eds)
Thinking Straight, Ottawa: Air Training Publications.
Preston, R. H. (1980) ‘Social Theology and Penal Theory and Practice: The Collapse of
the Rehabilitative Ideal and the Search for an Alternative’, in A. E. Bottoms and R.
H. Preston (eds) The Coming Penal Crisis, Edinburgh: Scottish Academic Press.
Sampson, R. J. and Laub, J. (1993) Crime in the Making: Pathways and Turning Points,
Cambridge, Mass.: Harvard University Press.
——, Raudenbush, S. W. and Earls, F. (1997) ‘Neighbourhoods and Violent Crime: A
Multi-Level Study of Collective Efficacy’, Science 277 (15 August): (page nos. not
known).
Sherman, L., Gottfredson, D., MacKenzie, J., Reuter, P. and Bushway, S. (1998)
Preventing Crime: What Works, What Doesn’t, What’s Promising, Washington, DC: US
Department of Justice.
Simon, J. (2000) ‘From the Big House to the Warehouse: Rethinking Prisons and State
Government in the 20th century’, Punishment and Society, 2(2): 213–34.
—— (2001) ‘Entitlement to Cruelty: Neo-liberalism and the Punitive Mentality in the
United States’, in K. Stenson and R. Sullivan (eds) Crime, Risk and Justice, Devon:
Willan.
Skynner, R. (1974) ‘The Minimum Sufficient Network’, Social Work Today, 14 July.
Sutherland, E. (1934) Principles of Criminology, Chicago: J. B. Lippincott and Co.
Thorpe, D., Smith, D., Green, C. and Paley, J. (1980) Out of Care, London: Allen &
Unwin.
Utting, W., Bright, J. and Hendrickson, B. (1993) Crime and the Family, Improving Child
Rearing and Preventing Delinquency, Family Policy Studies Centre, Paper no.16,
London: Family Policy Studies Centre.
Van Hooris, P. (1985) ‘Restitution Outcome and Probationer’s Assessments of Restitu-
tion: The Effects of Moral Development’, Criminal Justice and Behaviour, 12: 259–87.
Vanstone, M. (2000) ‘Cognitive-Behavioural Work with Offenders in the UK: A History
of Influential Endeavour’, The Howard Journal, 39(2): 171–83.
Warner Report (1996) Choosing with Care, London: Department of Health.
West, D. and Farrington, D. (1973) Who Becomes Delinquent?, London: Heinemann.
Wikstrom, T. and Loeber, R.(1997) ‘Individual Risk Factors, Neighbourhood SES and
Juvenile Offending’, in M. Tonry (ed.) The Handbook of Crime and Punishment, New
York: Oxford University Press.
Wilson, J. Q. and Kelling, G. (1982) ‘Broken Windows: The Police and Neighbourhood
Safety’, Atlantic Monthly (March): 29–38.
Young, J. (1999) The Exclusive Society, London: Sage Publications.
Youth Justice Task Force (1998) The Final Report of the Youth Justice Task Force, London:
Youth Justice Board of England and Wales.
8 Crime victimisation and
inequality in risk society
Tim Hope

Introduction
We live in both a risk society (Beck 1992) and an exclusive society (Young 1999).
At the same time as we orient ourselves around the risks and dangers which we
see surrounding us in our everyday lives, so also do our social and political
arrangements lend themselves to the magnification of inequalities in access to
those goods which reassure, protect or expose ourselves to risk. As our percep-
tion of the ‘bads’ increases, so do we seek to garner the ‘goods’ which would keep
them at bay, trading on our capital reserves and capacities across the various
spheres of economy, community and culture. The ‘ontological insecurity’ which
the condition of late modernity inspires in us (Giddens 1990; Young 1999) fuses
with the apprehension of mundane insecurities, pressurising us to invest in the
means of risk avoidance (Hope and Sparks 2000). As we feel increasingly that the
public sphere alone can no longer guarantee sufficiently the public goods of
everyday safety (see Garland 1996), so we are thrown upon our own individual
and collective resources and strategies to acquire the private goods which would
remedy our perceived security deficit. And our incapacity to protect ourselves
from risk leads to frustration with government – still seen as the primary
provider of safety in modern society – and with ourselves, as a reflection of our
own powerlessness in the face of the risks and harms that surround us.
Yet, access to capital in one sphere – for instance, through income and
wealth – remains an important commodity for acquiring capital in another
sphere – for example, in ‘community safety’ – promising to buy ourselves out of
risk and into security. That the dynamic of consuming private security itself may
be inherently and literally unsatisfying (Loader 1997) is not in itself a guarantee
that we will not continue to consume at an ever increasingly voracious rate.
And if security is a relative or ‘positional’ good – one that acquires value for its
possessor to the extent that others do not possess it to the same degree (Hirsch
1977) – then to the degree that those with capital successfully pursue their
private strategies of risk avoidance, those without will become increasingly
excluded from access to the goods of private security (Hope 1999; 2000).
Existing social inequalities will thus be reproduced in the newer spheres of risk
and risk avoidance.
194 Tim Hope

If such a process is at work in contemporary society we might expect to see

its consequences in the observed structure of outcomes – both of risk and of risk
avoidance. Even if we have little access to the decision-making processes of
individuals, and little chance of observing their pursuit of strategies of risk
avoidance in their everyday lives, we may still be able to infer their operation
from the observed structural patterns of risk. In this vein, this chapter essays an
‘actuarial’ analysis of the distribution of the risk to private citizens of household
property crime victimisation in England and Wales, as measured by the British
Crime Survey.

Crime risk in risk society

Actuarialism and criminology

Actuarialism in criminology has an ambivalent role. On the one hand,

The actuarial stance is calculative of risk, it is wary and probabilistic, it is

not concerned with causes but with probabilities; not with justice but with
harm minimisation. It does not seek a world free of crime but one where
the best practices of damage limitation have been put in place; not a utopia
but a series of gated havens in a hostile world.
(Young 1999: 66)

If actuarialism is ‘a major motif of social control in late modern society’ (ibid.), a

criminology whose methodology is actuarial may also collude with the newer
manifestations of domination and injustice inherent in the risk society,
elevating principles of societal risk management above those of individual
justice and liberty (Feeley and Simon 1993). Actuarial analysis thus becomes a
tool not only for key-players in the emerging markets of private insurance and
security but also more generally for many governance projects in social control
(Simon 1987) – an ‘administrative criminology’ for the risk society (Young
1999). Actuarial criminology may then become as complicit in social control as
did once a ‘modern’ (positivist) criminology in correctional penology (Hope
and Sparks 2000; Taylor et al. 1973).
Yet, on the other hand, in risk society, actuarialism itself has also become a
primary medium through which individuals, organisations and institutions
make sense of their world, communicate with one another and pursue strategies
of self-interest (Beck 1992). If society is becoming organised around considera-
tions of risk management and risk avoidance – not only in governmental
projects for the social control of dangerous classes (Feeley and Simon 1993), but
also profoundly in everyday lives and private practices (O’Malley 1992) – then
actuarial analysis must also constitute a major way of interpreting strategies of
risk management, including relations of inequality which may underlie them;
with the possibility that it could be critical of, as well as complicit in, strategies
of social control.
 Crime victimisation and inequality 195

In respect of household property crime – including, burglary, theft and crim-

inal damage – ‘risk factors’ such as where you live and how much you own, or
owe, have already become key factors informing insurers about your present and
future risk and what kind of contract they are willing to enter into with you to
deal with your apprehension of that risk. In risk society, people become increas-
ingly reflexive of the risks facing them (even if they cannot calculate their risks
precisely) and, especially, how their own risks are viewed by powerful institu-
tions which have the capacity to affect their individual, private security. For
instance, those who cannot get household insurance, or face high premiums,
rapidly acquire an everyday knowledge of how the crime risk of their commu-
nity is viewed by the master institutions of contemporary society. Similarly,
research has shown for some time that residents of ‘problem estates’ are often
acutely aware of the stigma attached to their community and will often hotly
resist the reinforcement of that stigma by external bodies, particularly their
social welfare landlords (Hope 1997). As research into the phenomenon of
neighbourhood watch suggests, the publication of local crime figures by the
police may not only stimulate increased security consciousness but also the
worry and anxiety which is supposed to be allayed (Hope 1995).
In light of the declining efficacy of the state to provide universal security
from crime, reflexivity towards crime risk becomes a more significant strategy of
everyday life. People may seek actively to provide themselves with private secu-
rity, and to ‘react to crime’ in their everyday lives, albeit with imperfect
knowledge of the risks confronting them or the means which might be available
for their protection (Skogan and Maxfield 1981). What is of interest, though,
are the collective consequences of these individual actions, not only for the level
but also for the distribution of private security.
The critique of ‘actuarial justice’ is concerned primarily to protect the
erosion of civil liberty which occurs when people are dealt with solely on the
basis of group characteristics. For example, where decisions to grant parole to
prisoners are based on the statistical probability of reconviction for the appli-
cant’s reference group (defined by age, gender, etc.) rather than on more
personal or moral criteria (Feeley and Simon 1993). Yet in its focus on indi-
vidual liberty, this critique may render itself less able to address genuinely
collective issues of social justice, such as class, race or gender bias in criminal
justice processes and outcomes, which are only likely to be revealed by aggre-
gate statistical analysis. So, in the interests of distributive justice, we would
need to know whether the emergent distribution of crime risk heightens ‘unjus-
tifiable inequalities’ in access to the rights and goods of private safety and how
the reflexive implementation of private security in response to risk is distributed
socially. And for this we need ‘actuarial’ criteria on which to base our moral and
political discourse about the new insecurities of the risk society.
Over the past twenty years, household crime victimisation surveys have
become a major technology for measuring crime risk and informing practice; and
their deployment has contributed both to the definition and management of
crime risk. The scale and resources necessary to conduct these have meant that
196 Tim Hope

they are a tool usually only available to large-scale organisations, primarily

governments. Their use in crime management seems set to grow. For instance, at
the time of writing, the Home Office for England and Wales has announced
plans to double the sample size of the British Crime Survey and to make its
coverage more consistent with police areas, presumably with the intention of
setting it alongside the current ‘official statistics’ of crimes recorded by the police
as another part of the state’s national accounting of crime risk. To be sure, domi-
nant modes of surveying may have major biases of response and coverage which
distort estimations of risk – most notably in the case of violence against women
(Stanko 2000) – but such distortions may be less to do with the technology per
se than with the practices and uses built into particular survey techniques
commonly used (see Mooney 1998). The burgeoning technology of risk control
per se is not necessarily dystopian – in this case, more actuarial knowledge of
crime risk is not an inevitable precursor of greater dominance. What is more at
stake in the politics of risk society is the reflexivity with which the actuarial anal-
ysis of risk is handled (Beck 1992). In particular, whether risk is treated as if it
were an accidental misfortune emanating from the seemingly ‘natural’ or given
environment of its potential victims (see Felson 1998) or as a manufactured risk
(Giddens 1990), embedded in the social structure, inherent in the social posi-
tions in which people are located and in the distributions of opportunities and
threats to which they are differentially exposed (Ewald 2000). Consequently,
before we conduct an actuarial analysis of victimisation risks, it is necessary to be
reflexive about the framework used to make sense of such analysis.

Risk positions and crime victimisation

As risks become regular or routine – or, alternatively, as risk-generating routines
proliferate – so risks become embedded into the structure of everyday life (see
Giddens 1984) and thus, consequentially, people come to acquire social risk posi-
tions in relation to them. Yet these positions are not necessarily identical to, or
consequential upon, ‘traditional’ social or class positions (Beck 1992). In the
social conditions of late modernity, neither risks nor social positions are fixed or
ascribed by traditions of community, culture or the inheritance of capital. Risks
fall widely amongst those who take part in modernity’s routines and activities.
Indeed, one of the most influential explanations for the growth of property
crime in contemporary society is that of the growth and change in the structure
of opportunities for crime produced by changes in everyday routine activities –
especially the growth of desirable and stealable goods – such as televisions, hi-fis
and CD players, computers and motor vehicles – and the increase of activities
away from the home and domestic environment (Felson 1998). That such
growth could be identified under macro-social conditions of apparently
increasing wealth, economic opportunity and equality (Cohen and Felson,
1979) presented an argument against traditional, class-based explanations of
vulnerability to crime risk. Moreover, while in industrial society, the distribu-
tion of crime risk primarily reflected the ecological patterns of urban growth
 Crime victimisation and inequality 197

and industrial production, produced chiefly as a consequence of social and

spatial proximity to populations vulnerable to offending, in late modern society,
crime risk patterns no longer so closely follow these imperatives, reflecting more
divergent patterns of distribution of communication and consumption rather
than of production (see Bottoms and Wiles 1997; Felson 1998).
Arguably, in the conditions of early modernity (industrial society) crime risk
flowed as a consequence of incumbency in positions ascribed by the social order:
since the lower classes were more prone to offending (as most prevailing expla-
nations predicted), lower-class people would generally be more vulnerable to
the risks of crime, in all its forms, since they were socially, culturally and
spatially proximate to and familiar with ‘offenders’. The higher classes would be
relatively free of crime risk since they were distant and socially segregated from
the lower (crime prone) classes and the places in which the latter were concen-
trated ecologically. In contrast, in late modern, risk society, incumbency in a
particular socio-economic position does not in and of itself guarantee safety or
generate risk. Opportunities for stealing and for being stolen from are abundant
in most walks of life; the ‘new ghettos’ are not necessarily segregated spatially
from the affluent suburbs (Logan and Molotch 1987); and society is open and
mobile to the extent that, in theory, ‘offenders’ are able to gain access to a
potentially large number of victims of all classes with fewer formal or practical
impediments to their mobility (Felson 1998).
In risk society, vulnerability to risk becomes individualised because people are
now presented with a proliferating range of options and choices which they are
in principle free to adopt in their everyday lives, freed from the proscribed folk-
ways of class, status or cultural identity (Beck 1992). Criminologists have
(rather unreflexively) conceptualised this individualisation of potential choice
and its consequences for crime risk in terms of a ‘lifestyle’ theory of victimisa-
tion. Here, it is not class or status per se which bestow risk or security; it is
individuals’ lifestyles which differentially expose or protect people from risk. Thus,
for instance, whether and when a person leaves their home unattended, where
they go to and for how long, whom they associate with and where, what they
own and consume, and in what quantities, and so on, all become variables
affecting individuals’ differential exposure to risk and the opportunities they
afford for self-protection or the protection available from other people. In such
explanations, other forms of difference and opportunity in addition to class – for
example, in gender, sexual preference and age – become more prominent, all
components of differential ‘lifestyles’ which may vary as much within income
groups as between them.
Nevertheless, as Beck (1992) notes, this does not necessarily imply that
every kind of risk is classless in its incidence, especially those risks, like crime
victimisation, which are not necessarily ‘global’ in nature, as are global warming
or pollution. The seeming ‘classless-ness’ of risk apparent in some writings on
modernity, such as Giddens (1998) and Beck (1992) himself, is perhaps a conse-
quence of a focus on the latter and a down-playing of the importance of the
former kind of risk. For more ‘mundane’ risks, like crime victimisation, it is not
198 Tim Hope

so much that risk society abolishes income- and status-based inequalities

through individualisation but that these embedded, structural and cultural
inequalities will re-emerge, despite the new arenas of life and opportunities for
self-expression which modernity has opened-up. Beck writes that:

The history of risk distribution shows that, like wealth, risks adhere to the
class pattern, only inversely: wealth accumulates at the top, risks at the
bottom. To that extent, risks seem to strengthen, not to abolish, the class
society. Poverty attracts an unfortunate abundance of risks. By contrast, the
wealthy (in income, power or education) can purchase safety and freedom
from risk.
(Beck 1992: 35, original emphases)

What differentiates the allocation of risks and safety between rich and poor
in risk society is precisely the individualisation which sunders traditional class-
based risks:

The ability to chose and maintain one’s own social relations is not an
ability everyone has by nature. It is, as sociologists of class know, a learned
ability which depends on special social and family backgrounds. The reflexive
conduct of life, the planning of one’s own biography and social relations,
gives rise to a new inequality, the inequality of dealing with insecurity and
reflexivity.
(Beck 1992: 98 original emphases)

As Pierre Bourdieu (1998) has shown in his various works, while the possession
of economic capital is a great enabler, the power of economic capital to shape
its possessors’ lifestyles and opportunities finds its expression and reinforcement
chiefly through its capacity to accumulate other forms of capital – social,
cultural, symbolic – which operate to achieve advantages for their possessors in
varying ways in diverse social spaces. In this respect, then, inequality in income
and status no longer ascribes a fixed quota of risk so much as it facilitates or
impedes the deployment of necessary and effective capital and capacities for avoiding
risk and/or acquiring safety. Though these acquisitions are achieved through indi-
vidualised strategies of everyday life, reflecting diversity and differentiation of
lifestyle and routine activity, their collective consequences and outcomes may
nevertheless reveal the aggregate inequalities of income and status that affect
the distribution of individual capacities and choices. The next part of this
chapter seeks to test this proposition in relation to the risk of property crime
victimisation.

Crime risk and the distribution of rich and poor

In conceptual terms, much of the difficulty in sorting out the risk factors of
crime victimisation as they are presented in the general population arises
 Crime victimisation and inequality 199

because the victimising act itself is the result of a convergence of two separate
processes:

A fundamental aspect of predatory crime is that it occurs in a social context

where there exists a convergence of victims and offenders in time and space.
(Miethe and Meier 1994: 44)

This convergence has been expressed also in terms of the conditions necessary
for a criminal event to happen – what has become known as the routine activity
approach, comprising the convergence of ‘suitable’ victims (or targets) and
‘motivated’ offenders in conditions (i.e., low ‘guardianship’) where it is unlikely
that a third party capable of exercising control over the situation will intervene
(Cohen and Felson 1979). Two basic distributions affecting the likelihood of
property crime victimisation risk can be distilled from this literature (for which
see Felson 1998; Miethe and Meier 1994):

1 Opportunities for crime, as presented by potential victims to putative

offenders.
2 Exposure to risk, as attributes of victims in relation to supplies of offenders.

The phenomenal growth of household property crime victimisation, espe-

cially burglary, during the second half of the twentieth century suggests the
proliferation of a ‘modernisation risk’ within post-industrial societies (Cohen
and Felson 1979). Broadly, this risk would seem to consist of a number of char-
acteristics of people’s homes and lifestyles (see Winchester and Jackson 1982)
which may have proliferated:

• environment: expressed in physical forms which combine privacy – for

example, the degree to which the property is not under the surveillance of
others – with accessibility – for example, the ease with which it is possible to
gain physical access to all parts of the residential property (Hope 1999);
• occupancy – the extent to which domestic property is unoccupied and
unguarded due to the routine activities of people’s lives in work and leisure
away from the home environment (Cohen and Felson 1979);
• value: the extent to which domestic property is ‘valuable’ as a commodity
which can be exchanged illicitly for other commodities (e.g., cash, drugs) –
the most valuable commodities in this regard (including cash itself) usually
have a high exchange value relative to their portability (Felson 1998);
• security: the extent to which domestic property is guarded by physical, elec-
tronic or human means which deny or increase the risk of unauthorised
access and appropriation (Winchester and Jackson 1982).

Nevertheless, there are different social positions which the ‘rich’ and ‘poor’
are able to take up in relation to these risks (I use these terms for ease of presen-
tation, though advisedly).
200 Tim Hope

Opportunities for crime

There are a number of ways in which opportunity theories of crime would
suggest that the ‘rich’ may be more vulnerable to household property crime
victimisation than are the ‘poor’.

1 The rich have more valuable and attractive property to steal – and may be
relatively more casual in safeguarding it if they have full household insur-
ance and relatively low premiums and excesses – i.e., the phenomenon of
moral hazard (Litton 1982).
2 As more household members are likely to be in full-time employment, they
are more likely to leave their homes unattended during weekdays and holidays.
3 Their dwellings may be easier to enter unobserved – for example, larger,
detached houses in their own grounds may afford more opportunities for
unobserved and undisturbed access to property (Hope 1984; 1999; 2000;
Winchester and Jackson 1982; Maguire 1982).

Nevertheless, the rich may be less vulnerable than the poor because:

1 They are more likely to purchase or adopt a greater range of household secu-
rity measures, especially more expensive home surveillance systems (Hope
2000).
2 They may take greater security precautions in their everyday lives because
they may be more ‘risk-averse’ (Field and Hope 1990) – for example, by
believing they have much to lose from crime victimisation, including the
peace and tranquillity of their communities (Loader et al. 2000).

The ‘poor’ may also be more vulnerable to household property crime than the
rich, because they are less able to afford, or have available, security measures and
protection (Hope 2000). But they may be less vulnerable to victimisation
because, being poor or unemployed, they may be more likely to spend time in
the home environment, thus exercising more guardianship over their household
property (Cantor and Land 1985).

Exposure to risk
In the context of household property crime, exposure to risk primarily means
exposure to active property crime offenders. Here, the likelihood is that the
poor will be more exposed to risk than the rich because they are more likely to
live in areas where active property offenders are also likely to live (Wikstrom 1998).
People who live near poor people are also likely to be active in property
crime against the poor because:

1 Having a greater likelihood of being poor themselves, they have (variously)

more or greater motivations to commit crime (especially if they are also young
and male).
 Crime victimisation and inequality 201

2 Young males in poor areas may also be subject to less supervision and social
control in the community, and spend more time in public in the company of
their peers, all of which are conducive to the development of active crimi-
nality (see Wikstrom 1998).
3 They may also be more frequent property criminals because they have a need
for a greater return from crime – for example, to supplement income, or
support a drug habit (cf. Sullivan 1989) – or because their (poor) victims
yield a lower per offence return of valuable items.
4 They are less likely to travel to other areas to commit crime because (a) they
lack transport or (b) are averse to committing crime in unfamiliar areas,
where they might stand out (Bottoms and Wiles 1997).

As a corollary, the rich may be less exposed to risk than the poor because
they are more likely to avoid crime by living in the kinds of area which are
orderly (i.e., they do not contain active resident offenders or keep such potential
persons under some kind of social control), distant (i.e., not close to places
containing offenders) and/or buffered from criminal areas (i.e., there are other
areas or obstacles lying between them and source of criminals – see Hope 1999).
Nonetheless, some of the rich may be more at risk than other rich, or even
than other poor, if they live in closer proximity to the poor (see above) by
living in ‘mixed class’ neighbourhoods or ‘gentrified enclaves’ in otherwise
deprived urban areas (see Hope 2000; Hirschfield and Bowers 1997).
In sum, various considerations of both opportunities for crime and exposure
to risk result in differing predictions about the risks of household property crime
for the ‘rich’ and the ‘poor’ respectively. These contradictory predictions may be
enough to satisfy those who regard only universal consistency as ‘proof’ of any
axiom (e.g., Wilson and Herrnstein 1985). Because the rich are, in this case,
sometimes more at risk than the poor, then victimisation it is suggested must be
classless. Nevertheless, as will be suggested below, while the above predictions
are apparent only with other things being equal (as they are in multivariate, statis-
tical analysis) the likely reality is usually that other things are not equal. And
one of the key inequalities is that rich and poor have, in sum, unequal propensi-
ties towards the various opportunity and risk factors of property crime
victimisation.

Models of risk
A certain kind of approach to causation tends to see these as separate questions.
In this kind of reasoning, the existence of opportunities and exposure processes
are thought to indicate particular explanations of crime. For example, that
offending is a matter of rational choice on the part of the offender such that the
task of explanation is to demonstrate that these are general processes, irrespec-
tive of their observed correlation with other distributions, such as income and
wealth which, likewise, are taken to be indicators of competing explanations of
crime, offender motivation, or the absence of social control. In some versions of
202 Tim Hope

this approach, the causal contenders are required to ‘slug it out’ to see which one
wins outright. In other versions, the task is to apportion the explanatory spoils –
so much to this one, so much to that – so as to develop and test ‘integrated’
theories. Methodologically, these causal strategies require clear specification of
expected effects, the development of measurement indicators which are unam-
biguous (i.e., which either measure opportunity/exposure processes or those of
motivation and control), reliable sampling strategies, and the use of appropriate
multivariate techniques of statistical analysis (see Miethe and Meier 1994, for
an elaboration of this research strategy).
Although in the past we had only officially-recorded crime data, now we
routinely possess one other source of data containing potential risk factors

Table 8.1 Household and area risk factors for household property crime

Household risk factors

The worse off?
council tenant (+)
young head of household (+)
children in household (+)
lone-parent household (+)
Asian sub-continent origin (+)
low level of household security (+)

The better off?

non-manual occupation (+)
house-dweller (+)
number of cars (+)
detached house (+)
income (+)

Area risk factors

Deprivation?
single parents (+)
children aged 5–15 (+)
private rented housing (+)
inner city (+)
‘rich’ ACORN groups (+)

Affluence?
cars per household (–)
deprivation index (–):
• overcrowded households
• large families
• housing association rental
• male unemployment
‘poor’ ACORN groups (+)
Source: 1992 British Crime Survey (BCS), 1991 UK Census (Osborn and Tseloni 1998; Ellingworth et al.
1997); 1996/1998 BCS, 1991 UK Census (Budd 1999).
Note: Household property crime includes: burglary, theft from dwelling and criminal damage to property
associated with the dwelling, its contents or grounds. A plus sign (+) = a positive risk factor (crime risk is
increased); a minus sign (–) = a negative risk factor (crime risk is reduced).
 Crime victimisation and inequality 203

(predominantly socio-demographic) – the ‘cross-sectional’ household victimisa-

tion survey. However, since these are surveys of reported victimisation events,
information on the likely activities of offenders is bound-up with information
on the characteristics and activities of victims. As a result, ambiguities remain
within the risk models developed, leading to inconsistent predictions of the
crime victimisation risk as it affects, respectively, the rich and the poor (for
examples, see Pantazis and Gordon 1999).
This can be illustrated, in Table 8.1, with reference to the risk factors identi-
fied in three recent efforts to model risks of household property crimes using
British Crime Survey (BCS) data (Budd 1999; Osborn and Tseloni 1998;
Ellingworth et al. 1997). Despite utilising differing sweeps of the BCS,
modelling assumptions, selections of explanatory variables, levels of measure-
ment of variables, and the inclusion of information on prior victimisation, the
three models come up with consistent findings regarding the socio-demographic
predictors of risk. Nevertheless, there is also inconsistency within each model as
to the predictors associated with rich and poor. Each of these models include
risk factors associated not only with households but also (contextual) factors
associated with the residential area in which the household is located; yet for
both levels there are contradictory patterns – both rich and poor areas and
households seem at risk.
Nevertheless, looking just at the ecological distribution of rich and poor (again
estimated from the BCS) we find a much simpler and straightforward relation-
ship with household property crime risk, as illustrated in Figure 8.1. Here,
victimisation risk seems unambiguously correlated with the distribution of rich
and poor areas. How, then, can we account for this paradox – an ambiguity of
risk at the individual level but a consistency of risk at the community level?
Notwithstanding the relative propensities of rich and poor respectively to be

Figure 8.1 Inequality and property crime

Source: 1992 British Crime Survey/1991 Census
204 Tim Hope

associated with particular opportunities and risks, part of the answer at least lies
in addressing two questions:

1 What is the appropriate level at which risk factors for household property
crime operate – are they a product of communities or individuals, and
which risk factors operate at which level?
2 How are rich and poor distributed within and between communities so that,
even if they had different propensities towards risk, the pattern of their
respective ecological distributions would result in the observed area pattern
of outcomes?

One strategy for research would be to allocate risk factors a priori to each
level – for example so that factors conjectured to be indexing the supply of
offenders would be allocated to the community level, while those indexing
opportunities would be allocated to the individual level (Miethe and Meier
1994). Yet while there may be some merit in this approach, again, it relies
upon unambiguous specification, not only in the measurement of indicators
but also in their allocation to the appropriate level. In practice, few data sets
are able to afford this, even if it were conceptually possible to do so with
respect to every risk factor. Yet even if the rich and poor had consistently
different types of risk propensity, we would still be left with how to handle the
possibility that the varying outcomes observed were due to the different distri-
butions of rich and poor within and between communities. In other words, we
need to consider how community and household risk interacts to produce
victimisation outcomes (Miethe and Meier 1994) and how such interaction is
conditioned by the respective distributions of rich and poor between and within
communities.
If we consider area level risk factors to have in common the possibility that
differences in victimisation represent differences in exposure to risk (in the
shape of active property offenders), then residents of poor areas will have more
exposure to risk than those in rich areas, because poor areas may contain or
attract more frequent and active property offenders than rich areas (Felson
1998: Chapter 2).
Consequently, how the different propensities of rich and poor to household
property crime victimisation are produced will therefore depend upon the
balance of opportunities to crime and exposures to risk attaching to individual
households, relative to the social context in which they are located. For example, we
might expect to observe some of the following interactions:

1 That the poor in poor areas have a higher risk than, say, the poor in rich areas
because, over and above the balance of opportunity and risk associated with
poverty, they are more exposed to risk by virtue of their area of residence.
2 The rich in poor areas also have higher risks than the rich in rich areas
because they are more exposed to offenders and have more goods to steal.
 Crime victimisation and inequality 205

3 The rich in rich areas may have a higher risk than the poor in rich areas
because they present more opportunities.

Nevertheless, whether we can observe these interactions will also depend upon
how the individual rich and poor are distributed between rich and poor areas. In
this respect, the overall pattern of outcomes of property crime victimisation risk
– whether in respect of individuals or communities – will depend upon the class
mix distribution of potential risk propensities attaching to the rich and the poor
as they are distributed, ‘ecologically’, across the range of communities (see
Braithwaite 1979, for an investigation of the class mix hypothesis with regard
mainly to offending propensities). Even though the ‘modernisation risks’ of
property crime victimisation may be similar for rich and poor, the specific posi-
tions which they may be able to take with regard to these risks may vary
according both to their personal and social circumstances. The next section
provides an empirical investigation of these possibilities.

The distribution of risk

Data in the subsequent analysis is taken from the 1992 British Crime Survey.
Information on households comes from the survey, while information on area
characteristics is taken from the 1991 Census and refers to the postcode sector
in which the household was located (i.e., areas of around 2,000 households
defined by the first half of the standard six-figure unit British postcode, plus the
next digit, for example AB1 2). The analysis is concerned with variation in the
number of household property crimes (defined above) experienced by
responding households over the twelve month BCS recall period. The specific
types of risk factors used in this analysis are those identified as having a statisti-
cally significant effect on the risk of household property crime in the
multivariate model estimated by Ellingworth et al. (1997) (see also Table 8.1,
above), to which the reader is referred for more details about data, measure-
ment and model specification. In this analysis, the aim was to construct four
different indices of affluence and deprivation applying to each respondent.
These were:

• an area deprivation index

• an area affluence index
• a household deprivation index
• a household affluence index.

These are listed in Table 8.2. None of these indices constitutes a perfect
measure of affluence or deprivation, and some relevant BCS data are left out.
For instance, household income in the 1992 BCS has a high non-response rate
and may be unreliable – pace Pantazis and Gordon (1999) who conclude on the
basis of this variable that the rich experience more crime than the poor.
Nevertheless, the variables used aim to capture primary aspects of economic
206 Tim Hope

opportunity (e.g., educational level) and consumption (e.g., car access and
housing tenure) which have come to be seen increasingly as important surrogate
measures for income and quality of life (JRF 1995). Additionally, these variables
have already been used in multivariate models of risk (see Table 8.1). Finally,
given the binary nature of the indicators – and the observation that the
frequency distribution of victimisation is not normally distributed (Osborn and
Tseloni 1998), relatively simple descriptive and distribution-free statistical tech-
niques were used to investigate the possibility of interaction effects of the kind
noted above.
Table 8.3 shows the correlation between the various indices and property
crime victimisation. As a check on the validity of our indices, there is a strong,

Table 8.2 Indices of individual and area deprivation and affluence

Area deprivation index*

proportions of households with:
• overcrowding (⬎ one person per room)
• three or more children
• housing association accommodation
proportion of adult males:
• unemployed

Area affluence index*

average number of cars per household
proportion of detached dwellings

Household deprivation index**

rental tenure (household)
not in employment (respondent)
children under 5 years old (household)
lone-parent household (household)
under 30 years old (respondent)
no use of car (household)
completed full-time education under 17 years old (respondent)

Household affluence index**

owner-occupier (household)
aged 30–59 (respondent)
in full-time work (respondent)
two or more cars (household)
completed education at 20 years old or later (respondent)
detached dwelling (household)
Source: 1991 Census (area deprivation and affluence indices); 1992 British Crime Survey (household
deprivation and affluence indices)
Notes:
In the ‘Area deprivation index’ and the ‘Area affluence index’, variables were identified for the area level
indices using principal components analysis – the variables in the two scales showing collinearity
(Ellingworth et al. 1997: 213)
*Variables standardised, additive scale
**Variables binary coded, additive scale
 Crime victimisation and inequality 207

negative correlation between the respective indices of deprivation and affluence

at both the area (⫺.815) and individual (⫺.770) levels, so we can be confident
that the indices are measuring the different dimensions of rich and poor. Table
8.3 also shows that residents in poor areas face a higher risk, and those in rich
areas a lower risk; and also that poor households have a higher risk and rich
households have a lower risk of victimisation. Nevertheless, these correlations,
though statistically significant, are relatively weak. What are also relatively
weak are the correlations between area and household deprivation and afflu-
ence. Though a definite tendency, by no means all the rich live in rich areas
(.319) nor do all the poor in poor areas (.248), at least as defined in this anal-
ysis. These weak inter-correlations provide some grounds for investigating
whether there are stronger patterns of victimisation risk evident from the inter-
action between household deprivation (or affluence) and area deprivation (or
affluence).
Such patterns are indeed evident from the findings presented in Table 8.4
which shows the correlation between household deprivation (or affluence) and
property crime victimisation within different levels of area deprivation (arranged
in quintiles from the 20 per cent least deprived areas to the most deprived quin-
tile of areas). Clearly, property crime risk increases appreciably according to the
level of deprivation of the community (see Figure 8.1). Nevertheless, as antici-
pated, there is also evidence of the conditioning of individual risk according to the
level of area deprivation. Thus, in the least deprived (most affluent) areas the
rich have higher risk rates than the poor. Indeed, contrary to the overall pattern
(Table 8.3), in the least deprived areas (first quintile), the more deprived the
household on our measure, the less at risk they seem to be. The pattern of higher
risk rates for the ‘rich’ relative to the ‘poor’ seems to hold generally in the 40
per cent least deprived areas in the national sample. However, the pattern of
risk seems to reverse when one begins to approach the ‘other Britain’ – the half
of the country with the more deprived or less advantaged communities. Here, it
is the poor who are more likely to be at risk than the rich.

Table 8.3 Property crime victimisation, deprivation and affluence

Rank-order correlation coefficients (Spearman’s Rho)

Household Area Area affluence Household
victimisation deprivation deprivation
incidents (n.)
Area .133
deprivation
Area affluence –.154 –.815
Household .051 .248 –.284
deprivation
Household –.029 –.269 .319 –.770
affluence
Source: 1992 British Crime Survey, 1991 UK Census
Notes: All coefficients significant at p<.01. Weighted data. Base no. 11,730.
208 Tim Hope
Table 8.4 Correlation between property crime victimisation, (a) household
deprivation (b) household affluence, within levels of areas deprivation

Rank-order correlation (Spearman's Rho)

Area deprivation Property crime Correlation Correlation
incidence rate per household household affluence
100 households deprivation and and property crime
property crime
1st quintile 17 –.043* .064*
(lowest 20 %)
2nd quintile 26 –.022 .043*
3rd quintile 31 .055* –.028
4th quintile 38 .063* –.038
5th quintile 49 .034 .021
(highest 20 %)
Source: 1992 British Crime Survey, 1991 UK Census
Notes: Weighted data. Base no. 11,739
* Correlation statistically significant at p⬍.05

Table 8.5 summarises patterns of risk according to the individual risk factors
used in our overall indices. In the least deprived parts of the country (first and
second quintiles), those most at risk are younger households, and those with
young children and/or two or more cars. Those least at risk are persons not in
work, including the elderly, and those who do not own a car. In the most
deprived parts of the country (fourth and fifth quintiles), those most at risk
include renters, families with young children, lone-parent households and both
young adults and the middle-aged. In both rich and poor Britain, then, there
would seem to be a clear divide in household property crime risk between the
younger population and the older, less mobile, and retired populations.
Arguably, much the same sort of risk processes apply: for instance, younger
households are more likely to leave their homes unoccupied and to have fewer
adults around the home environment to supervise activity (Felson 1998).
Additionally, the relatively more affluent households may also be more vulner-
able because of their possession of more desirable goods – in richer Britain these
may be the two-car-plus households; in poorer Britain they are likely to be the
middle-aged. Yet at the extremes of income, households may be able to buy (or
not buy) themselves out of risk. Thus, in affluent Britain, owners of detached
houses may be able to afford sufficient security in order to offset their attractive-
ness to burglars at the level of risk present in their communities (see Hope
1999; 2000); something perhaps unachievable for detached house dwellers in
poorer neighbourhoods, as their relatively greater risk in the fourth quintile
suggests. Similarly, in poorer Britain, renters may be more vulnerable to victimi-
sation than owners, not only because they are living in the same
neighbourhoods (e.g., social housing estates) as offenders but also because they
are unable to take many security precautions (Hope 2000).
Finally, how does the class mix of an area affect its crime rate? Looking at the
distribution of property crime rates between neighbourhoods, we can estimate
that around one-fifth of the victims of household property crime live in the 10
 Crime victimisation and inequality 209
Table 8.5 Risk factors for household property crime victimisation by level of area
deprivation

Level of deprivation of area (quintiles, from lowest to highest)

Risk factor Q1 Q2 Q3 Q4 Q5
renter 0 0 + + +
not in work – 0 0 0 0
young + 0 + + +
children
lone parent 0 0 + 0 +
young adult + 0 + + 0
no car – 0 0 0 0
low 0 0 0 0 0
education
middle- + + 0 0 +
aged
two-plus + + 0 0 0
cars
detached 0 – – + 0
house
Source: 1992 British Crime Survey, 1991 UK Census
Notes: + signifies a positive risk; – signifies a negative risk; 0 signifies no significant risk. All differences in
means significant at p ⬍ .055 (Mann-Whitney U-Test)

per cent of the residential areas with the highest crime rates, and suffer over one-
third of the total of household property crime (which itself constitutes just under
one half of all victimisation recorded in the BCS). Over half of all property
crime – and over one-third of all property crime victims – are likely to be found
in just one-fifth of the communities of England and Wales. Conversely, the 50
per cent of communities of England and Wales with the lowest crime rates suffer
merely 15 per cent of household property crime victimisation, spread between a
quarter of victims (Hope 1996; 1997). Inequality in both individual and area
distributions go together – victims who are multiply (or repeatedly) victimised in
a given period are more likely to live in high crime rate areas than are other, less
frequently victimised victims (Osborn and Tseloni 1998; Trickett et al. 1992).
Figure 8.2 suggests, broadly, that this inequality in the distribution of commu-
nity crime risks is correlated with the ratio of the poor to the rich within the
community. It suggests not only that as the crime rate of the community
increases, so does the ratio between poor and rich within the community but also
that the concentration of the crime rate coincides with the increasing rate of
concentration of the poor. Thus, the 20 per cent of communities with the highest
crime rates not only have appreciably higher crime rates than other communities
but also have much greater numbers of the ‘poor’ relative to the ‘rich’.
Although the data presented here are relatively crude measures of risk and
social position they derive nevertheless from a representative national survey and
provide some guidance about crime victimisation risk, albeit roughly calibrated.
In general, they provide some evidence on the likely social risk positions of the
‘rich’ and the ‘poor’ with respect to property crime victimisation. Although the
210 Tim Hope

Figure 8.2 Property crime rates and poverty concentration

‘modernisation risks’ of residential crime – that is, the risk attaching to environ-
ment, occupancy, value and security – may be generally applicable to most
instances of victimisation and are therefore, in a broad sense, ‘classless’, the
specific social risk positions which rich and poor adopt in respect of these risks
vary considerably. Moreover, there does not seem to be any intrinsic risk
attaching to individual class position; rather, the risk of victimisation associated
with being rich or poor depends upon the social (community) context in which
individuals are located. And a primary feature of that social context is its class
mix: specifically, the concentration of the ‘poor’ relative to the ‘rich’. The
concluding section of this chapter considers some of the political ramifications.

The politics of risk avoidance

Risk strategies of rich and poor

While the state may claim to have the power to prevent crime by being autho-
rised to take action directly against ‘offenders’, private citizens are able to do no
more than try to pursue individualised strategies of risk avoidance. Clearly,
contemporary governments hope that individualised strategies will also reduce
crime in the aggregate – an expectation that has come to be more fervently
wished for by politicians as they come to terms with the declining capacity of
the ‘sovereign state’ to deliver overarching (public good) security at a level
 Crime victimisation and inequality 211

expected by its citizens (Garland 1996). Yet, this responsibilisation strategy on the
part of the state vis-à-vis its citizens does not necessarily contain within it a
conception of the public good. Citizens are not being asked to carry out altru-
istic acts of intervention, such as to mount guard on other people’s homes, to
engage in self-denial, or to give up the desire for or possession of CD players or
portable televisions. Rather, they are being expected to attain security both for
themselves and in the aggregate, largely through the pursuit of their own indi-
vidual self-interest in the avoidance of risk. This is not unlike developments in
other areas of social welfare provision and may be part of an emerging ‘new
orthodoxy’ of welfare (Jordan 1998). Yet a core belief in rational self-interest is
no guarantee that the resulting outcomes in terms of the distribution of risk and
safety will be equitable nor, as important, deducible simply from the summation
of individual rational strategies. Rather, it will involve:

An understanding of the interactions between the (often unintended)

collective consequences of individual actions; the emergence of new kinds
of risks, and new methods of managing them; changing technologies in the
creation of exclusive ‘clubs’ for sharing risks and collective benefits; and the
role of nation states as providers of public goods in a global economic envi-
ronment.
(Jordan 1998: 113)

Class differences in income and wealth may result in different risk positions
occupied respectively by rich and poor via the resources and strategies which
affect individuals’ ability to avoid risk. Yet the collective consequences of individ-
uals’ strategies may produce inequalities in access to safety and the avoidance of risk.
We can illustrate some of the resulting problems with reference to the data
presented in this chapter.
Simply put, the likely most effective strategy of property crime risk avoid-
ance on the part of the poor is for them to remove themselves from high-risk areas
and live amongst richer neighbours. Reference to Table 8.4 suggests that if they
can get into the richest neighbourhoods their risk will greatly reduce, possibly
to the extent that they become less at risk than the rich. Unfortunately for
them, the most rational individual strategy for the rich to avoid property crime
is to avoid the poor. Risk seems to increase the greater the ratio of poor to rich in
a community (see Figure 8.2), and the rich in poor areas face a higher risk than
that of the rich in rich areas. Other things being equal (which they rarely are),
neither poor nor rich have intrinsic crime victimisation propensities which
they carry around with them (Table 8.4). So, in principle, the individual rich
do not have anything to fear from the individual poor by allowing them to live
alongside them (see Table 8.5). But they do seem to have something to fear by
way of increased risk if they allow too many poor to be their neighbours (Figure
8.2), presumably because this would increase the chances of living in closer
proximity to active offenders. Whatever the niceties of the situation, the strate-
gies of risk avoidance for rich and poor end up as radically opposed because,
212 Tim Hope

rationally, each would see the other as part of their solution to their individual
problems of risk avoidance.

The risk pools of rich and poor

Over and above individual class propensities – which this analysis shows not to
be invariant across social situations – property crime risk varies according to the
community of residence – specifically, the nature of the people amongst whom
one lives. On this basis, other things being equal, the most rational strategy for
the rich would seem to be to insulate themselves from the poor. Neither the
individual rich nor the individual poor can control or impose limits on their
neighbours and only the very rich can insulate (fortify) themselves and their
property to the degree that it does not matter in terms of their perceived secu-
rity how many poor surround them. In lieu, insulation from risk can be achieved
most cost-effectively through a collective strategy of pooling risk (Jordan 1998) –
by ‘voting with one’s feet’ to ensure that one is living amongst ‘low-risk’ people.
By pooling risk with others a security ‘club’ is created with the advantage that
the low risks enjoyed by your neighbours are also enjoyed by you, and vice
versa. In this respect, the ‘club’ offers collective goods to members – including
sharing in the benefits of other people’s individualised strategies of risk avoid-
ance – which are not available to those who are excluded (Hope 2000; see also
Jordan 1996).
Entry into the private security ‘club’ of the better-off neighbourhood is, of
course, determined by property prices – the capacity which provides individuals
with capital to buy into more affluent neighbourhoods also allows the purchaser
to buy into the collective security goods of the suburb (Hope 1999). Yet the
collective security goods available to residents may be undermined unless their
exclusive nature can be maintained. Our analysis suggests that this is because
private security resembles what Hirsch (1977) called a positional good – some-
thing which has value to its possessor only because others cannot possess it to
the same degree and whose supply cannot be increased greatly by changes in the
basic factors of production. The positional advantage offered by the ‘exclusive’
suburb itself diminishes through congestion unless external demand for its goods
can be resisted (Hirsch 1977). Positional goods, which residents might once
have enjoyed exclusively, loose their value if everyone has equal access to them.
If the cost of private security can be reduced by avoiding risk in the first place –
particularly by avoiding proximity to a ‘supply’ of offenders – then the private
security value of the suburb will depend upon the extent to which it can main-
tain its exclusivity by preventing others, including would-be offenders, from
encroaching upon it (Hope 1999). Having done so, it can then provide
members with the collective benefits of individuals’ security actions. That is,
keeping the collective externality benefits generated by individual private security
actions within the club (Hope 2000). It is thus in the interests of the rich
within their (low-) risk pool to resist encroachment by the less rich; ironically
for the poor, the positional advantage of the exclusive suburban risk pool dimin-
 Crime victimisation and inequality 213

ishes the more that the poor successfully overcome the barriers to sharing the
positional security advantages of the rich. As Table 8.4 and Figure 8.2 show, the
fewer the rich who live among them, the more their risk increases.
Thus, the defence of the better-off community against ‘encroachment’
becomes an important weapon in the preservation of neighbourhood values and
amenities, including security. Much community action in suburbs is ‘preserva-
tionist’ (Savage et al. 1992), not least with regard to anti-crime efforts (Skogan
1988), and better-off communities do have some advantages in generating
resources to defend themselves. The ‘weak’, overlapping ties which characterise
social relationships in better-off suburbs (see Baumgartner 1988; Loader et al.
2000) provide linkages between sources of power and influence within a
community (Granovetter 1973). These also provide opportunities for network
‘closure’, creating a collective reciprocity of social obligation from which social
capital can be generated (Coleman 1990). In turn, social capital can be
mobilised to create collective goods like community organisations and residents
associations. These groups can then cash-in the (symbolic and cultural) capital
which the neighbourhood holds with extra-communal sources of power, such as
public police and local government, to mobilise resources to preserve neigh-
bourhood amenities, including security (Skogan 1988). Without external
intervention, the market in private security becomes a struggle for positional
advantage amongst competing risk pools – a struggle to avoid the congesting
and crowding of positional security advantages. As Mike Davis observes of the
‘home-owner’ politics of the communities of Los Angeles:

The market provision of ‘security’ generates its own paranoid demand.

Security becomes a positional good defined by income access to private
‘protective services’ and membership in some hardened residential enclave
or restricted suburb. As a prestige symbol … ‘security’ has less to do with
personal safety than with the degree of personal insulation, in residential,
work, consumption and travel environments, from ‘unsavoury’ groups and
individuals, even crowds in general.
(Davis 1990: 224)

In contrast to the rich, increasing poverty in the social mix of the commu-
nity seems to be associated with a greater victimisation of the poor (Figure 8.2).
Further, a concentration of poverty in the community results in a greater concentra-
tion of victimisation amongst the poorer sectors of such communities (Table 8.4).
And to the extent that the poor are excluded, whether consciously or not, from
the security pools of the rich, they face the danger of becoming trapped in the
risk pools of poor neighbourhoods and neighbours. In late modernity a key
characteristic of the communities of concentrated poverty is the concentration of
social dislocation (Waquant 1995; Lash and Urry 1994; Wilson 1987). The expe-
rience of destabilised neighbourhoods caught in a spiral of economic and social
deterioration is increasingly one where adverse circumstances ratchet together
to produce compounded social dislocations for both residents individually and
214 Tim Hope

communities collectively – for instance, the compounding of offending with

other dislocations including youth poverty and drug abuse (Stewart and
Stewart 1993, cited in Smith and Stewart 1997). Such compounding of disloca-
tion within the community not only intensifies the harm suffered by residents
but also has knock-on effects on the ‘institutions’ of the community, which
themselves might comprise some of the social defences against such harm
(Hope 1995; 1997).
Severe economic decline and resource inequality may be producing a vicious
spiral in which not only does crime and disorder impinge on the effectiveness of
local institutions – for example, schools, youth services and voluntary organisa-
tions – but also cripples their ability to provide the kind of framework of order
which might ameliorate or offset the dislocations and heightened risks of disad-
vantage. In sum, the risk pools of concentrated poverty:

may be due to a failure to prevent the process of criminal embeddedness of

local youth [see Hagan 1994] – fuelled by low economic opportunity –
which escalates their offending activities, with certain vulnerable groups of
residents coming to be repeatedly victimised as a consequence, particularly
as social controls progressively diminish and the neighbourhood tips into
increasing disorderliness. Each element in the process may interact dynami-
cally with the others to account for the sharp growth of crime in particular
communities caught in the spiral of deterioration.
(Hope 1996: 178)

Conclusion: the politics of risk avoidance

As has been said often enough before, the rich live in communities of choice,
the poor in communities of fate. The capacity or incapacity to avoid risk under-
pins the social risk positions taken up by rich and poor and leads to distributive
injustice – simply put, the rich have unequal access to security from crime rela-
tive to the poor. Although once the meanings of a social good are clarified it
becomes possible to work out how and on what criterion the good ought to be
distributed justly, there is equally nothing intrinsic to the meaning of the good
which guarantees how, and by what means, it can be distributed feasibly in prac-
tice. In the case of private security, as with the delivery of many other social
goods, that becomes a political task (Walzer 1983).
The political problem for governments in the face of high crime rates has
been seen as that of ensuring, by one means or another, a level of provision of
security goods sufficient to meet at least some minimum level of need
(demand) for the collective good. For these reasons, Garland (1996) identifies
two strategies which have been pursued by governments: one consists of
greater investment in the punitive, deterrent and incarcerative services
provided by the state; the other, the responsibilisation strategy, summarises
efforts to generate more ‘private’ (including citizen) investment in the produc-
tion of collective security. An optimistic prognosis is the hope that by raising
 Crime victimisation and inequality 215

the overall (national) aggregate level of provision of private security goods of

any kind, the collective need for security will be met. For example, British
governments over the past couple of decades have pursued campaigns and
programmes of exhortation, demonstration and (modest) investment in crime
prevention with the principal aim of raising security levels for and amongst
private citizens. And governments remain attached to the ‘performance indi-
cator’ of achieving reductions in national reported crime rates (Barr and Pease
1990). In a potent political sense, the reduction of the national crime rate is
what governments of nation-states see as their mandate from the electorate as
a whole (Home Office 1999).
Nevertheless, it appears that the threat to the collective good of private
security, when it is disaggregated as individual risk, is inequitably distributed
amongst members of the population. Not everyone seems actuarially ‘at risk’ of
property crime victimisation to the same degree. This apparent inequality in
need poses problems for the project of maintaining the collective provision of
private security, particularly through growth in the total amount of privately-
provided security. It is not that inequality in the distribution of a social harm
itself poses a problem for social justice, but that the available goods to remedy
the harm are not distributed on the basis of need but on some other criteria (say
income and wealth) which comes to dominate their provision, irrespective of
the way in which the meaning of the good implies that it ought to be
distributed (Walzer 1983). Thus, we may be far more egalitarian about the
distribution of the collective good of private safety than we are about the distri-
bution of private wealth. So it becomes unjust if we allow the distribution of
wealth to dominate the distribution of safety – the inequality which we tolerate
in one sphere becomes intolerable when it comes to determine the distribution
of another, separate and more equitably regarded, sphere. It is not so much that
the ‘commodification’ of access to private security is unjust in an intrinsic sense
– that fetishistic, private consumption corrupts the nature of the collective good
– but rather that the injustice lies in the consequences of commodification – that
is, the facility which commodification allows (i.e., money purchase) for those
with greater capital to distort the basis on which the good might otherwise be
distributed (see also Waldron 1995).
Tackling inequality in the distribution of private safety, then, lies not in the
direction of redistributing insecurity (see Barr and Pease 1990), since it is not
feasible to assume that members of a society predicated on the pursuit of indi-
vidual self-interest would accept or let such a redistribution of harm take place,
but in safeguarding the autonomy of the just criteria for distributing private
security. To paraphrase Walzer (1983: 20): ‘private security should not be denied
to people who do not possess some other good – for example, wealth – merely
because they do not possess wealth and without regard to the way we think
private security ought to be distributed’. The problem in implementing this
principle, however, is that the distribution of private insecurity is intimately
connected to the distribution of private security. That is, the strategies which
are adopted by private actors to provide security for themselves usually at best
216 Tim Hope

do not benefit others less protected than themselves and, at worst, reinforce the
latter’s insecurity.
The collective consequences of crime risk avoidance in risk society is the
creation of risk pools dictated according to one’s social position, either by
choice or fate. The collapse of the state’s universal insurance policy against
crime victimisation risk (Garland 1996) may have induced a dynamic which
encourages the clubbing of private security (Hope 2000). As Jordan (1996) argues
with respect to social security in general, the logic of collective action – albeit
the unintended consequence of individualised strategies of risk pooling and risk
avoidance – may act to reinforce social exclusion – the transformation of indi-
vidual disadvantage into an ascribed social position. In a great many respects, in
risk society the threat of crime victimisation is not just a consequence of social
exclusion but also a contributory cause. The challenge facing government
remains whether and how to intervene – not just to reduce risk but also how to
distribute fairly the means for dealing with risk.

References
Barr, R. and Pease, K. (1990) ‘Crime Placement, Displacement and Deflection’, in M.
Tonry and N. Morris (eds) Crime and Justice, volume 12, Chicago: University of
Chicago Press.
Baumgartner, M. P. (1988) The Moral Order of a Suburb, New York: Oxford University
Press.
Beck, U. (1992) Risk Society, London: Sage.
Bourdieu, P. (1998) Practical Reason, Cambridge: Polity Press.
Bottoms, A. E. and Wiles, P. (1997) ‘Environmental Criminology’, in M. Maguire, R.
Morgan and R. Reiner (1997) The Oxford Handbook of Criminology, 2nd edition,
Oxford: Clarendon Press.
Braithwaite, J. (1979) Inequality, Crime and Public Policy, London: Routledge and Kegan
Paul.
Budd, T. (1999) Burglary of Domestic Dwellings: Findings from the British Crime Survey,
Home Office Statistical Bulletin, no. 4, London: Home Office.
Cantor, D. and Land, K. C. (1985) ‘Unemployment and Crime Rates in the Post-World
War II United States: A Theoretical and Empirical Analysis’, American Sociological
Review, 50: 317–23.
Cohen, L. E. and Felson, M. (1979) ‘Social Change and Crime Rate Trends: A Routine
Activities Approach’, American Sociological Review, 44: 588–608.
Coleman, J. S. (1990) Foundations of Social Theory, Cambridge, MA: Belknap Press.
Davis, M. (1990) City of Quartz, London: Verso.
Ellingworth, D., Hope, T., Osborn, D. R., Trickett, A. and Pease, K. (1997) ‘Prior Victimi-
sation and Crime Risk’, Journal of Crime Prevention and Risk Management, 2: 201–14.
Ewald, U. (2000) ‘Criminal Victimization and Social Adaptation in Modernity: Fear of
Crime and Risk Perception in the New Germany’, in T. Hope and R. Sparks (eds)
Crime, Risk and Insecurity, London: Routledge.
Feeley, M. and Simon, J. (1993) ‘The New Penology: Notes on the Emerging Strategy of
Corrections and its Implications’, Criminology, 30: 449–74.
Felson, M. (1998) Crime and Everyday Life, 2nd edition, Thousand Oaks, CA: Pine Forge
Press.
 Crime victimisation and inequality 217

Field, S. and Hope, T. (1990) ‘Economics, the Consumer and Under-Provision in Crime
Prevention’, in R. Morgan (ed.) Policing, Organised Crime and Crime Prevention,
British Criminology Conference 1989, volume 4, Bristol: Bristol Centre for Criminal
Justice.
Garland, D. (1996) ‘The Limits of the Sovereign State: Strategies of Crime Control in
Contemporary Society’, British Journal of Criminology, 36: 445–71.
Giddens, A.(1984) The Constitution of Society, Cambridge: Polity Press.
—— (1990) The Consequences of Modernity, Cambridge: Polity Press.
—— (1998) The Third Way, Cambridge: Polity Press.
Granovetter, M. S. (1973) ‘The Strength of Weak Ties’, American Journal of Sociology,
78: 1360–80.
Hirsch, F. (1977) Social Limits to Growth, London: Routledge.
Hirschfield, A. and Bowers, K. J. (1997) ‘The Development of a Social, Demographic
and Land Use Profiler for Areas of High Crime’, British Journal of Criminology, 37:
103–20.
Home Office (1999) The Government’s Crime Reduction Strategy, London: Home Office
Communication Directorate.
Hope, T. (1984) ‘Building Design and Burglary’, in R. Clarke and T. Hope (eds) Coping
with Burglary: Research Perspectives in Policy, Boston, MA: Kluwer Nijhoff.
—— (1995) ‘Community Crime Prevention’, in M. Tonry and D. P. Farrington (eds)
Building a Safer Society: Strategic Approaches to Crime Prevention, Crime and Justice,
volume 19, Chicago: University of Chicago Press.
—— (1996) ‘Communities, Crime and Inequality in England and Wales’, in T. Bennett
(ed.) Preventing Crime and Disorder: Targeting Strategies and Responsibilities, Cambridge:
Institute of Criminology.
—— (1997) ‘Inequality and the Future of Community Crime Prevention’, in S. P. Lab
(ed.) Crime Prevention at a Crossroads, American Academy of Criminal Justice
Sciences Monograph Series, Cincinnati, Ohio: Anderson Publishing.
—— (1999) ‘Privatopia on Trial? Property Guardianship in the Suburbs’, in K. Painter
and N. Tilley (eds) Surveillance of Public Space, Crime Prevention Studies, volume 10,
Monsey, NY: Criminal Justice Press.
—— (2000) ‘Inequality and the Clubbing of Private Security’, in T. Hope and R. Sparks
(eds) Crime, Risk and Insecurity, London: Routledge.
—— and R. Sparks (2000) (eds) Crime, Risk and Insecurity, London: Routledge.
Jordan, B. (1996) A Theory of Poverty and Social Exclusion, Cambridge: Polity Press.
—— (1998) The New Politics of Welfare, London: Sage.
JRF (1995) Joseph Rowntree Foundation Inquiry into Income and Wealth, volumes 1 and 2,
York: Joseph Rowntree Foundation.
Lash, S. and Urry, J. (1994) Economies of Signs and Space, London: Sage.
Litton, R. A. (1982) ‘Crime Prevention and Insurance’, Howard Journal, 21: 6–22.
Reprinted in P. O’Malley (1998) (ed.) Crime and the Risk Society, Aldershot:
Ashgate/Dartmouth.
Loader, I. (1997) ‘Thinking Normatively About Private Security’, Journal of Law and
Society, 24: 377–94.
——, Girling, E. and Sparks, R. (2000) ‘After Success? Anxieties of Affluence in an
English Village’, in T. Hope and R. Sparks (eds) Crime, Risk and Insecurity, London:
Routledge.
Logan, J. H. and Molotch, H. (1987) Urban Fortunes: The Political Economy of Place,
Berkeley: University of California Press.
218 Tim Hope

Maguire, M. (1982) Burglary in a Dwelling, London: Heinemann.

Miethe, T. D and Meier, R. F. (1994) Crime and its Social Context, Albany, NY: SUNY
Press.
Mooney, J. (1998) Gender, Violence and the Social Order, London: Macmillan.
O’Malley, P. (1992) ‘Risk, Power and Crime Prevention’, Economy and Society, 21:
252–75. Reprinted in P. O’Malley (1998) (ed.) Crime and the Risk Society, Aldershot:
Ashgate/Dartmouth.
Osborn, D. R. and Tseloni, A. (1998) ‘The Distribution of Household Property Crimes’,
Journal of Quantitative Criminology, 14: 307–30.
Pantazis, C. and Gordon, D. (1999) ‘Are Crime and Fear of Crime More Likely to be
Experienced by the Poor?’, in D. Dorling and S. Simpson (eds) Statistics in Society,
London: Arnold.
Savage, M., Barlow, J., Dickens, P. and Fielding, T. (1992) Property, Bureaucracy and
Culture, London: Routledge.
Simon, J. (1987) ‘The Emergence of a Risk Society: Insurance, Law and the State’,
Socialist Review, 95: 93–108.
Skogan, W. G. (1988) ‘Community Organisations and Crime’, in M. Tonry and N.
Morris (eds) Crime and Justice: A Review of Research, volume 10, Chicago: University
of Chicago Press.
—— and Maxfield, M. G. (1981) Coping with Crime, Beverly Hills: Sage.
Smith, D. and Stewart, J. (1997) ‘Probation and Social Exclusion’, Social Policy and
Administration, 31: 96–115.
Stanko, E. A. (2000) ‘Victims R US: The Life History of “Fear of Crime” and the Politi-
cization of Violence’, in T. Hope and R. Sparks (eds) Crime, Risk and Insecurity,
London: Routledge.
Sullivan, M. L. (1989) Getting Paid: Youth Crime and Work in the Inner City, Ithaca, NY:
Cornell University Press.
Taylor, I., Walton, P. and Young, J. (1973) The New Criminology: For a Social Theory of
Deviance, London: Routledge and Kegan Paul.
Trickett, A., Osborn, D., Seymour, J. and Pease, K. (1992) ‘What is Different About
High Crime Areas?’, British Journal of Criminology, 32: 81–90.
Waldron, J. (1995) ‘Money and Complex Equality’, in D. Miller and M. Walzer (eds)
Pluralism, Justice and Equality, New York: Oxford University Press.
Walzer, M. (1983) Spheres of Justice: A Defence of Pluralism and Equality, Oxford: Black-
well.
Waquant, L. J. D. (1995) ‘The Ghetto, the State and the New Capitalist Economy’, in P.
Kasinitz (ed.) Metropolis: Centre and Symbol of Our Times, London: Macmillan.
Wikstrom, P. -O. H. (1998) ‘Communities and Crime’, in M. Tonry (ed.) Handbook of
Crime and Punishment, New York: Oxford University Press.
Wilson, J. Q and Herrnstein, R. (1985) Crime and Human Nature, New York: Simon and
Schuster.
Wilson, W. J. (1987) The Truly Disadvantaged: The Inner City, the Underclass and Public
Policy, Chicago: University of Chicago Press.
Winchester, S. and Jackson, H. (1982) Residential Burglary, Home Office Research Study.
London: HMSO.
Young, J. (1999) The Exclusive Society, London: Sage.
9 Distributive justice and crime
Paul Wiles and Ken Pease

Distributive justice concerns the dispensation of benefits and services to people,

and the tax and other burdens which make that possible. It lies at the core of
most political values. In the literature on distributive justice, the central
notions include justifications of deviations from equality, means whereby such
deviations may be understood, and how fair distribution may be achieved and
recognised once achieved. Much sophisticated work has been carried out within
this tradition. Notions of justice as impartiality, mutual advantage, primary
social goods, ‘offensive tastes’ and fetishist handicaps (Sen 1980; Nussbaum and
Sen 1993) feature in a diverse, complex and important literature. The levels of
discord among leading scholars reflect both the complexity of the area and the
passion which scholars bring to it. This chapter takes up a few of the ideas of
leading thinkers, notably Rawls’ (1971) difference principle and Walzer’s (1983)
elaboration of ‘spheres of justice’ thinking, to suggest some ways in which crime
policy could be shaped with distributive concerns in mind.
The questing criminologist, having become somewhat familiar with (not to
say confused by) the leading theories, will seek examples which speak directly
to crime issues. These are not plentiful. S/he will find the distributive justice
literature to be full of examples featuring goods like education, income and
power. Crime tends not to feature explicitly, even as its inverse, safety. Roemer
(1996: 64) comments that: ‘If one idea must be singled out as the most promi-
nent in contemporary theories of distributive justice, it is that personal
responsibility justifiably restricts the degree of outcome equality’. However, in
the literature, discussion of benefits and services are typically distanced from
the actions (such as crime) of those amongst whom the distribution is to occur,
although discussion of pre-existing attributes (such as ability) is plentiful.
Thus, while personal responsibility is mentioned as a reason for fairly permit-
ting departure from equality, it is difficult to make explicit links between this
and crime policy.
Although there is little in the literature we have read which speaks directly
to the distribution of crime, much can be adapted to that purpose. In this
chapter, we will:

1 Describe some of the ideas which can be put to work by criminologists.

220 Paul Wiles and Ken Pease

2 Discuss the linkage between distributive justice and its sub-types, retribu-
tive and restorative justice, and how this linkage has been invoked in crime
policy.
3 Outline the current distribution of crime opportunities, and conclude that
their extreme inequality, and linkage with other social indices, suggests
distributive unfairness.
4 Set out some policy suggestions which would have the effect of remedying
such unfairness.

Key ideas in distributive justice

Perhaps the best-known heuristic device in work on distributive justice is that
used by John Rawls (1971). In this, the fair solution is the one which would be
chosen from behind a veil of ignorance by rational people knowing nothing of
their own situation and confronting an abstract set of goods. One of the princi-
ples which Rawls deemed would apply from behind the veil of ignorance – the
difference principle – is that inequalities are justified only if they are designed to
bring the greatest benefit to the least advantaged social grouping. Dworkin
(1981a and b) also used the veil of ignorance device. His concern was rightly to
identify the thing to be equalised (equalisandum) for an egalitarian theory of
justice. He distinguished transferable resources, like wealth and power, from
non-transferable resources, like talent. He argued that the just solution would
involve equalisation of transferable resources as well as the kind of insurance
arrangements against deficits in non-transferable resources, which would be
taken out by someone knowing their preferences but not their non-transferable
resources.
Many philosophers have been sceptical of a single distributively fair solution,
and concentrate instead on the permeability of advantage conferred by the
possession of goods of certain types. This perspective is known as ‘complex
equality’ and Pascal (1658/1961) is cited as its first clear proponent. Its essence
is stated by Walzer (1983) thus:

The regime of complex equality is the opposite of tyranny. It establishes a

set of relationships such that domination is impossible. In formal terms,
complex equality means that no citizen’s standing in one sphere or with
regard to one social good can be undercut by his standing in some other
sphere, with regard to some other good. Thus, citizen X may be chosen over
citizen Y for political office, and then the two of them will be unequal in
the sphere of politics. But they will not be unequal generally so long as X’s
office gives him no advantages over Y in any other sphere – superior
medical care, access to better schools for his children, entrepreneurial
opportunities and so on. So long as office is not a dominant good, is not
generally convertible, office holders will stand, or at least can stand, in a
relation of inequality to the men and women they govern.
(Walzer 1983: 20)
 Distributive justice and crime 221

The notion of parallel ‘spheres of justice’ has been elaborated and applied to
specific policy choices by Elster (1989; 1992) and Young (1994). The same
notion is of potential importance to a resolution of the opposition taken by
Fraser (1997) to exist between justices of recognition and redistribution. In
brief, complex equality permits the elaboration of a differentiated view of what
should be equalised, and specifies relationships between status of different types.
Recognition is achieved when the attribute recognised carries no penalty in
other spheres, when for example gender carries no cost in violent victimisation
and pension entitlements. Young’s (1990) first attribute of ideal city life, ‘social
differentiation without exclusion’, describes a state of independence of status in
different spheres.
Fraser (1997: 13) makes a clear distinction between justices of redistribution
and recognition. The first is socio-economic injustice, ‘which is rooted in the
political-economic structure of society’. Examples include exploitation;
economic marginalisation and deprivation. This is contrasted with cultural or
symbolic injustice. Here injustice is rooted in social patterns of representation,
interpretation and communication. Examples include cultural domination,
non-recognition and disrespect. The problem of the Marxian working class
instances the first type of injustice. That of the gay community instances the
second. Fraser points out that the first kind of injustice is remedied by denying
the specificity of the disadvantaged group. The second kind of injustice is reme-
died by asserting its specificity:

The upshot is that the politics of recognition and the politics of redistribu-
tion often appear to have mutually contradictory aims. Whereas the first
tends to promote group differentiation, the second tends to undermine it.
Thus, the two kinds of claim stand in tension with one another; they can
interfere with, or even work against, each other.
(Fraser 1997: 16)

The central distinction made by Fraser (and the elaborated view by Young
[1990; 1999]) must either shape the whole of this chapter, or be reconciled with
another approach. We believe that it is readily reconcilable with a ‘spheres of
justice’ approach to redistribution. Complex equality is achieved according to
Walzer when status is not transferable across spheres. The formulation of sphere
is the key to reconciliation with Fraser. In a differentiated society, spheres
proliferate. People occupy spheres by race, sexuality, profession and sporting
club allegiance. Walzer’s approach requires that status is not correlated across
spheres. Thus sexuality must not be translatable into professional success.
Attempts to do so are unfair (indeed, actionable under employment legisla-
tion). Thus a group suffering from injustices of recognition acts not by
de-differentiation but by asserting and seeking to establish the de facto separa-
tion of spheres. This approach requires similar action taken in respect of all
those in a disadvantaged set, not ad hoc action in respect of those who are in
the intersecting set of two forms of disadvantage. For example, it may involve
222 Paul Wiles and Ken Pease

remediation of disadvantage incidental to disability, and remediation of disad-

vantage incidental to a minority sexual orientation, but not towards the
disabled of minority sexual orientation. This becomes crucial (and familiar)
when we think of, say, the dyslexic offender. Action may not be taken in
respect of the dyslexic offender because it is unfair to dyslexic non-offenders
and non-dyslexic offenders. In this way it resolves the ‘goodies for baddies’
dilemma in criminal justice, described later in this chapter.
Fraser’s approach to achieving both distributive justice and justice of recog-
nition, which she collectively terms ‘transformation remedies’, proceeds by
destabilising group differentiation, for example in the case of sexuality by
nuancing general notions of sexuality. This is translatable into Walzer’s formula-
tion by seeking to change perceptions of spheres of justice, perhaps to the point
at which a human characteristic is no longer seen usefully to constitute a viable
sphere of justice, and be relegated to an incidental attribute. Fraser’s work
certainly helps one think in terms of the repertoire of spheres of justice as
culture and time limited, and that perceptions of justice will depend upon
consistency in distinguishing between human attributes which are salient in
judging fairness, and those which are incidental. In sum, Fraser’s distinction is
seen as helpful in advancing the ‘spheres of justice’ argument and as being
entirely consistent with it. Thus, an informing theme throughout the chapter
will be spheres of justice, rather than the central distinction between justices of
distribution and recognition, advocated by Fraser.

Crime contingent distributive justice

Before considering the implications of ‘spheres of justice’ for crime, we must first
establish the relationship between distributive justice and the evocations of
justice familiar in the criminological literature. It will be concluded that
retributive and restorative justice, together with victim support and compensa-
tion schemes, are sub-types of distributive justice whose uniting feature is that
they are triggered by the crime event.
Criminology takes as its starting point the crime event. Much work has
concerned itself with social differences in the probability of an act being classi-
fied as criminal, and retribution ensuing. Race has been the social
categorisation featuring in much of this work (see, for example, Hood 1992;
von Hirsch and Roberts 1997). Retributive justice involves a consideration of
the expected probability and degree of sanction consequent on a criminal act
and the scaling of penalties in proportion to harm inflicted. This is achieved
(however imperfectly) by the specification of maximum allowable penalties and
the maintenance of proportionality between crime seriousness and punishment
below that maximum. By disadvantaging offenders, it attempts to offset gains
achieved by the commission of crime. Restorative justice involves negotiation
between the offender and victim to repair and rearrange matters, economically
and/or psychologically, between them. Both retribution and restitution seek,
however crudely, to restore an approximation of the status quo ante of the distri-
 Distributive justice and crime 223

bution of resources as between victim and offender. Victim support and crim-
inal injuries compensation also represent attempts to promote distributive
justice and this remains possible even when an offender’s identity is not known.
Retribution, restoration, victim compensation/support and distribution as
modes of justice are linked through their attempts to establish a balance of
effort and advantage between citizens. Viewed thus, retribution, restoration and
compensation are simply means of moving towards distributive justice in the
wake of an offence. Distributive justice can be seen as the overarching notion,
with retribution, restoration and compensation becoming relevant (in the
service of promoting distributive justice) once a crime has been committed. The
debates on race and sentencing note that the formulation of the criminal law
itself can generate patterns of ethnicity among offenders, on the basis of differ-
ential criminalisation of lifestyles. This well illustrates the limits of distributive
justice achievable through the operation of a criminal justice system.
Classifying retribution as a form of distributive justice seems contentious. It
could be argued that it is precisely the appeal of these responses that they do not
involve any redistribution of resources. But the sanctions of criminal justice do
involve a redistribution of resources. A two-year prison sentence delays the
development of a career and family life by two years, in comparison with imme-
diate colleagues. An eye for an eye redistributes visual impairment. A fine
redistributes disposable income. The problem appears to us to be a consequence
of equating the search for distributive justice with policy choices which improve
the lot of people. Criminal justice is perhaps unique in seeking distributive
justice by deliberately disadvantaging people, but it is no less redistributive for
that. One can as well equalise height by chopping off the legs of the tall as by
good nutrition of the short.
In what follows, the distinction will be made between initial distributive
justice and contingent distributive justice. The former concerns allocation
across society. The second concerns re-allocations, which follow disruptions of
distribution occasioned by an offence. It should be stressed that the distinction
concerns the circumstances of application of fair distribution, not the point at
which fairness is to be decided. Fairness should be determined in advance. Only
its application is contingent. The principles of retributive justice are general.
Their application is contingent on case facts.

Tough on crime, tough on the causes of crime

Because it seems implicit in much thinking about crime, we should explicitly
reject the notion that addressing material inequality directly is a sensible way of
reducing crime. In a spheres of justice approach, we should strive to make them
independent. Some of the key relationships are counter-intuitive, and the
research on the penal effectiveness of providing benefits to offenders does not
inspire much hope. In Field’s (1991) analysis, violence appears linked with
affluence. Property crime is linked inversely with GDP/head, which is not a
measure of distribution. Moreover, the association is direct rather than delayed,
224 Paul Wiles and Ken Pease

so must therefore be more a function of anticipated than current poverty. The

link between violent victimisation and wealth goes against the notion that the
poor are most victimised by violence (Skogan 1991). In the analysis of the
British Crime Survey, it is the better-off households within the worst areas
which are most prone to victimisation (Trickett et al. 1995). Improving the job
prospects of offenders by training or subsidised employment is not one of the
favoured approaches identified by the major review of what works in crime
reduction – although there are isolated instances of success (Sherman et al.
1997). Crime and deprivation do co-vary at the areal level, but the relationship
is imperfect. The simple presumption that economic inequality causes crime is
seductive but far too simple. Remedying material inequality is not the royal road
to crime reduction.
While remedying material inequality would not be the finding of an
evidence-led approach to crime reduction, it is one which seems politically
ingrained in parties of the Left. For this reason we will, for the moment, take
the debate on these terms. What is the policy link between unfairness in mate-
rial distribution and the appropriate response to crimes committed by those
disadvantaged in the initial shake-out? This link is central to the New Labour
mantra heading this section: ‘Tough on crime, tough on the causes of crime’.
Let us assume causes of crime are taken to include early material and social
disadvantage (but, as noted above, the reality is complex). If this constitutes
distributive unfairness to the pre-delinquent child, being tough on the causes of
crime means reducing material and social disadvantage. If being tough on crime
means being tough on the criminal (it can mean nothing else), one is tough on
offenders even though distributive unfairness led to their offending. If one is
tough on the (presumed) causes of crime, it becomes unnecessary to be tough
on offenders. If people offend because of their exposure to the causes of crime, it
is unfair to be tough on them. So being tough on crime and tough on the causes
of crime at the same time is cruel or dishonest. The problem can perhaps be
resolved by being tough on the causes of crime now and later becoming tough
on offenders who wilfully offend despite the removal of the circumstances
which may contribute to crime.
In practice, the interactive link between toughness towards criminals and
toughness towards the causes of criminality, between distributive justice and its
sub-types, retribution and restoration, is confused, and most of what is contro-
versial about criminal justice stems from that confusion. The fact that the New
Labour mantra has survived without frequent identification of its central
tension is itself remarkable.
The nub of the confusion stems from the assumption that if we had delivered
distributive justice to someone before he or she embarked upon a criminal
career, we would not need to dispense retributive or restorative justice now. For
example, many pre-sentence reports and most pleas in mitigation attribute
offenders’ misdeeds to their failure to have received a fair deal initially. If they
had received a fairer share of love, abuse, education, training or the like, they
would not be in the dock. When sentences are deemed unduly lenient, this is
 Distributive justice and crime 225

often because initial distributive injustice has been taken into account. When
sentences are deemed unduly harsh, this is often because of a failure to take
initial injustice into account. Because these are matters of judgement, the same
sentence may be judged both too severe and too lenient. This is typically the
case where women have killed abusive partners. Also, probation officers and
social workers invite criticism from the press when they use the decisions of
retributive justice as a springboard in the attempt to remedy actual or perceived
distributive injustice. These hoped-for remedies dispense ‘goodies for baddies’
where holidays, driver education and the like are offered to the errant and
denied to the righteous.
In a society where no one is deprived by initial distribution, retribution (at
least towards adults) could be dispensed without any angst on the part of the
punisher,1 and ‘goodies for baddies’ would be perverse. In the real world, where
systemic or individual distributive injustice is presumed to intrude, and the
language of the court conspires in that presumption, all criminal justice
personnel have to make judgement calls about how the degree of a sanction must
be adjusted in individual cases. Arguments about juvenile justice and means of
dealing with mentally disordered offenders are difficult because of differing
presumptions about the proper scope of remediation of past distributive injustice.
This is not to say that offenders who have experienced initial distributive
justice (in so far as society can dispense it) are easy to sentence. Initial injustice
starts at conception. Are genes distributed in a just way? Are we the victim of
injustice if we have genes making for risk-taking and shortness of temper? Is
short stature, and consequent feelings of inferiority and tendency to violence,
the product of parental smoking and poor nutrition? The Dworkin formulation
described above, and as developed by Arneson (1989) and Cohen (1989), offers
at least a way of thinking about these issues. The central shift is from welfare
equalisation to resource equalisation. Since resources include native talents,
compensation is necessary for inequalities in resources for which individuals are
not responsible. This requires a sense of community or fraternity, and requires
the talented not to be self-regarding.
In sum, the language, pronouncements and decisions surrounding caution,
prosecution, sentence imposition, sentence administration and release – in fact,
all decisions dealing with criminal offenders – become difficult in so far as issues
of initial distributive justice are taken to interact with contingent justice.
Apologetic retribution follows when punitive intervention is moderated by
reason of presumed initial distributive injustice. Distribution blindness occurs
when punitive intervention is greater by the neglect of such injustice.
Apologetic retribution and distribution blindness are the Scylla and Charybdis
of offence contingent justice.
In the foregoing, there was too easy a transition from discussions of contigent
justice to concerns about what to do with offenders. Leslie Wilkins (1999),
instancing his work in preventing aircraft accidents, repeatedly bemoaned the
reduction of the problem of crime (paralleling the crash) to the problem of the
offender (paralleling the pilot or ground crew). Historically, offence contingent
226 Paul Wiles and Ken Pease

justice has been centred on the offender. Sanctions are generally tempered or
wholly by-passed on the basis of the offender’s failure to receive distributive fair-
ness in the past. Offender-oriented crime reduction methods are based upon
belated attempts to afford distributive justice to perpetrators. This translates
into those applications for crime prevention funding providing facilities for
offenders. It is interesting that all long-term person-oriented crime prevention
programmes are focused on offenders, or those at risk of becoming offenders,
rather than victims or those for whom victimisation is a future hazard.2 Even if
effective, this assigns to crime victims a secondary role, whose future protection
is dependent upon the response of perpetrators to efforts made on their behalf.
The fact of a crime leads to two sets of responses, one through the detection and
processing of putative offenders, the other to direct protection of the person and
property of the victim. Distributive justice is perceived to be important for the
offender and to be achieved to benefit the victim only through the achievement
of change to the perpetrator.
Is it a fact of nature that the allocation of long-term person-oriented crime
prevention has to be centred upon the perpetrator rather than the victim? It is
not. The sad fact is that we know little of how victim careers work. We know a
great deal about how offender careers work. Victimisation surveys such as the
British Crime Survey (see, for example, Hales and Stratford 1997) are cross-
sectional, and do not allow us to view the trajectory of crime victimisation
across a lifetime. Just as notions of fairness in criminology have been over-
whelmingly discussed in relation to the offender, so the inequities of crime
distribution from the victim perspective have been neglected. Farrell et al.
(2001) propose an accelerated longitudinal design using data from the US
National Victimisation Survey, to yield some early answers.3 The introduction
of a longitudinal element into the British Crime Survey will in due course show
something about how victim careers develop, but the modest extent of the
longitudinal element means that it will be many years before a picture emerges
to rival the proposed accelerated longitudinal design from the American data.
The fact that we are having to play catch-up on a topic of such importance is
astonishing.

Distributive justice and crime victimisation

Perhaps the first problem in recognising victimisation as an issue in distributive
justice is linguistic. One talks about inequalities in health and education rather
than illness and illiteracy/innumeracy. This is because it is the protective and
remedial services against illness and illiteracy whose distribution is at issue. To
be comparable, should one cast the victimisation debates in terms of safety
rather than risk? The objection may be raised that we cannot come close to
guaranteeing safety. However, the same is true in the health arena. Preventive
and appropriate remedial treatment can reduce the risks of illness, but some
illnesses are intractable, and in many cases, citizen lifestyle may offset the
health benefits of preventive measures deployed.
 Distributive justice and crime 227

A second problem, more apparent than real, concerns the wilfulness of

actions taken to make citizens unsafe – actions taken by the criminally inclined
and by those who carelessly dispense crime opportunities as part of their busi-
ness. Should such actions be excluded from the range of convenience of
distributive justice? Why is there no just distribution of directed malice?
Safety has been chosen as the currency of choice only to be rejected. This is
because it focuses primarily on the potential victim. One can think more gener-
ally about the distribution of crime opportunities. This has the effect of
simultaneously regulating victim safety and offender temptation. By choosing
crime opportunities, one recognises that it may be as unfair to present someone
with a host of criminal opportunities by which to be tempted, as to allow a
person or situation to be chronically victimised by those who succumb to that
temptation. In short, by choosing the supply of opportunities over safety one
considers the welfare of both potential victim and potential offender at the
same time.
Considering the distribution of crime opportunities in terms of the well-
being of both offenders and victims creates new problems. They should be
addressed, but doing so is beyond the scope of this chapter. Probably all human
action has an ethical ‘range of convenience’. When a society is grossly unfair,
the removal of temptations is not defensible. ‘Give us this day our daily bread’
in the Lord’s Prayer comes before ‘lead us not into temptation’. Target hard-
ening the baker’s shop is not acceptable in a city of starving and penniless
people. Target hardening stores selling electronic goods in affluent societies is
acceptable. Where is the crossover point? This is not an argument for omitting
temptations and delivery from evil from the list of variables to which distribu-
tive justice is relevant, rather a balancing of the harms inflicted upon the
starving, those who lust after electronic gizmos, PC World and bakers. This is
no different from organ transplantation, where insensitive harvesting of the
organs from the dead causes their loved ones suffering which has to be set
against the suffering involved in repeated dialysis or slow death.
A second apparent problem lies in the interpretation of diminishing crime
opportunities as diminished freedom of action. In particular, CCTV is
demonised as the sign of the surveillance society in headlong pursuit of crime
reduction. In fact, more sophisticated crime reduction measures, such as
biometric access control (e.g., your computer logs on in response to your finger-
print on the screen) and location-aware electronic goods (which cease to work
when they are moved from where they ‘ought’ to be), do not reduce the non-
criminal options available to a citizen. How one engineers a society which
favours such options over cruder forms of opportunity reduction is a debate in
itself. For the moment, it should be asserted that controlling the supply of
opportunities will not necessarily reduce freedom of action generally.
Before addressing the fairness of current distributions of victimisation by
crime, it may be appropriate to summarise the argument to this point. First, it
has been suggested that retribution, compensation and restoration are sub-types
of distributive justice applied crudely in the wake of an offence to restore the
228 Paul Wiles and Ken Pease

balance of effort and advantage between citizens. Second, some of the most
difficult issues in criminal justice come with the interaction between retribu-
tive/restorative justice and presumptions about the causal role of distributive
unfairness in causing offending. Third, while notions of fairness to offenders are
well theorised, no similar development has taken place with respect to crime
victims. This is reflected in the absence of research on victim careers, by
contrast with the extensive research on criminal careers; and finally, thinking
about the distributive fairness of crime opportunities (in preference to the fair
distribution of safety or risk) incorporates both victims and offenders as people
to whom justice must be done.

Is the distribution of crime fair?

Crime is distributed very unequally across even small areas, and across people
within areas with similar aggregate levels of crime (Farrell et al. 1996; Hope
1996; Pease 1998). One of the major developments in criminology over the last
two decades has been the recognition of these facts and their implications. This
has occurred by a number of routes, including the recognition of crimes whose
rate of repetition is huge. These include domestic violence, bullying and racial
harassment. The smallest geographic unit to be found in the standard volumes
on crime, the annual Criminal Statistics, will be seen to be the police force area.
Adding up crimes and expressing them as a total for a command unit or police
force area obscures how crime experience varies street by street, home by home
and pub by pub. The oldest joke about statistics is that if you have got your head
in the oven and your feet in the freezer, on average you will be comfortable.
Combining quiet areas like Bramhall and difficult ones like Wythenshawe to
produce an overall figure for Greater Manchester Police and expecting the
resulting total to be meaningful makes no more sense for the corporate body of
the police than the oven-freezer predicament does for the human body.
Furthermore, the fact that the numbers are routinely expressed in this way
shows an apparent indifference to the distribution of suffering which they
entail.
Figure 9.1 shows the total number of crimes experienced in each sampling
unit of the British Crime Survey. Readers will be aware that this survey reports
an enquiry, which now takes place every year, of crime, both reported and unre-
ported to the police, in England and Wales. The data in Figure 9.1 come from
the 1988 Survey and refer to property crime, but the picture is much the same
in different years and for crime against the person too. The survey employed
some 600 sampling points, at each of which some thirty people were inter-
viewed. The 600 areas were then divided into the 10 per cent of areas with the
most crime, the 10 per cent with the next most crime, and so on, up to the 10
per cent with least crime. Figure 9.1 simply takes the total number of crimes
captured in the 10 per cent (decile) of areas with most crime, the 10 per cent
with next most, and so on, and makes evident the ratios between these areas
(i.e., how many times more crime is suffered by the worst than the best areas).
 Distributive justice and crime 229

Figure 9.1 Property crime by area crime levels

Source: 1988 British Crime Survey

The shape of the curve shows that the 10 per cent of areas with most crime has
more than double the amount of crime in the next most crime prone 10 per
cent of areas. Looking at the worst 10 per cent of areas against the best 10 per
cent, the worst 10 per cent suffer forty-three times as much crime as the least
victimised 10 per cent. Another way of saying this is that some 40 per cent of
all crime experienced happens in the 10 per cent most crime prone areas. Well
over half occurs in the 20 per cent most crime prone areas. Calls logged from
the public will also show a massive range of crime densities.
The 10 per cent of areas with most crime can get to be that way via one or
both of two different routes. First, increasing proportions of people living in the
area may become victims, so that perhaps 80 or 90 per cent of people in the
worst areas suffer some crime. Alternatively, the same proportion of people
could become victims of crime in the worst and best areas, but those victimised
in the worst areas would suffer crime more often. The old and unfunny joke is
that in England and Wales somebody is getting robbed every thirty minutes –
and she’s getting sick of it. Unfunny as it is, it shows that more people do not
necessarily have to fall victim to crime for the crime rate to rise.
Figure 9.2 shows the proportion of people victimised in each area decile.
Thus in decile 10 (the 10 per cent of areas with most crime) some 28 per cent
of people are victims of crime over the year. The proportion increases steadily to
230 Paul Wiles and Ken Pease

Figure 9.2 Property crime by area victim prevalence

Source: 1988 British Crime Survey

that figure from 3 per cent in the least crime-ridden areas. The point to stress,
however, is that even if you live in one of 10 per cent most crime prone areas,
you have a 72 per cent chance of not being a victim over a one-year period.
However, Figure 9.3 shows the number of crimes which people who are victims
each suffer. It will be seen that in the worst areas, each victim of property crime
suffers an average of nearly five.
The same pattern is even more marked for crimes of violence, where 15 per
cent of people in the worst areas suffer crimes against the person, but each of
those who do suffer crime has to endure six crimes, on average. Most people in
the worst areas do not suffer crime over the course of a year. Those who do,
suffer it several times each. Since areas which suffer much property crime are by
and large also those which suffer most personal crime, looking at all crime
combined yields an even more marked difference between areas in levels of
chronic victimisation.4
One should not rely on the precision of the numbers presented to measure
crime concentration. There are conventions in the British Crime Survey
concerning the maximum number of victimisation forms which a victim may
complete, the maximum number of events recordable as a series against a
chronic victim, and the like. Such conventions change the exact numbers
arrived at. Neither should one stereotype all crime victims as the innocent and
put upon. As yet unpublished research by the first author with Andrew Costello
 Distributive justice and crime 231

Figure 9.3 Property crime by area victim concentration

Source: 1988 British Crime Survey

suggests recent criminal involvement is a characteristic of some repeat victims.5

Insurance fraud is another possibility to explain some repeat victimisation.
However, there is no challenge to the fact that crime is heavily concentrated,
both by area and within area by victim and household. If one wants to address
the problem of crime, repeat victims in high crime areas would be where most
effect is to be had, whatever the dynamics of the problem.
The reason for developing the point about concentration at such length is
that questions about how crime opportunities are distributed is the single fact
about crime that one would wish to know if notions of fairness of that distri-
bution were regarded as a salient feature of crime. Yet such statistics are not to
be found in Criminal Statistics, and have in the past had to be dragged out of
British Crime Survey data after much effort. As criminologists, we have
largely behaved as though indifferent to the concentration of crime opportu-
nities.
Even now, criminologists have paid less attention than they should to risk
differences across non-geographic dimensions. Over the last two decades, much
attention has been given to the development of tools to aid precise spatial loca-
tion of crime. This has not been accompanied by equivalent attention being
given to the social mapping of crime. Even now, the allocation of resource to
high crime areas by the Home Office takes the police Operational Command
Unit (OCU) to be the appropriate unit defining a high crime area. Given what
we know about variation within OCUs, social patterning of crime risks, and the
role of prior victimisation in marking out those at high risk, this seems crass.
Many people and places outside the designated ‘high crime’ OCUs will be at
greater risk than many of those within it. Illustrations of the social patterning of
victimisation opportunities will now be given.
232 Paul Wiles and Ken Pease
Table 9.1 Risks of domestic burglary by household type

Household is: Relative risk

head of household 16–24 2.71
one adult living with children 2.00
head of household is single 1.73
head of household is separated 1.63
respondent is Asian 1.77
head of household is unemployed 1.80
head of household is economically 1.70
inactive
home is privately rented 1.73s
respondent resident for less than one 1.75
year
home has no security measures 2.71
home in inner city 1.52
home in area with high levels of physical 2.14
disorder
Source: calculated from Budd 1999

Table 9.1 is calculated from Budd (1999) and represents the risk of domestic
burglary victimisation in relation to a national average of 1. It will be seen that
certain social groups suffer much more than their ‘fair’ share of burglary, with
young householders suffering some 2.7 times the national rate, the recently
moved some 1.75, and so on.

What gets counted, counts

The primacy of geographic over other forms of classification reflected in
Criminal Statistics defines the scope of what we can regard as distributively
just.6 We cannot dispense justice when we do not know the variables in terms
of which distribution is unfair, however fairness is defined. We cannot know
whether health care is distributed fairly in terms of gender, ethnicity or what-
ever without having breakdowns of health care by these variables. We are
ill-equipped to begin to dispense justice because of the paucity of analyses along
non-geographic lines. Counting new things may be the most important change
one could advocate. Allocating events according to the variables across which
the equalisandum must be equalised may be the most fundamental necessary
condition for achieving fairness.
Even when thinking geographically, the unit of analysis defines the degree of
concentration. Hence even when counting old things (places and people
victimised) we must do so in ways which reflect crime experience, rather than
aggregated by administrative boundaries. As hinted earlier, the smaller the unit
of analysis the higher the degree of concentration. Figure 9.4 shows Lorenz
curves for crime data disaggregated at force and OCU level. The most crime
prone 2.5 per cent of forces accounts for some 8 per cent of crime, whereas the
 Distributive justice and crime 233

Figure 9.4 Inequality at force and OCU level

most crime-prone 2.5 per cent of OCUs account for some 17 per cent of crime.
The difference between the curves remains along their length. The smaller the
unit, the higher the degree of concentration.

Displacement
In the late 1980s, the geographer Robert Barr and one of the writers of this
chapter (Pease) became interested in the terms in which the debate about crime
displacement was being conducted. The debate centred upon whether displace-
ment happened, the extent to which it did so, and the limits on crime
prevention efficacy which that entailed. Absent from the literature was any
suggestion that some distributions of crime were to be preferred to others. Barr
and Pease (1990) expressed the view that:

‘Displacement’ is discussed as a limit on the efficacy of crime prevention. In

the gloom that attends such discussion, it is too readily forgotten that for a
crime to be displaced, it must first be unplaced; in other words, that a crime
at a particular place and time must have been prevented. The word
‘deflected’ gives a better sense of the achievement at the heart of the
process. … [W]e should ask ourselves what kinds of crimes, perpetrated
against whom, would we most like to prevent? Such a question does not
admit of a sensible, unqualified answer.
(Barr and Pease 1990: 280–1)
234 Paul Wiles and Ken Pease

The notion of benign displacement was introduced by Barr and Pease (1990).
Benign displacement is said to exist when the distribution of crime after deflec-
tion is judged to be fairer than its distribution beforehand. Malign displacement
is its opposite. There is a judgement to be made about the fair distribution of
unavoidable crime, a point which the lack of differentiation of the concept of
displacement obscures.
Barr and Pease’s argument has been extended by Barnes (1995). The impor-
tance of the argument lies simply in the recognition that the previous literature
had been written as though ‘a crime is a crime is a crime’, with no apparent
recognition that the way in which the crime burden is shared is matter of policy
concern. To speculate, the literature is as it is because the focus has been on the
administrative unit recording crime rather than the crime victim. Police
commanders can get into trouble for large numbers of crimes. How they are
distributed is not of central relevance. Crimes will be regarded as equal only
when their sole existence are as units to be totted up by the police for onward
transmission to the Home Office for inclusion in Criminal Statistics. It is not as
though crimes are treated that way by front-line police officers, where, for
example, particular effort may be given to preventing burglaries against the
elderly, or against those who may make the most vociferous complaints to the
Chief Constable.7 Although the literature on displacement is naive about crime
distributions, front-line police officers are not.
There are some social changes which may impact upon emerging patterns of
crime concentration. One dimension along which people may vary is that of
information-richness to information-poverty. Informed decisions taken by the
information-rich in relation to where to live, how to protect themselves and
their families, and so on, may well tend to concentrate crime victimisation yet
further.

So what?
By what means may concerns for the fair distribution of crime opportunities be
brought recognisably to bear upon crime control practice? What follows is a
speculative, certainly flawed, but hopefully useful attempt to make some of the
connections.

Fairness feeds on facts

The first policy implications have already been alluded to. For distributive fair-
ness to be achieved, one must know the distribution of crime events across the
social variables of interest. This is true whatever criterion of fairness one
favours. If fairness involves limits on variation in the number of crimes suffered
per place or household, one needs to know the number of crimes per victim.
This is still difficult to determine from police crime recording systems (see
Pease 1998).
Knowing the distribution of crime opportunities (whether reflected in
 Distributive justice and crime 235

crimes committed or crimes suffered by area residents) will not be enough to

act fairly. Difficult decisions will be required as to the extent to which
chosen lifestyle characteristics drive crime rates. This will require the devel-
opment of an approach to crime based upon the Dworkin (1981a and b)
distinction between transferable and non-transferable goods. Perhaps the
development of such an approach should come before the choice of variables
in terms of which crime experience must be known as a necessary condition
of fairness.

Crime and Rawls’ difference principle

Rawls’ difference principle asserts a moral mandate to adopt those institu-
tions which will maximise the amount of primary social goods enjoyed by
those who enjoy the least of them. Restricting our attention to the supply of
criminal opportunities, what would this mean for policy? Policing (as other
public services) currently suffers from a surfeit of performance indicators. The
difference principle might be applied by substituting one performance indi-
cator for many of the others. This would be the ratio of crime in, say, the
worst 25 per cent of households within a force area to crime in the least
crime-prone 25 per cent. That ratio can (self-evidently) be reduced by
increasing crime in the least crime-prone areas, or by reducing it in the most
crime-prone areas. While there would be some limited scope for criming
more of the events coming to police attention in the least crime-prone areas,
the more plausible tactic would be the reduction of crime in the most crime-
ravaged areas. Underpinning this approach is the awareness that the least
crime-prone areas largely limit their own crime, by residents installing protec-
tions to their homes, ferrying their children to places rather than entrusting
them to public transport, mobilising policing effort by persuasive articulation
of their concerns, backed by political influence. If leverage on policing
services is indeed greatest in less crime-prone homes, manipulating the ratios
would perhaps be an elegant way of giving effect to the Rawls’ difference
principle.
This proposal, while radical in its implications, could be readily imple-
mented, albeit in a crude form. A more refined form would be only slightly
more difficult, and would rely on using individual households/premises as the
unit of analysis for most property crime, and individual people for crimes of
violence. Its starting point is the calculation of some form of coefficient of
variability across OCUs, police forces, offence types or whatever. Offence
types are chosen here. While of least direct relevance to the comparison of
policing areas, there was not available data on police areas disaggregated
enough to be meaningful. What is presented here is little more than an indi-
cation of what should be done. The standard form of coefficient of variability
is SD/mean. This is presented for offence types in Table 9.2. The unit of ana-
lysis is the OCU. It will be noted that robbery and violence rates are most
variable across OCUs, with burglary rates least variable. It is thus in terms of
236 Paul Wiles and Ken Pease

Table 9.2 OCU rate variability by offence types

Offence type Coefficient of variability

violence 2.53
sexual offences 1.65
robbery 3.51
burglary 0.62
theft of motor vehicle 1.84
theft from motor vehicle 1.70

violence that citizen experience is most varied according to where they live.
By the Rawls’ difference principle, areas with the highest rates of violence
should be given some priority, since they define the most extreme variation
by area.
It is perhaps worth noting that standard scores (mean/SD) are commonplace
in the criminological literature, whereas the coefficient of variability
(SD/mean) is extremely rare. In microcosm, this eloquently states that crimi-
nology has neglected the measures in terms of which distributive justice is
always conceived.

Combining or uncoupling forms of disadvantage

There are two approaches to considering crime opportunities alongside the
supply of other forms of good or service. One is to incorporate them all into
externally occasioned hazard, as the common currency of distributive justice.
This will be examined next, and followed by some consideration of crime
opportunities within a ‘spheres of justice’ framework.

Hazard integration
This approach has been most fully developed within criminology, as the length
of its exposition here suggests (see also Wiles and Pease 2000). Ironically, it will
find much less favour amongst philosophers of equity. Its central element is to
integrate the consideration and response to diverse externally-caused hazards.
The separation of crime and disorder as kinds of harm should be avoided. This
would require the repeal of the Crime and Disorder Act (1998) and its replace-
ment by an Act which did not distinguish between harms resulting from human
malice and other harms. Current arrangements will lead to, for example, inten-
sive action to reduce pub assaults while more people are being killed on the
roads. Harms should not be divided into criminal and other for the purposes of
prevention. The elevation of crime to a privileged position among harms risks
the ‘criminalisation of social policy’, and the cynical justification of programmes
which may be progressive in other ways by reference to their supposed crime
prevention potential.
 Distributive justice and crime 237

Spheres of justice
Reducing harms to a common currency goes against the recent arguments of
communitarians. Different goods being incommensurable, fairness will be
achieved when status on one dimension is not associated with status on others.
Communities that now suffer high rates of crime are often also characterised by
poor housing, deprivation, exclusion, poor educational achievement, pollution
and the like. This is not coincidence. A wide range of social and physical vari-
ables contribute to and are contributed to by crime risk, both singly and in
interaction (see, for example, Tseloni et al. forthcoming). These variables
include affluence, region, area type, parity, household composition, land use and
physical attributes of a home, car or person. The establishment of causal routes
through which these allied ills manifest themselves is technically difficult and
probably pointless. Social variables may both drive and be driven by crime and
perceptions of crime, or be manifestations of other variables, overt or subtle. For
example, concentration of single parent households may be associated with
crime for many reasons, including the following:

• because crime changes population composition with the more desperate

moving to less desirable areas;
• because the lifestyle of single parents changes natural surveillance and
other area attributes which increase crime; or
• because both single parent households and crime are attributes of an area’s
economic decline, which drives both; or
• because the spending power of lone parents distorts local economies,
leading to the decline or change of the business profile.

The intriguing (even mind-blowing) question is: what would crime policy
look like if its single aim were to uncouple risks from the risks of other types of
disadvantage? How could it be monitored? It would mean that effort at crime
reduction would be concentrated on those who were most crime-prone. This
could most easily be achieved by a major emphasis on the prevention of repeat
victimisation, since repeat victims are disproportionately to be found in areas
and groups suffering most crime (see Trickett et al. 1995; Johnson et al. 1997;
Pease 1998). The emphasis solely on prior victims will lead the reader to ques-
tion, ‘And what else?’ The answer is nothing else. Why not prior victimisation
occasioning high trauma? Because repeats can escalate in trauma, from robbery
to murder, and because the sheer grind of chronic victimisation, even events
which may dispassionately be regarded as individually trivial, is itself traumatic
(see Shaw 2001). Preventing repeat crimes is the most elegant means of
applying Rawls’ difference principle. Not classifying victims further buys
simplicity of implementation.
Prior victims are disproportionately future victims. Any other means of
dispensing crime prevention help will be less efficient in identifying and
helping those at risk. Let us assume that help is dispensed by geographic area.
As noted above, crime is less concentrated the more aggregated an area or
238 Paul Wiles and Ken Pease

group. Thus an areal definition of areas needing crime reduction help will
exclude individuals at high risk and include those at low risk. Dispensing crime
reduction according to prior victimisation is the closest we have to a means of
reducing differences in crime suffering. Recent research (Everson 2000) also
suggests that it denies opportunities to those prolific offenders to whom we most
wish to deny opportunities.
Parallel attention to those who are disadvantaged in other respects will
reduce associations between harms. As stated earlier, this must not be restricted
to the intersecting set of those suffering both harms. When the key associations
approach zero, we have a distributively fair society within a set of independent
spheres of justice (see Walzer 1983), and hopefully the reconciliation of justices
of redistribution and recognition (Fraser 1997). Functionally, this approach and
the required reduction in range of experiences of crime within an area (see
‘Crime and Rawls’ Difference Principle’ above) will tend to the same result, if
the attempt to reduce best-worst ratios as indices of performance are applied
across forms of disadvantage.

Conclusion
Of the ideas presented above, the idea of best-worst ratios in crime as the basis
of performance measurement seems to have immediate promise. The recogni-
tion that statistics centring on variation are rare in criminology, whereas
statistics centring on central tendency are common, felt like an insight. The
crude calculations of OCU and force variability contained in this chapter show
that the technique is easy, the only difficult issue being the collection of the
most appropriate data. Application to more disaggregated data will make the
technique more useful and truer to the Rawls’ difference principle.
Notions of uncoupling measures of disadvantage, taken from Walzer (1983)
and consonant with the ideas of Fraser (1997; 2000), are intriguing in their
implications, and could with profit be elaborated. The details of how the distri-
bution of opportunities from the victim perspective translates into effects on
criminal careers will not be simple. Single offenders commit offences against
many victims, alone or with others. The impact of different tactics of reducing
crime opportunities will likely be different in limiting victim harm and offender
vulnerability to sanction. Past neglect of these issues makes the task a daunting
one, albeit worthwhile.

Notes
The opinions expressed in this chapter are not necessarily those of the Home Office.

1 This is to assume that the ascription of offending to prior distributive injustice is

more than an unfounded assumption.
2 Rape alarms and target hardening of locations prone to victimisation are usually
categorised as situational prevention. The argument of this chapter may require a
reconsideration of crime prevention typologies.
 Distributive justice and crime 239

3 In conventional longitudinal research, one has to wait a lifetime to get a lifetime’s

longitudinal data, having started with a birth cohort. Sadly, criminologists get old as
fast as do those whom they study. In the accelerated longitudinal design, one begins
with a series of cohorts, with starting age of, say, 0, 5, 10, 15, etc. After five years,
one has data on 0–5 year olds from the cohort which started at 0, data on 5–10 year
olds from the cohort which started at 5, and so on. Thus, placing the cohort data
together, a lifetime of data (with some qualifications) is available.
4 The smaller the unit of analysis, the greater is the inequality. This is because looking
at large areas effectively averages diverse levels of risk within the area. The ultimate
level of analysis relates to individual households or people. Because individual
victimisation risks have not been analysed over time, this approach will overstate
the concentration of victimisation somewhat, since it will include real risk differ-
ences with chance fluctuations.
5 This strengthens the case for considering the distribution of crime opportunities
focusing on victims and perpetrators at the same time. Often, they are the same person.
6 In the recent rash of publications from the Home Office and the Inspectorate of
Constabulary about crime statistics, place as the primary classifying variable remains
unquestioned.
7 None of this should be taken to imply that displacement is typically a major problem
for crime reduction programmes (see Hesseling 1994). It is rather to suggest that
displacement research would (and should) have developed differently if informed by
notions of distributive fairness.

References
Arneson, R. (1989) ‘Equality of Opportunity for Welfare’, Philosophical Studies, 56:
77–93.
Barnes, G. C. (1995) ‘Defining and Optimising Displacement’, in J. E. Eck and D. Weis-
burd (eds) Crime and Place, Monsey, NY: Criminal Justice Press.
Barr, R. and Pease, K. (1990) ‘Crime Placement, Displacement and Deflection’, in M.
Tonry and N. Morris (eds) Crime and Justice, volume 12, Chicago: University of
Chicago Press.
Budd, T. (1999) Burglary of Domestic Dwellings: Findings from the British Crime Survey,
Home Office Statistical Bulletin, 4/99, London: Home Office.
Cohen, G. A. (1989) ‘On the Currency of Egalitarian Justice’, Ethics 99: 906–44.
Dworkin, R. (1981a) ‘What is Equality? Part 1: Equality of Welfare’, Philosophy and Public
Affairs, 10: 185–246.
—— (1981b) ‘What is Equality? Part 2: Equality of Resources’, Philosophy and Public
Affairs, 10: 283–345.
Elster, J. (1989) Solomonic Judgements: Studies in the Limitations of Rationality, Cambridge:
Cambridge University Press.
—— (1992) Local Justice: How Institutions Allocate Scarce Goods and Necessary Burdens,
New York: Sage.
Everson, S. (2001) Repeat Victims and Prolific Offenders, Ph.D. thesis, University of
Huddersfield.
Farrell, G., Ellingworth, D. and Pease, K. (1996) ‘High Crime Rates, Repeat Victimisa-
tion and Routine Activities’, in T. Bennett (ed.) Preventing Crime and Disorder,
Cambridge: Institute of Criminology.
Farrell, G., Tseloni, A., Wiersema, B. and Pease, K. (2001) ‘Victim Careers and “Career
Victims”? Towards a Research Agenda’, in G. Farrell and K. Pease (eds) Repeat
Victimisation, Monsey, NY: Willow Tree Press.
240 Paul Wiles and Ken Pease

Field, S. (1991) Trends in Crime and their Interpretation: A Study of Recorded Crime in Post-
war England and Wales, Home Office Research Study 119, London: HMSO.
Fraser, N. (1997) Justice Interruptus, London: Routledge.
—— (2000) ‘Rethinking Recognition’, New Left Review, 224: 107–19.
Hales, J. and Stratford, N. (1997) 1996 British Crime Survey (England and Wales), Tech-
nical Report, London: SCPR.
Hesseling, R. B. P. (1995) ‘Displacement: A Review of the Empirical Literature’, in R. V.
Clarke (ed.) Crime Prevention Studies 3, Monsey, NY: Criminal Justice Press.
Hood, R. (1992) Race and Sentencing, Oxford: Oxford University Press.
Hope, T. J. (1996) ‘Communities, Crime and Inequality in England and Wales’, in T.
Bennett (ed.) Preventing Crime and Disorder, Cambridge: Institute of Criminology.
Johnson, S. D., Bowers, K. and Hirschfield, A. (1997) ‘New Insights into the Spatial and
Temporal Distribution of Repeat Victimisation’, British Journal of Criminology, 37:
224–41.
Nussbaum, M. and Sen, A.(1993) The Quality of Life, Oxford: Clarendon Press.
Pascal, B. (1658/1961) Pensées, Harmondsworth: Penguin.
Pease, K. (1998) Repeat Victimisation: Taking Stock, London: Home Office Policing and
Reducing Crime Unit.
Rawls, J. (1971) A Theory of Justice, Cambridge: Cambridge University Press.
Roemer, J. E. (1996) Theories of Distributive Justice, Cambridge, Mass.: Harvard Univer-
sity Press.
Sen, A. (1980) ‘Equality of What?’, in S. McMurrin (ed.) Tanner Lectures on Human
Values,volume 1, Salt Lake City: University of Utah Press.
Shaw, M. (2001) ‘Time Heals All Wounds?’, in G. Farrell and K. Pease (eds) Repeat
Victimisation, Monsey, NY: Willow Tree Press.
Sherman, L. W., Gottfredson, D., MacKenzie, D., Eck, J., Reuter, P. and Bushway, S.
(1997) Preventing Crime: What Works, What Doesn’t and What’s Promising, Wash-
ington, DC: Office of Justice Programmes.
Skogan, W. (1991) ‘Redesigning the National Crime Victimisation Survey’, Public
Opinion Quarterly, 23: 1–24.
Trickett, A., Osborn, D. R. and Ellingworth, D. (1995) ‘Property Crime Victimisation:
The Roles of Individual and Area Influences’, International Review of Victimology, 3:
273–95.
Tseloni, A., Osborn, D. R., Trickett, A. and Pease, K. (forthcoming) ‘The Lifestyle and
Socio-economic Environment of Households Experiencing High Property Crime
Rates in England and Wales’, British Journal of Criminology, in press.
Von Hirsch, A. and Roberts, J. V. (1997) ‘Racial Disparity in Sentencing: Reflections on
the Hood Study’, Howard Journal, 36: 227–36.
Walzer, M. (1983) Spheres of Justice: A Defence of Pluralism and Equality, New York: Basic
Books.
Wiles, P. and Pease, K. (2000) ‘Crime Prevention and Community Safety: Tweedledum
and Tweedledee?’, in V. McLaren, S. Ballantyne and K. Pease (eds) Key Issues in
Community Safety and Crime Prevention, London: IPPR.
Wilkins, L. T. (1999) Did I Really Become a Criminologist?, unpublished manuscript.
Young, H. P. (1994) Equity, Princeton: Princeton University Press.
Young I. M. (1990) Justice and the Politics of Difference, Princeton: Princeton University
Press.
—— (1999) ‘Unruly Categories: A Critique of Nancy Fraser’s Dual Systems Theory’,
New Left Review, 223: 147–60.
Index

Note: Page numbers in bold type refer to figures

Page numbers in italic type refer to tables
Page numbers followed by ‘n’ refer to notes

actuarialism 194–6 capital punishment 131

anomie 32 CCTV 227
anti-social behaviour 70; anti-social chain gangs 175
behaviour orders (ASBOs) 64, 77n Childline 172
area deprivation/affluence index 206n children 169, 176; rights 175
arrests and searches, London (1997–9) civilising process 101
150 Clarke, R. 3
Asians 147–8, 154 CMOCs 92–3, 94
Association of Chief Police Officers 155 cognitive-skills training 182–3, 187
Atkins, S. et al. 89 Cohen, S. 10
Audit Commission 8, 69, 70, 77n, 81, 167, Colley, H. 187–8
173, 174 communitarianism 58, 107–14
community 26–53; deterritorialisation
Barr, R. 233–4 36–7; late modernity 38; organic 31,
Bauman, Z. 13, 42–3, 48 33, 35; phantasmagoric 35; policing
Bayley, D. 105 132; safety 3–6, 11, 15, 20, 98–119,
Beck, U. 197–8 167–8
Blair, T. 43, 47, 54, 55, 60, 172, 178 compensation 227
Boateng, P. 173 complex equality 220, 221
boot camps 175 conditional discharges 170
Bottoms, A. 7, 39, 40, 41 Cooley, C. 40
Bourdieu, P. 198 Cooper, A. 7
Box, S. 103 correctionalism 167–89
Braga, A. 137 Court Orders 174
Braithwaite, J. 108 Crawford, A. 8, 26, 31, 32
Bratton, W. 127 crime: audit 68–9; control 15, 20;
British Crime Survey (BCS) 155–6, 160, distribution 228
196, 205, 224 Crime and Disorder Act (1998) 4, 7, 56,
‘Broken Windows’ 7, 14, 32, 120–44 61, 69, 81, 103, 170, 175; ethnicity
Buchanan, A. 109 164; Section (37) 167; youth justice
Bulger, James 176 provisions 168–9
burglary 150, 185, 232 crime prevention 3–6, 92, 105
Crime Reduction Programme 81
Campbell, B. 178 Crime Sentences Act (1994) 171
cannabis 150 Criminal Justice System 122–4, 146
242 Index

Crown Prosecution Service (CPS) 156 gangs, chain 175

culture wars 124–5 Garland, D. 75, 76, 214
curfews 176 Gendreau, P. et al. 185
Currie, E. 169, 179 ghetto 18, 42
Giddens, A. 19, 29, 35, 54, 55, 61, 113
Dahrendorf, R. 16 Giuliani, Mayor R. 120
Davis, M. 15, 213 Gladwell, M. 140
Deakin, N. 63 good parenting 169
decentralisation 9–12 Gordon, D. 205
decriminalising 123 Gorz, A. 49
dejuvenilisation 175–6 government: joined-up 59–60; policy 114
delinquency 111 graffiti 125
Delors, J. 17 Grieve, J. 157
deprivation: and affluence 206; relative gun: control 131; culture 126–8
27, 29, 33, 41, 106
Detention and Training Order (DTO) 171 Hale, C. 103
Diallo incident 129 Hannah-Moffat, K. 20
disorder 12–15, 100, 125–39 harassment, sexual 45
displacement 3, 233 Harcourt, B.E. 121
disproportionality 151–4 hate crime 157
distributive justice 27, 219–39; and crime Hebdige, D. 39
vicitimisation 226–8 Her Majesty’s Inspectorate of
doli incapax 176 Constabulary (HMIC) 81
domestic burglary 232 Hirsch, F. 212
domestic violence 84, 85, 87, 95 Hobsbawm, E. 34
Downes, D. 55, 102 homophobic incidents 157
drugs 44, 103, 150, 154; dealers 137; Honneth, A. 109
legislation 131 Hood, R. 30, 164
Dworkin, R. 220, 225 Hope, T. 19, 37, 112
household crime: and area risk factors 202;
electronic tagging 174, 176 victimisation surveys 195
Elejabarrieta, F. et al. 184 household security measures 200
Elias, N. 101 housing estates: London 32–3; problem
essentialism 30–1 195
ethnicity 145–66, 223; age structure 153 Hughes, G. 98
ethnography 42 Hutton, W. 29
Etzioni, A. 58, 107
Evans, K. 32 identity crisis 30–1
evidence-led policy 81–97 incarceration 171; see also prisons
exclusion, social see social exclusion inequality 220; force and OCU level 233;
see also social exclusion; underclass
families: failing 177–9; lower class 178 injuries, compensation 223
fare beating 127, 141n institutional racism 146, 162
Farrell, G. et al. 226 interpersonal violence 103
Farrington, D. 8, 178
feminists 2, 48 Johnstone, G. 175
Field, S. 223 joined-up government 59–60
FitzGerald, M. 17 Jones, K. 30
foot patrols, police 125 Jordan, B. 211, 216
Foster, J. 32, 33 justice: distributive 184, 222, 223, 227
Franke, H. 102
Fraser, N. 27, 45, 221, 222 Kansas City Gun Experiment 127
Fukuyama, F. 124 Kelling, G. 14, 73
 Index 243

Kirkholt burglary reduction programme New Public Management (NPM) see

89–90, 185 managerialism
Knuttson, J. 89 New York 127, 128
Nightingale, C. 41
Labour Party: New Labour 7, 8, 167–93; non-custodial preventive disposals 170
youth crime 169 Northamptonshire Youth Diversion
Lawrence, Stephen 145 Scheme 184–5
Levitas, R. 44 Nutall, C. 181
liberalism 107–14
Liddle, R. 179–80 offensive weapons 150
lifestyles 197 ontological insecurity 27, 30, 193
lighting 89 Operational Command Unit (OCU) 231;
Lipsey, M. 185 rate variability by offence types 236
local authorities 11, 114 organic community 31, 33, 35
Local Authority Child Protection Register
174 Painter, K. 89
longitudinal design 239n Pantazis, C. 205
Los Angeles 15, 213 Parenting Order 178
Lousada, J. 7 parents 108; good 169; single household
111, 237
Macpherson Report (1999) 17, 145, 146, parole 195
160, 161, 162 Pascal, B. 220
managerialism 6–9, 59–63 Pawson, R. 86, 90, 182, 188
Mandelson, P. 179–80 Pease, K. 102, 233–4
Maple, J. 127 Perfect, M. 169
Marcuse, H. 108 performance indicators 6, 235
Matthews, R. 132, 133, 134 Peters, A.A.G. 66
Mead, G.H. 440 phantasmagoric community 35
Medway Secure Training Centre 172 Phillips, M. 177, 178
Meier, R.E. 199 Pitts, J. 18
mentoring 186; methodological problems police 5, 13, 127, 155, 160; cautioning rate
187–8 173; foot patrol 125; OCU 231; zero
Metropolitan Police Survey (MPS) 158, tolerance 73, 98, 129; see also
161 Metropolitan Police Survey
Meyrowitz, J. 36, 37, 40–1 Police and Criminal Evidence Act
middle class, rich 204 (PACE)(1984) 147
middle-range theories 92–3 policy: evidence-led 81–97, 114;
Miethe, T.D. 199 government 114
Minneapolis Domestic Violence politics: parties 105; risk avoidance
Experiment 84 210–14; youth justice 175; see also
Mitterand, F. 16 Labour Party
Morgan, R. 55 positional good 212
Morgan Report (1991) 4, 10, 63, 105 post-Fordism 29
multi-agency approaches 135 poverty 210
multiple victimisation 5 Power, M. 67, 68, 69
murder 137 pre-court preventive interventions 170
Murray, C. 16, 177 pre-emptive interventions 169
pre-sentence reports 224
National Institute of Justice 122 prisons 92
neighbourhood watch schemes 105 private goods 193
Neil, A. 177 private security 212; collective provision
neo-tribalism 111 215
New Labour 7, 8, 167–93 privatisation 9–12
244 Index

probation services 94 Scottish Children’s Hearings 171

progressive minimalism 169, 173, 180 searches: and arrests (London 1997–9)
property crime 196, 199, 207, 223, 229; 150; offensive weapons 150;
area crime levels 229; area victim population and suspects (London
concentration 231; area victim 1998–9) 152; unrecorded 149;
prevalence 230; inequality 203; and weapons 150
poverty 210; risk factors 209; Secure Estate 168, 169, 172–3
victimisation 207, 208 Secure Training Centres (STCs) 171, 172,
property marking 89 180
prostitution 13, 107; see also Wolfenden security, private 212, 215
Report (1957) security club 212
public goods 193 seemocks (CMOCs) 92–3, 94
public tolerance 12–13, 104–6 Sennett, R. 28, 34, 39
punishment: capital 131; Netherlands 102 sexual harassment 45
shaming 112
Quantum Opportunities Programme 187 Shapland, J. 62
Shaw, J.W. 127
Race and Violent Crimes Task Force Sherman, L. 85, 87, 95, 127, 182
(1998) 157, 158 Shipman, H. 90
racial incidents 156, 158; England and Siegel, F. 130, 131
Wales, (1991–99) 157; harassment Simon, J. 175
146, 155–62, 159; MPS (1997–2000) single parent households 111, 237
158 Skogan, W.G. 121
racism 105; institutional 146, 162 Smith, J. 47
Ramsbotham, Sir D. 189 social capital 213
Raudenbush, S.W. 132, 137, 138, 139 social class 178, 208, 211; see also middle
Rawls, J. 219, 220; difference principle class; underclass; working class
235–6 social control 29, 32, 110, 151; districts 15
realistic evaluation 90–3, 93–4 social exclusion 15–18, 26–53, 43
recidivism 182, 183 Social Exclusion Unit (1999) 17, 37, 43,
referral order 171 56, 62
rehabilitation 180 Social Services Inspectorate 173
relative deprivation 27, 29, 33, 41, 106 Social Trends (2000) 37
restorative justice 184, 222, 227 spatial location of crime 231
retributive justice 222, 223, 227 squeegeeing 125, 132, 141n
Rhodes, R. 62, 64, 66, 71 statistics 6, 163, 236
rights: children 175; gay 48 stop and search 146, 147–55
risk: distribution 205–10; exposure 200–1; Straw, J. 58, 64, 98, 172, 174, 176
manufactured 196; models 201–5 surveillance 174, 227
risk avoidance, politics 210–14, 214 Sus laws 154
risk factors 8; property crime victimisation
209 tagging, electronic 174, 176
risk society 19–20, 29, 193–218; crime risk Tam, H. 58
194 target hardening 227, 238n
robbery 137 Taub, R. 106
Roemer, J.E. 219 television 37, 39, 42
Rogan, D.F. 127 Third Way 7, 54–80, 113
Room, G. 16, 17 Thomas, C. 45
Ross, R. 182 Thompson, J. 36
Runciman, W. 29 Tilley, N. 12, 182, 188
tolerance 99–101; level of 106; public
Samaritans 172 12–13
Sampson, R.J. 132, 137, 138–9 Tomlinson, M. 36
 Index 245

underclass 15–18, 32, 41–3, 178 welfare state 28, 177

Utting, W. et al. 178 Wellman, B. 37
West, D. 8, 178
Van Hooris, P. 185 Wheen, F. 49
Vanstone, M. 183 Wiles, P. 39, 40, 41
vehicle searches 126 Wilkins, L. 225
verbal abuse 156, 158 Wilson, D. 185
vice squads 13 Wilson, J.Q. 14, 73, 120, 124
victimisation 184, 193–218, 226; and Wilson, W.J. 16, 18
distributive justice 226–8; multiple 5; Wittgenstein, L. 43
poor/rich households 207; property Wolfenden Report (1957) 13
crime 207, 208 working class 178; poor 200, 204
victimology 2
victims 13, 62, 159, 160, 161; support 223 Young, I.M. 35, 110, 221
videotapes 136, 180 Young, J. 194
vigilante squads 13 youth crime 169
violence 223, 230; domestic 84, 85, 87, 95; youth culture 39
interpersonal 103 Youth Justice: administrative structure
voice verification systems 174 169–70; Board of England and Wales
Vulnerable Persons Units 157 (YJB) 168; community safety 167–93;
and Criminal Evidence Act (1999) 56,
Walklate, S. 32 170; Diversion Scheme
Walsh, K. 63 (Northamptonshire) 184–5; politics
Walzer, M. 215, 219, 220, 221 175; Task Force (1998) 167, 172
Warner, Lord 59 Youth Offending Teams (YOTs) 62, 173–4
weapons 127; searches 150 Youth Treatment Centres 168, 173
Weatheritt, M. 70
Webster, F. 39 zero tolerance 98, 129; policing 73, 98, 129