ISSUES IN POLITICAL THEORY

Political Theory has undergone a remarkable development in

recent years. From a state in which it was once declared dead,
it has come to occupy a central place in the study of Politics.
Both political ideas and the wide-ranging arguments to which
they give rise are now treated in a rigorous, analytical fashion,
and political theorists have contributed to disciplines as diverse
as economics, sociology and law. These developments have made
the subject more challenging and exciting, but they have also
added to the difficulties of students and others coming to the
subject for the first time. Much of the burgeoning literature in
specialist books and journals is readily intelligible only to those
who are already well-versed in the subject.

Issues in Political Theory is a series conceived in response to this

situation. It consists of a number of detailed and comprehensive
studies of issues central to Political Theory which take account
of the latest developments in scholarly debate. While making
original contributions to the subject, books in the series are
written especially for those who are new to Political Theory.
Each volume aims to introduce its readers to the intricacies of
a fundamental political issue and to help them find their way
through the detailed, and often complicated, argument that that
issue has attracted.

PETER JONES
ALBERT WEALE
ISSUES IN POLITICAL THEORY
Series editors: PETER JONES and ALBERT WEAL£

Published

David Beetham: The Legitimation of Power

Christopher J. Berry: Human Nature
Tom Campbell: Justice
Tim Gray: Freedom
Michael Lessnoff: Social Contract
Richard Lindley: Autonomy
Susan Mendus: Toleration and the Limits of Liberalism
Andrew Reeve: Property

Forthcoming
John Horton: Political Obligation
Peter Jones: Rights
Raymond Plant: Equality
Hillel Steiner: Utilitarianism

Series Standing Order

If you would like to receive future titles in this series as they are published.
you can make use of our standing order facility. To place a standing order
please contact your bookseller or, in case of difficulty. write to us at the
address below with your name and address and the name of the series.
Please state with which title you wish to begin your standing order.
(If you live outside the United Kingdom we may not have the rights for your
area, in which case we will forward your order to the publisher concerned.)

Customer Services Department. Macmillan Distribution Ltd,

Houndmills. Basingstoke. Hampshire. RG21 2XS. England.
The Legitimation of Power

David Beetham

M
MACMILLAN
© David Beetham 1991

All rights reserved. No reproduction, copy or transmission

of this publication may be made without written permission.

No paragraph of this publication may be reproduced, copied

or transmitted save with written permission or in accordance
with the provisions of the Copyright, Designs and Patents Act
1988 or under the terms of any licence permitting limited
copying issued by the Copyright Licensing Agency,
33-4 Alfred Place, London WC1E 7DP.

Any person who does any unauthorised act in relation to

this publication may be liable to criminal prosecution and
civil claims for damages.

First published 1991

Published by
MACMILLAN EDUCATION LTD
Houndmills, Basingstoke, Hampshire RG21 2XS
and London
Companies and representatives
throughout the world

Phototypeset by Intype, London

ISBN 978-0-333-37539-6 ISBN 978-1-349-21599-7 (eBook)

DOI 10.1007/978-1-349-21599-7

A catalogue entry for this book is available from the British Library
Contents
Preface and Acknowledgements vii

PART I THE CRITERIA FOR LEGITIMACY

1 Towards a Social-scientific Concept of Legitimacy 3
The different dimensions of legitimacy 15
The significance of legitimacy 27
Conclusion: a social or political concept? 37

2 Power and its Need of Legitimation 42

The concept of power 43
The social organisation of power 46
Power and legitimacy 56

3 The Normative Structure of Legitimacy 64

Legitimacy derived from rules 64
Legitimacy as justifiability of rules 69
Legitimacy through expressed consent 90
Conclusion 97

4 The Social Construction of Legitimacy 100

PART II LEGITIMACY IN THE CONTEMPORARY STATE

5 Dimensions of State Legitimacy 117
The elusiveness of legality 121
Constitutional rules and their justification 126
Modes of popular consent 150
Conclusion 158

v
VI The Legitimation of Power

6 Crisis Tendencies of Political Systems 101

Liberal democracy 163
Liberal democracy in developing countries 171
The Communist model 179
Types of Islamic state 191

7 Modes of Non-legitimate power 205

Revolutions and legitimacy 213
Revolutionary regimes and relegitimation 221
Coups d'etat and military regimes 228
Conclusion: the dilemmas of political legitimacy 237

8 Legitimacy in Political Science and Political Philosophy 243

Bibliography 251
Index 260
Preface and
Acknowledgements
Unlike other volumes in the series, this book is more a work of
social science than of normative philosophy. Its ambition is to
rescue the concept of legitimacy from the confusion into which
it has sunk, so that its place in the analysis of power relations
can be securely located. Fulfilling such an ambition is necessarily
both a theoretical and a philosophical task. The first part of
the book identifies the different components of legitimacy, and
explores the processes of legitimation across the dimensions of
gender, class and political power. The second part offers an
account of legitimacy in relation to the contemporary state. The
argument of the book is cumulative, so that each part is necessary
to the other. I hope readers will forgive such a presumption on
their continuing interest.
I have incurred many debts in writing the book, and learnt a
lot from the comments of others. Draft chapters have been dis-
cussed by Diana Coole, Malcolm Harrison, Veronica Munoz,
Ben Rogers and John Schwarzmantel. The bulk of the manu-
script has been read by Zygmunt Bauman, Margaret Beetham,
Diemut Bubeck, David Held, Steven Lukes, Peter Jones and
Albert Weale, the last two in their capacity as editors of the
series. I am grateful to them all for the trouble they have taken.
Initial work on the project was made possible by a grant from
the Economic and Social Research Council, and I acknowledge
their support for research that is more interpretative than utili-
tarian in purpose. I must plead (who can't?) the pressure of
administrative and teaching responsibilities for not having com-
pleted it sooner. Among the benefits of procrastination must be
counted the fact that the manuscript was not ready when the
European revolutions of 1989 took place, nor yet when the Iraqi

vii
VIJI Preface and Acknowledgements

army invaded Kuwait. Although details of my analysis will be

further overtaken by events, the main thrust of the argument
will, I hope, prove sound. Finally, as one who remains wilfully
devoted to the pen in preference to the keyboard, I salute once
more the skills of Jeanne Bellovics.

David Beetham
 PART I
THE CRITERIA FOR
LEGITIMACY
1 Towards a Social-
scientific Concept of
Legitimacy
The exercise of power by one person over others, or by one
group over another, is a basic and recurrent feature of all societ-
ies. Those who are subordinate experience it as constraining,
often humiliating and sometimes life-threatening; and many
would escape it if they could. Those who hold power, or seek
to do so, are themselves frequently at odds with one another
over the scope of their power and the control over their subordi-
nates, with potentially damaging consequences. Power, in other
words, is a highly problematical, as well as recurrent feature of
human societies. And because it is so problematical, societies
will seek to subject it to justifiable rules, and the powerful them-
selves will seek to secure consent to their power from at least
the most important among their subordinates. Where power is
acquired and exercised according to justifiable rules, and with
evidence of consent, we call it rightful or legitimate.
How far power is legitimate, what makes it so, and why it
matters: these are all inherently difficult and contentious ques-
tions. They have at various times and places seriously exercised
those involved in power relations, especially in periods of legal
uncertainty, moral disagreement or intense social and political
conflict. These questions have also been the special concern of
different groups of professionals - legal experts, moral or politi-
cal philosophers, social scientists, to name but three - who have
each approached them from a different focus of interest, and
have tended to employ different conceptions or definitions of
legitimacy according to their respective professional standpoint.
So in addition to the inherent difficulty of deciding what makes

3
4 The Legitimation of Power

power legitimate, there is the extra complication of divergent

definitions offered by different groups of professionals. It is this
double layer of complexity that makes the subject of legitimacy
so confusing.
In unravelling this complexity, I shall begin with the different
professionalisms, or academic specialisms as we could call them,
since this will provide the best point of departure for understand-
ing what the subject of legitimacy is about. To begin with, legal
experts, and especially constitutional lawyers, are concerned with
the resolution of legal disputes about power: how it is validly
acquired, who is entitled to exercise it, within what limits, and
so on. Their special expertise and focus of interest is with the
definition and interpretation of legal rules, and with how these
are initiated, revised and enforced. For them, power is legitimate
where its acquisition and exercise conform to established law.
For them legitimacy is equivalent to legal validity.
Now there is an obvious sense in which we would all recognise
legal validity as an important aspect of legitimacy. A recurrent
context in which the terms 'legitimate' and 'illegitimate' have
traditionally been used has been in relation to children, and
whether they have been conceived within a legally authorised
marriage or not. This question has historically derived its signifi-
cance from its relation to issues of power, and in particular to
disputes about the succession to property and position. In socie-
ties where the chief mode of access to power is through birth,
it becomes of supreme importance to determine who is the legit-
imate, i.e. legally valid, heir to an estate or a dynasty. And what
applies to power based on heredity applies with equal force to
any other system of power; the mode of access to it, as well
as its scope and duration, require social regulation, and the
characteristic means of doing so is through legal rules. To say
that the power a person has was legally acquired and is exercised
within the law, is a first condition of its legitimacy.
However, if legal validity is a recognisable element in legit-
imacy, it cannot by any means exhaust it. Disputes about the
legitimacy, or rightfulness, of power are not just disputes about
what someone is legally entitled to have or to do; they also
involve disagreements about whether the law itself is justifiable,
and whether it conforms to moral or political principles that are
rationally defensible. Are the relations of power, of dominance
Social Science and Legitimacy 5

and subordination, which the law sustains, are the rules that
determine access to positions of power or the means of exercising
it, themselves rightful? If not, are we obliged to obey them, or
to cooperate with the powerful, on any other grounds than a
prudent calculation of our own self-interest? These moral ques-
tions and practical dilemmas about power go deeper than the
question of its legal validity; they concern the justification for
the law itself. It is not what the law actually prescribes, but what
it ought to prescribe, that is here the central issue of legitimacy.
Now, the question of how power relations within a society
ought to be arranged, and what would count as a sufficient
justification to require the support of the subordinate for them,
has been the special concern of moral and political philosophy.
Some philosophers have been primarily apologetic, seeking justi-
fications for an existing structure of power; others have been
critical, exposing its deficiencies, and advocating revisions or
alternatives to it. Yet what they share is the project of elucidating
the most general principles - of justice, of right, of social utility
- necessary to the justification of power relations; and also, in
the modern period at least, of specifying what conditions have
to be met if those subject to power can be said to have consented
to it, and so be morally obliged to obey or support it. For the
moral and political philosopher, power is legitimate where the
rules governing it are justifiable according to rationally defensible
normative principles. And as with any moral principles, these
embody a universalising claim; it is not the principles that happen
to pertain in a given society that are sufficient, but those that
any rational person, upon considered and unbiassed reflection,
would have to agree to. What is 'legitimate' to the philosopher,
then, is what is morally justifiable or rightful; legitimacy entails
the moral justifiability of power relations.
If we turn, finally, to the social scientist, we see a different
focus of interest from those of the legal expert or moral philos-
opher. The social scientist's concern is not with solving legal
disputes or moral dilemmas about power, or helping others to
do so; it is rather with identifying the empirical consequences
that legitimacy has for the character of power relations, for the
different ways in which they are organised, and for the extent
to which the powerful can actually count on the obedience or
support of those subordinate to them. The social scientist's pur-
6 The Legitimation of Power

pose, in other words, is primarily an explanatory one. An under-

standing of legitimacy helps explain, for example, why people
have the expectations they do about a power relationship, why
institutions of power differ systematically from one type of
society to another, why power is exercised more coercively in
some contexts than in others. Above all, it helps explain the
erosion of power relations, and those dramatic breaches of social
and political order that occur as riots, revolts and revolutions.
It is not just because these events are particularly dramatic and
fateful that they interest the social scientist. As with so much
else about society, it is only when legitimacy is absent that we
can fully appreciate its significance where it is present, and where
it is so often taken for granted.
What, then, is legitimacy for the social scientist? What makes
power legitimate? Social scientists, unlike moral or political phil-
osophers, are concerned with legitimacy in particular historial
societies rather than universally; with legitimacy in given social
contexts rather than independent of any particular context; with
actual social relations rather than ideal ones. They are only too
aware that what makes power legitimate in one society may
differ from others, and that the criteria for legitimacy in one
may be rejected by another. Moreover, they are trained to stand
back from their own values and beliefs so as better to understand
those of others. As individuals they may be convinced by philo-
sophical anarchism, but this position will not help them under-
stand the legitimacy of the modern state, or distinguish where it
is, from where it is not, legitimate. As individuals they may
regard all religious beliefs as rationally unfounded and indefen-
sible, but such a judgement will not help them understand the
legitimacy of the late Ayatollah Khomeini. What matters for an
adequate understanding is not what they personally believe, but
what is believed in the society they are studying. For this reason
most social scientists in the twentieth century have followed Max
Weber in defining legitimacy as the belief in legitimacy on the
part of the relevant social agents; and power relations as legit-
imate where those involved in them, subordinate as well as domi-
nant, believe them to be so (Weber, 1968, p.213).
At this point in the discussion it is important that I make clear
what sort of book this is going to be. It is not intended as a
work of legal theory or political philosophy, as I have character-
Social Science and Legitimacy 7

ised them above. In particular, it is not about political obligation

as philosophers would treat it. My purpose is not to clarify the
moral dilemmas of citizens faced with the choice of whether or
not to break the law. This is not because I consider such issues
unimportant or unamenable to rational argument. It is because
my concern is with the different, social-scientific question: what
are the causes and consequences of people's disobedience? That
is to say, I am concerned with legitimacy as a problem for social
science rather than for political philosophy.
There are a number of reasons for my choice. The first is that,
although I am a political philosopher as well as a social scientist,
I believe that the two activities have a different purpose and a
different logic, and neither is best served by confusing it with
the other. This is not to say that there is no significant connection
between the two activities; but the connection can only properly
be defined, as I hope to show, by first carefully observing the
distinction between them. My second reason is that handling
normative topics, which is the everyday business of the philos-
opher, provides the severest challenge to the social scientist,
and confronts him or her with the most acute methodological
difficulties. Nowhere is it more essential to overcome these diffi-
culties than in the study of power relations, which because of
their importance carry such a potential for obfuscation. On both
methodological and substantive grounds, then, the subject of
legitimacy must count as one of the central issues of social sci-
ence. My third reason is that social scientists have in fact been
thoroughly confused about legitimacy, and their confusion has
its starting point in their failure to conceptualise it adequately,
or to offer a coherent account of what makes power legitimate
in particular societies. If that judgement sounds like arrogance
on my part, it is also a confession: I have been thoroughly
confused myself. The following discussion, which seeks to ident-
ify the confusion and offer a convincing resolution of it, is also
the record of a personal journey out of the maze which this
subject represents.
The source of the confusion lies with the work of Max Weber.
It is one of the most remarkable features about the study of
legitimacy in politics departments, at least in the Anglo-Saxon
world, that it is suspended between two separate bodies of litera-
ture that have absolutely no connection with one another. If you
8 The Legitimation of Power

are studying legitimacy as a subject in political philosophy you

will probably start with Thomas Hobbes, if not earlier, and
proceed through the great tradition which includes Locke, Rous-
seau, Hegel and others. If you are studying it as a subject in
political science or political sociology you will most likely begin
with Max Weber, and may not discuss other thinkers at all, but
proceed to a series of empirical case studies of power relations
and theoretical explanations for obedience and disobedience. I
will return to the problems exposed by this extraordinary disjunc-
tion later; but for the moment my starting point as a social
scientist has to be with Weber, rather than with the earlier
theorists.
Max Weber is rightly regarded as one of the 'founding fathers'
of twentieth-century social science, and his influence across a
range of disciplines and subjects has been enormous, and usually
beneficial, even where later thinkers have disagreed with him.
On the subject of legitimacy, however, it has to be said that his
influence has been an almost unqualified disaster. The starting
point of what is wrong lies in his definition of legitimacy. For a
social scientist to say that a given power relation is legitimate,
Weber argues, is not to make a moral judgement about it in the
manner of the philosopher; it is rather to make a report (which
may be empirically true or false) about other people's beliefs.
Power is legitimate where those involved in it believe it to be
so; legitimacy derives from people's belief in legitimacy. So, he
writes, legitimacy is equivalent to 'Legitimitatsglaube' (a belief
in legitimacy); and legitimate power is power 'als legitim ange-
sehen' (that is regarded as legitimate) (Weber, 1956, pp.23, 157,
659; 1958, p.493).
What is wrong with this formulation can be seen most clearly
from what other social scientists have made of it. Social scientists
are, by training, dismissive of universal truths and values; they
are only too aware of the variety of beliefs held from one society
to the next, and from one historical epoch to another. They
become sceptical about the possibility of any rational grounding
for normative ideas or value systems, which they frequently label
as 'myths', 'prejudices' or 'ideologies'. Such scepticism leads
them in turn to concentrate on the processes of socialisation, the
structures of influence, the agencies of dissemination whereby
ideas come to be acquired and reproduced. The beliefs people
Social Science and Legitimacy 9

hold are thus explained as the product of the cumulative influ-

ences to which they have been exposed.
Such an explanation appears particularly plausible in an age
of propaganda and public relations, when the public sphere is
dominated by an emphasis on presentation over reality. If people
believe in the legitimacy of power, is this not because the power-
ful have been successful in the public relations campaign, because
they have managed to convince people that they are legitimate,
because their 'legitimations' have been accepted? Is the question
of their legitimacy not therefore in the hands of the powerful
themselves? So S.M. Lipset, in typically Weberian vein, defines
the legitimacy of a political system as its capacity 'to engender
and maintain the belief that the existing political institutions are
the most appropriate ones for the society' (1958, p.86). And R.
Merelman calls legitimacy 'a quality attributed to a regime by a
population. That quality is the outcome of the government's
capacity to engender legitimacy' (1966, p.548). Taken to their
logical conclusion, such definitions would imply that the reason
for the collapse of the communist regimes in Eastern Europe in
1989 lay in a deficiency of public relations, rather than anything
actually wrong with the system of rule itself.
At this point political philosophers become indignant with
social scientists and their Weber-inspired definitions, which trans-
form the issue of legitimacy from a question about the actual
characteristics of a system of power into one concerning the
beliefs people hold about it. In so doing, it is argued, they are
emptying the concept of legitimacy of any objective reference or
moral content, and in effect acquiescing in the very manipulations
of the powerful that they are concerned to describe. 'The new
definitions,' writes J .H. Schaar, 'all dissolve legitimacy into belief
or opinion. If a people holds the belief that existing institutions
are "appropriate" or "morally proper", then those institutions
are legitimate. That's all there is to it.' (1969, p.284). Hannah
Pitkin in turn draws attention to the epistemological conse-
quences of such a definition: 'In seeking to insulate the sociol-
ogist from the context of judging and taking a position,' she
writes, 'Weber in effect made it incomprehensible that anyone
might judge legitimacy and illegitimacy according to rational,
objective standards' (1972, p.283). And Robert Grafstein points
to the abandonment of moral judgement entailed by the Weber-
10 The Legitimation of Power

ian approach: 'The concept should properly signify a normative

evaluation of a political regime: the correctness of its procedures,
the justification for its decisions, and the fairness with which it
treats its subjects. In Weber's hands, however, legitimacy no
longer represents an evaluation of a regime; indeed it no longer
refers directly to the regime itself' (1981 p.456).
These criticisms seem to me entirely justified in pointing to the
reductionist conclusions that can be drawn from the Weberian
definition of legitimacy. Yet they are also mistaken if they imply,
as they seem to, that the social scientist ought to become a
moral or political philosopher, and engage in evaluating a regime
against independent normative standards. The problem with the
Weberian definition is not that it fails to meet the criteria of
normative philosophy, but that it encourages bad social science.
And one reason it does so is that it leaves the social scientist
with no adequate means of explaining why people acknowledge
the legitimacy of power at one time or place and not another.
The social scientist, it seems, is someone who must always be
taken by surprise when people stop treating power as legitimate
and take to the streets in protest.
Now it can be argued in Weber's defence that his social-
scientific definition of legitimacy as the 'belief in legitimacy' need
not have the reductionist and manipulative implications I have
described above, and that all I have offered is a caricature of
his position. What is mistaken, it could be said, is to divorce
people's beliefs about legitimacy from their grounds or reasons
for holding them; and these are to be found precisely in the
actual characteristics of a regime, such as its conformity to their
values, its ability to satisfy their interests, and so on. Did not
Weber himself explore the different grounds for people's belief
in legitimacy in his analysis of the rational-legal, traditional and
charismatic principles of authority? And would not a social scien-
tist who was alert to the actual inadequacies and processes of
degeneration of a regime be able, if not to predict, then at least
adequately to explain, the erosion of belief in its legitimacy? The
mistake, in other words, is not Weber's, but that of those social
scientists who have reduced the explanation of beliefs to the
processes and agencies of their dissemination and internalisation,
rather than an analysis of the factors which give people sufficient
grounds or reasons for holding them.
Social Science and Legitimacy 11

I have some sympathy with such a reply. In particular, I shall,

myself, at various points in this work be analysing people's beliefs
and examining their reasons for holding them. Yet the above
reply is still handicapped by the Weberian definition of legitimacy
as the 'belief in legitimacy', within whose orbit it remains. What
is wrong with this definition is, first, that it misrepresents the
relationship between beliefs and legitimacy; and, secondly, that
it takes no account of those aspects of legitimacy that have little
to do with beliefs at all. This brings me now to the heart of the
matter.
The first objection to the Weberian definition of legitimacy -
which could be called the 'received' definition, since almost all
social scientists have adopted it - is that it misrepresents the
relationship between legitimacy and people's beliefs. A given
power relationship is not legitimate because people believe in its
legitimacy, but because it can be justified in terms of their beliefs.
This may seem a fine distinction, but it is a fundamental one.
When we seek to assess the legitimacy of a regime, a political
system, or some other power relation, one thing we are doing
is assessing how far it can be justified in terms of people's beliefs,
how far it conforms to their values or standards, how far it
satisfies the normative expectations they have of it. We are
making an assessment of the degree of congruence, or lack of
it, between a given system of power and the beliefs, values and
expectations that provide its justification. We are not making a
report on people's 'belief in its legitimacy'.
An example will clarify the point I am making. It is argued
that the British electoral system, with its first-past-the-post rules
determining who shall be elected in each constituency, is losing
its legitimacy, and to an extent therefore also weakening that of
the governments elected under it. This is not because of any
shift in people's beliefs, but because the rules have increasingly
delivered results that diverge, both regionally and nationally,
from the proportion of votes cast, and hence from accepted
notions about the representative purpose of elections in a democ-
racy. It is the increasingly unrepresentative character of the elec-
toral system, and its consequent vulnerability to attack in a
society that believes in representation, that is the basis for the
weakening legitimacy of governments appointed under it. The
vulnerability was there before it was exploited, and the weaken-
12 The Legitimation of Power

ing of legitimacy took place before people publicly acknowledged

it. It may have taken the poll-tax legislation to bring the issue
to the forefront of public attention. But the potential for doing
so was already present in the growing discrepancy between the
rules and the beliefs or values underpinning them. It is this
discrepancy that is important to an analysis of legitimacy; what
has occurred cannot be made intelligible in terms of a shift in
people's beliefs about legitimacy or 'belief in legitimacy'.
The Weberian definition not only misrepresents the role that
beliefs play in legitimacy. In making legitimacy primarily a matter
of belief, it also ignores those elements which are not really to
do with beliefs at all. I have already mentioned legality as an
important component of legitimacy. Whether power is or is not
acquired and exercised within the law is a question quite indepen-
dent of people's beliefs; it is a matter of judicial determination.
Or consider another element in legitimacy that I have so far only
touched on: that of consent. Despite the confusion that some
political philosophers have created with the notion of 'tacit con-
sent' (e.g. Locke, 1967, pp.365-7), what is important for legit-
imacy is evidence of consent expressed through actions which
are understood as demonstrating consent within the conventions
of the particular society, such as: concluding an agreement or
entering into a contract with a superior party; swearing an oath
of allegiance; joining in acclamation; voting in an election or
plebiscite; and so on.
Why are such actions important? It is not that they provide
evidence of people's 'belief in legitimacy'. Consent can be given
from a variety of different motives, including considerations of
personal self-interest. What is important about these actions is
that they confer legitimacy; they contribute to making power
legitimate. They do this both through the public demonstration
of people's consent to the power relationship, and through the
resulting obligations that derive from them on the part of both
dominant and subordinate alike. They possess simultaneously a
symbolic and a normative force.
Contributing to legitimacy, then, are to be found a number of
different factors, operating at different levels. There is the legal
validity of the acquisition and exercise of power; there is the
justifiability of the rules governing a power relationship in terms
of the beliefs and values current in the given society; there is
Social Science and Legitimacy 13

the evidence of consent derived from actions expressive of it.

These factors, successively and cumulatively, are what make
power legitimate. To the extent that they are present, it will be
legitimate; to the extent that they are absent, it will not. Together
these criteria provide grounds, not for a 'belief in legitimacy',
but for those subject to power to support and cooperate with its
holders; grounds, that is to say, not for belief, but for obligation.
It follows that the social scientist, in concluding that a given
power relationship is legitimate, is making a judgement, not
delivering a report about people's belief in legitimacy. The Web-
erian definition not only misconceives the nature of legitimacy;
it also proposes a quite misleading research strategy for determin-
ing whether power is legitimate: that of asking people whether
they believe it is. Apart from the problem of expecting ordinary
people to understand what legitimacy means, when social scien-
tists have such difficulty themselves with the concept, this strat-
egy involves looking in the wrong place. Is power valid in terms
of the law? Is the law justifiable in terms of the beliefs and
values established in the society? Is there demonstrable evidence
of consent to the given relations of power? All these questions
can in principle be answered from evidence in the public domain.
This is not to say that the answers may not on occasion prove
contradictory, or that the evidence will not need careful interpret-
ation. But the point is that the evidence is available in the public
sphere, not in the private recesses of people's minds. And when
we have answered the questions given above, the further ques-
tion 'do people believe in the legitimacy of a given power?'
becomes redundant.
If the social scientist, then, in considering the legitimacy of a
given power relation or system, is necessarily involved in offering
an assessment, in making a judgement, what kind of judgement
is it, and how does it differ from those of the lawyer and the
philosopher? Unlike the lawyer, the social scientist is interested
in much more than legal validity; he or she is interested in
the normative standing of the power arrangements that the law
validates. Unlike the moral or political philosopher, on the other
hand, the social scientist assesses these arrangements not against
independent or universal criteria of the right or the good, but
against those that pertain within the society in question; he or
she does not assess actions expressive of consent against ideal
14 The Legitimation of Power

conditions or ideal criteria for consent, but in relation to the

conventions of the particular society. Legitimacy for social scien-
tists is always legitimacy-in-context, rather than absolutely, ide-
ally or abstractly. It is this that enables them to give a coherent
account of legitimate power in societies other than their own,
and to assess the degree of legitimacy of political systems far
removed in time and space (Lowenthal, 1979, pp.401-2).
However, although the criteria of legitimacy the social scientist
employs are different from those of the lawyer and the philos-
opher, they are also connected to them, as the account I have
given should make clear. The social scientist will need to extend
the lawyer's criterion of legal validity in the same way as the
philosopher does; he or she will in turn need to contextualise
the criterion of normative justifiability used by the moral or
political philosopher. Yet in each case it will be the same kind
of criteria that will be looked for. Because of this, the account
of legitimacy that I have offered, unlike that of Weberian social
science, can where relevant draw upon different traditions of
legal and political theorising respectively, rather than remain
completely divorced from them. It will also be able to give a
coherent account of the roles of the lawyer and the philosopher
at key moments of uncertainty and dispute about the legal or
moral bases of legitimacy in given historical societies.
Let me sum up the argument so far. The account I have given
seeks to distinguish a properly social-scientific judgement about
legitimacy-in-context from both the misleading Weberian strategy
of reporting people's belief in legitimacy on the one side, and
the normative-philosophical project of elucidating independent
criteria of justifiability, or ideal conditions for consent on the
other. The inadequacies I find in both, from the social-scientific
point of view, are similar to those discussed by Habermas in his
article 'Legitimation problems in the modern state', in which he
rejects both what he calls 'the empiricist' and the 'normativist'
concepts of legitimacy:

'The one can be employed in the social sciences but is unsatis-

factory because it abstracts from the systematic weight of
grounds for validity; the other would be satisfactory in this
regard but is untenable because of the metaphysical context
in which it is embedded.' (1979, p.204)
Social Science and Legitimacy 15

At the same time, he goes on, a Rawlsian approach which seeks

to define the procedural conditions (or ideal speech context)
necessary to a rational consensus cannot provide the basis for
analysing legitimacy in given historical societies:

'Every general theory of justification remains peculiarly

abstract in relation to the historical forms of legitimate domi-
nation. If one brings standards of discursive justification to
bear on traditional societies, one behaves in an historically
"unjust" manner. Is there an alternative to this historical injus-
tice of general theories, on the one hand, and the standardless-
ness of mere historical understanding, on the other?' (ibid.
p.205)

Habermas' own answer to this question - the key question

about legitimacy for the social scientist - involves constructing a
developmental sequence of historical forms of legitimation after
the pattern of cognitive developmental psychology. I find his
solution unsatisfactory, at least as a starting point for an analysis
of legitimacy. This is because, in concentrating on the differences
between different historical forms, rather than also on what they
have in common, Habermas fails to give an account of the under-
lying structure and logic of legitimation in general, which must
form the necessary basis for an exploration of what is historically
variable and specific. My own starting point lies with this under-
lying structure of legitimacy, which I have already sketched out,
and will consider more systematically in the section that follows.
The reader who has experienced difficulty in following the dis-
cussion so far will, I hope, find the issues becoming progressively
clearer as I proceed.

The different dimensions of legitimacy

The key to understanding the concept of legitimacy lies in the

recognition that it is multi-dimensional in character. It embodies
three distinct elements or levels, which are qualitatively different
from one another. Power can be said to be legitimate to the
extent that:
16 The Legitimation of Power

i) it conforms to established rules

ii) the rules can be justified by reference to beliefs shared by
both dominant and subordinate, and
iii) there is evidence of consent by the subordinate to the particu-
lar power relation.

The first level is that of rules; the second that of justifications

grounded on beliefs; the third that of actions. The three levels
are not alternatives, since all contribute to legitimacy; all provide
the subordinate with moral grounds for compliance or cooper-
ation with the powerful. Each, however, is different, and has its
own characteristic form of non-legitimacy. I shall say something
further about each of them in turn.

i) The first and most basic level of legitimacy is that of rules,

corresponding to the legal definition already discussed. Power
can be said to be legitimate in the first instance if it is acquired
and exercised in accordance with established rules. For con-
venience I shall call the rules governing the acquisition and exer-
cise of power the 'rules of power'. These rules may be unwritten,
as informal conventions, or they may be formalised in legal codes
or judgements. Pressure towards formalisation arises in most
societies from the need to resolve disputes about power by
making the rules both precise and strictly enforceable, but there
still remains a considerable role for convention, or 'custom and
practice', even where legal formalisation is well advanced. There
also remains considerable scope for dispute about the law,
though a characteristic feature of legal systems is the presence
of an ultimate authority whose rulings are acknowledged as final.
The opposite of legitimacy according to the rules is, simply,
illegitimacy; power is illegitimate where it is either acquired in
contravention of the rules (expropriation, usurpation, coup
d'etat), or exercised in a manner that contravenes or exceeds
them. The illegal acquisition of power usually has more pro-
found, because more all-pervasive, consequences for legitimacy
than some breach or contravention in its exercise, though that
depends upon the seriousness of the breach, and whether it is
repeated. Where the rules of power are continually broken, we
could speak of a condition of chronic illegitimacy.
Social Science and Legitimacy 17

ii) On its own, legal validity is insufficient to secure legitimacy,

since the rules through which power is acquired and exercised
themselves stand in need of justification. This is the second level
of legitimacy: power is legitimate to the extent that the rules of
power can be justified in terms of beliefs shared by both domi-
nant and subordinate. What kinds of justification and what kinds
of belief are needed? To be justified, power has to be derived
from a valid source of authority (this is particularly true of
political power); the rules must provide that those who come to
hold power have the qualities appropriate to its exercise; and
the structure of power must be seen to serve a recognisably
general interest, rather than simply the interests of the powerful.
These justifications in turn depend upon beliefs current in a given
society about what is the rightful source of authority; about
what qualities are appropriate to the exercise of power and how
individuals come to possess them; and some conception of a
common interest, reciprocal benefit, or societal need that the
system of power satisfies.
No society is characterised by a complete uniformity of beliefs.
Indeed, one of the distinctive features of power relations is the
difference of circumstances, opportunities and values between
dominant and subordinate groups. Yet without a minimum of
the appropriate beliefs defined above being shared between the
dominant and the subordinate, and indeed among the subordi-
nate themselves, there can be no basis on which justifications for
the rules of power can find a purchase. Naturally what counts
as an adequate or sufficient justification will be more open to
dispute than what is legally valid, and there is no ultimate author-
ity to settle such questions; nevertheless clear limits are set by
logic and the beliefs of a given society to what justifications are
plausible or credible within it.
This second level or dimension of legitimacy has its corre-
sponding negative or opposite. Rules of power will lack legit-
imacy to the extent that they cannot be justified in terms of
shared beliefs: either because no basis of shared belief exists in
the first place (e.g. slavery, 'artificial' or divided communities);
or because changes in belief have deprived the rules of their
supporting basis (e.g. hereditary rule or male power, in face of
a declining belief in the superior qualities supposedly ascribed
by birth or sex); or because changing circumstances have made
18 The Legitimation of Power

existing justifications for the rules implausible, despite beliefs

remaining constant (the example of the British electoral system
discussed on pages 11-12). These different situations clearly have
widely differing significance, but they can all be described as
examples, not so much of illegitimacy, as of legitimacy deficit or
weakness.

iii) The third level of legitimacy involves the demonstrable

expression of consent on the part of the subordinate to the
particular power relation in which they are involved, through
actions which provide evidence of consent. As I argued earlier,
the importance of actions such as concluding agreements with a
superior, swearing allegiance, or taking part in an election, is
the contribution they make to legitimacy. They do this in two
ways. The first is that they have a subjectively binding force for
those who have taken part in them, regardless of the motives
for which they have done so. Actions expressive of consent, even
if undertaken purely out of self-interest, will introduce a moral
component into a relationship, and create a normative commit-
ment on the part of those engaging in them. Secondly, such
actions have a publicly symbolic or declaratory force, in that
they constitute an express acknowledgement on the part of the
subordinate of the position of the powerful, which the latter are
able to use as confirmation of their legitimacy to third parties
not involved in the relationship, or those who have not taken
part in any expressions of consent. They are thus often associated
with impressive forms of ceremonial.
Now some theorists would argue that consent is a distinctively
modern component or condition of legitimacy, and one that is
specific to the liberal or individualist tradition; it therefore cannot
be relevant to an understanding of legitimacy in other historical
societies. However, this is to narrow the concept of consent to
the criteria pertaining in an individualist culture: namely, that it
should be linked to an individual choice between alternatives
(e.g. between alternative possible husbands, employers or
governments), and that all adults are in principle qualified to give
consent. The fact that the criterion of choice between alternative
'superiors' does not necessarily pertain in other societies does
not mean that there are not other conventional forms of express-
ing consent appropriate to those societies. And in most historical
Social Science and Legitimacy 19

societies only some among the subordinate have been qualified

to give consent; as I shall show later, the qualification is typically
related to the category of the 'free' in the sphere of social and
economic power relations, and to those who count as members
of the political community in the sphere of politics. Moreover,
in a pre-individualist age it was taken for granted that some adult
males could give consent on behalf of other people (their wives,
children, clients, tenants, even descendants) in a manner that
would be unacceptable in a more individualist culture, unless
those involved were acting in an explicitly representative
capacity. What counts as consent, therefore, and from whom it is
required to confer legitimacy on the powerful, is itself a culturally
specific matter, determined by the conventions of a given society,
rather than definable absolutely. What is common to legitimate
power everywhere, however, is the need to 'bind in' at least the
most significant members among the subordinate, through actions
or ceremonies publicly expressive of consent, so as to establish
or reinforce their obligation to a superior authority, and to dem-
onstrate to a wider audience the legitimacy of the powerful.
It is in the sense of the public actions of the subordinate,
expressive of consent, that we can properly talk about the 'legi-
timation' of power, not the propaganda or public relations cam-
paigns, the 'legitimations' generated by the powerful themselves.
And if the public expression of consent contributes to the legit-
imacy of the powerful, then the withdrawal or refusal of consent
will by the same token detract from it. Actions ranging from
non-cooperation and passive resistance to open disobedience and
militant opposition on the part of those qualified to give consent
will in different measure erode legitimacy, and the larger the
numbers involved, the greater this erosion will be. At this level,
the opposite or negative of legitimacy can be called delegi-
timation.

For power to be fully legitimate, then, three conditions are

required: its conformity to established rules; the justifiability of
the rules by reference to shared beliefs; the express consent of
the subordinate, or of the most significant among them, to the
particular relations of power. All three components contribute
to legitimacy, though the extent to which they are realised in a
given context will be a matter of degree. Legitimacy is not an
20 The Legitimation of Power

ali-or-nothing affair. For this reason I used the formula above:

power can be said to be legitimate to the extent that ... etc. Every
power relation knows its breaches of the rules or conventions; in
any society there will be some people who do not accept the
norms underpinning the rules of power, and some who refuse to
express their consent, or who do so only under manifest duress.
What matters is how widespread these deviations are, and how
substantial in relation to the underlying norms and conventions
that determine the legitimacy of power in a given context. Legit-
imacy may be eroded, contested or incomplete; and judgements
about it are usually judgements of degree, rather than ali-or-
nothing.
Above all, the analysis I have given above demonstrates that
legitimacy is not a single quality that systems of power possess
or not, but a set of distinct criteria, or multiple dimensions,
operating at different levels, each of which provides moral
grounds for compliance or cooperation on the part of those
subordinate to a given power relation. By the same token, power
can be non-legitimate in very different ways, which I have sig-
nalled by the different terms: illegitimacy, legitimacy deficit and
delegitimation. The erosion of justificatory norms, slavery, con-
quest, dictatorship, coup d'etat, separatist agitation, revolution-
ary mobilisation - all are examples where power lacks some
element of legitimacy, but does so in very different ways. The
accompanying diagram summarises in tabular form the different
dimensions of legitimate and non-legitimate power that I have
distinguished, to reinforce the argument of the text.

Table 1.1 The three dimensions of legitimacy

Criteria of Legitimacy Form of Non-legitimate Power

conformity to rules (legal illegitimacy (breach of rules)
validity)
11 justifiability of rules in terms of legitimacy deficit (discrepancy
shared beliefs between rules and supporting
beliefs, absence of shared
beliefs)
111 legitimation through expressed delegitimation (withdrawal of
consent consent)
Social Science and Legitimacy 21

In analysing legitimacy into its component elements, I am

claiming that these constitute basic criteria for legitimacy in all
historical societies, past and present. In other words, if we want
to know what makes power legitimate anywhere, it is to these
criteria that we must look. Such a claim can only ultimately be
vindicated by the persuasiveness of the explanatory analysis that
follows in the rest of the volume. Here it is important to clarify
what this claim does, and does not, entail.
At first sight it might seem that the claim that there are criteria
of legitimacy universally applicable must contradict what I said
earlier about the social scientist's task being to assess the legit-
imacy of power in its context, i.e. against the norms and values
of a given society. In fact there is no contradiction. The criteria
distinguished above constitute only the most general framework,
the specific content or substance of which has to be 'filled in' for
each historical society. Is power valid according to the rules?
The relevant rules have to be specified, their conventional or
legal form established, the mode of adjudication pertinent to the
given society determined, and so on. Are the rules justifiable in
terms of the beliefs and norms of the particular society? Again
the content of these beliefs will have to be specified; but the
analysis given above, and elaborated later in Chapter 3, provides
a guide to what we must look for: beliefs about a) the valid
source of authority (tradition, the people, divine authorisation,
etc.); b) how people come to possess the qualities appropriate
to the exercise of power (heredity, achievement, etc.); c) the
ends that power should serve (variable according to gender, class,
political power, etc.). Moreover the form in which these beliefs
are reproduced and authenticated will also vary systematically
from one age or society to another: myths and story telling,
divine revelation, philosophical argument, scientific 'proof'. Is
there, finally, evidence of expressed consent on the part of those
qualified to give it? Again, who counts as qualified, and what
actions count as appropriate, will be determined by the conven-
tions of the given society or system of power.
What we have, then, is a set of general criteria for legitimacy,
the specific content of which is historically variable, and must
therefore be determined for each type of society. This combi-
nation of a general category, with variable and historically spec-
ific forms, is a common feature of comparative analysis in the
22 The Legitimation of Power

social sciences. Consider for example the Marxian concept of

class. Marx was insistent that it was the historical differentiation
between different forms of class society (slave, feudal, capitalist)
that was crucial to understanding the specific dynamic of each.
But the differentiation was only possible in the first place by
being based upon a theoretically elaborated general concept of
class, as a relationship between the direct producers and the
owners of the means of production, in which the latter appropri-
ated a surplus from the labour of the former. An understanding
of this general category, according to Marx, provided the key to
the analysis of class relations in each historically specific mode
of production (Marx, 1966, vol.3, pp.790-2).
A closer parallel would be that provided by Barrington Moore
in his book Injustice, where he identifies, beneath the variability
of social norms and arrangements in different historical periods,
'recurring elements' in their moral codes (Moore, 1978, ch.1).
In a similar manner I am proposing that there is an underlying
structure of legitimacy common to all societies, however much
its content will vary from one to the other. What explains this
common structure? As I shall show more fully later, it has its
origin in the different ways in which power that is not legitimate
offends our moral sense; in an underlying logic common to moral
argument everywhere, however diverse its actual content; and in
the needs that are shared by all societies, however varied the
social relations and organisational arrangements by means of
which they are met. It is these common features - of human
beings as rational moral agents seeking to ensure that their social
relations and arrangements meet their needs and conform to
their moral sense - that provide the underlying structure to
legitimacy everywhere.
The structure that I have so far merely outlined, comprising
rule-conformity, the justifiability of the rules in terms of shared
norms and beliefs, expressed consent on the part of those quali-
fied, provides the social scientist with the framework to under-
take two different tasks. The first is a systematic comparison
between different forms of legitimacy appropriate to different
historical types of social and political system. The assumption
made here is that rules or arrangements of power embody justifi-
catory beliefs or norms, and conventions about consent, even if
they may come to diverge from them over time, or the beliefs
Social Science and Legitimacy 23

themselves become atrophied. Here is the second task which the

above structure provides the basis for: that of assessing the
degree of legitimacy-in-context of a given power relationship, as
a necessary element in explaining the behaviour of those involved
in it. This is the immanent judgement or assessment of the social
scientist, as opposed to the normative philosopher's assessment
of a power relationship against independent, external standards
or criteria. And this judgement also becomes an immanent cri-
tique when the social scientist is able to show the features internal
to a system of power that, on the one hand, sustain and repro-
duce its legitimating beliefs, or, on the other, systematically
undermine them over time.
The value of the multi-dimensional conception of legitimacy
in facilitating the fulfilment of such an agenda stands in marked
contrast to the Weberian conception considered in the previous
section. Although Weber himself made a modest contribution to
the first part of the agenda outlined above, the conception of
legitimacy that he bequeathed to twentieth-century social science
was inadequate for the purpose, and quite useless for the second,
more critical, task of analysis and explanation. This is because
the conception that power is legitimate if people believe it to be
so, is mistaken on a number of grounds, which can now be
summarised. It reduces legitimacy from a complex of factors
which give people good grounds for compliance, to a single
dimension: their 'belief in legitimacy'. It misconceives the
relationship between legitimacy and the beliefs that provide the
justificatory basis for rules of power. It fails to recognise that,
although prudential and normative reasons for obedience are
indeed distinct, nevertheless, people's interests can be harnessed
to legitimacy through actions expressive of consent. Finally, it
leaves the social scientist helpless in the critical task of analysing
an erosion of legitimacy in power relations, by proposing a report
on people's 'belief in legitimacy' rather than identifying a
developing discrepancy between rules of power and the norms
that provide their justification.
I said earlier that Weber's definition of legitimacy was only the
starting point of what was wrong with his theory. The problem
of his definition works through into his threefold typology of
legitimate authority: traditional, rational-legal and charismatic
(Weber, 1968, pp.215-16). If I also continue my critique of
24 The Legitimation of Power

Weber at this point it is not because of any unreasoned animus

on my part, but because of my conviction that it is necessary for
social science to be freed from the whole Weberian legacy if it
is to make sense of the subject of legitimacy. Social scientists
since Weber have, if anything, been even more transfixed by his
threefold typology than by his definition of legitimacy itself;
indeed it has become a straightjacket into which, either singly
or in combination, every example of legitimate power has, willy-
nilly, to be forced.
True to his definition of legitimacy as the 'belief in legitimacy',
Weber makes each of his types of legitimate authority - tra-
ditional, rational-legal, charismatic - dependent upon a different
type of belief: in the sanctity of tradition, in rule-conformity and
procedural correctness, and in the charismatic qualities of the
individual leader, respectively. What is wrong with this, to put
a complex matter briefly, is that it elevates each of the three
contributory components of legitimacy, that I have distinguished
above, into a separate and fully self-sufficient type of legitimacy
(see Beetham, 1991). Thus, the first level of legal validity
becomes a 'rational-legal' type, based upon a belief in rule-
conformity and procedural correctness; in the process it becomes
detached from any substantive beliefs or principles in relation to
which the legal rules and procedures can be justified. Weber's
traditional type, on the other hand, represents one example of
a second level basis for the justification of rules (belief in the
sanctity of the past); but the contrast with the rational-legal type
obscures its true status, and conveys the misleading impression
that power in a traditional order is not validated in terms of
rules at all. At the same time the typology offers no account of
the beliefs that have replaced traditionalism as a second-level
basis for justifying rules of power in the modern world. The
charismatic type, thirdly, represents the rare case of a legitimacy
deriving solely from consent, in the absence of rules or justifying
beliefs, a consent expressed in the act of acknowledging and
following a leader. However, Weber's typology once more
obscures its character, by making the basis of legitimacy not the
act of recognition by a following, but their subjective belief in
the 'charisma' of the leader, a concept that has had a particularly
confused and unfortunate career.
In each case, Weber's typology represents the elevation of a
Social Science and Legitimacy 25

different level of legitimacy into a self-sufficient type, in a way

that obscures the status of each as but one element in a totality.
Once this process is understood, it becomes clear why the
typology has proved both plausible to later social scientists, and
at the same time such a potent source of confusion. Its plausi-
bility derives from the fact that the typology embodies three
genuine components of legitimacy; the confusion from the fact
that it does so in a distorted form, whereby the proper status of
each component has become obscured. True to Weber's concep-
tion of legitimacy as constituted by subjective beliefs, each
element with its distinctive characteristics has been transposed
into a different type of belief: in rules and procedures, in tra-
dition, and in charisma respectively.
Other critics of Weber have had a sense of unease about his
threefold typology, and the concept of charismatic authority in
particular has met with repeated objections (e.g. Friedrich, 1961;
Wolpe, 1968; Bensman and Givant, 1975). Yet the typology
continues to hold the field, with or without qualification, because
the source of Weber's error has not been clearly identified, and
no convincing alternative has been proposed. I hope I have said
enough, if not more than enough, to convince the reader that
there is indeed an alternative approach to the subject; and that
the whole Weberian theory of legitimacy has to be left behind
as one of the blindest of blind alleys in the history of social
science, notable only for the impressiveness of the name that it
bears, not for the direction in which it leads. That at any rate
is the course that I propose to follow in the remainder of this
book, in which the name of Weber will hardly be mentioned
again. The final introductory task will be to explain more fully
why the analysis of legitimacy, such as I have offered, is so
important to an understanding of power relations.

The significance of legitimacy

Legitimacy, as we have seen, comprises the moral or normative

aspect of power relationships; or, more correctly, the sum of
these aspects. A social-scientific analysis of legitimacy is con-
cerned with the effect it has on the character of a given relation-
ship, and on the behaviour of those involved in it. It is the
26 The Legitimation of Power

importance of legitimacy- its character and degree - to explain-

ing people's behaviour that concerns the social scientist.
To consider first the behaviour of those subordinate within a
power relationship: its legitimacy provides them with moral
grounds for cooperation and obedience. Legitimate power or
authority has the right to expect obedience from subordinates,
even where they may disagree with the content of a particular
law or instruction; and subordinates have a corresponding obli-
gation to obey. This obligation is not absolute - hence the
dilemmas that occur when people are required by a legitimate
superior to do things that are morally objectionable to them, as
opposed to inconvenient or merely stupid. But it is the right
that legitimacy gives those in authority to require obedience
in principle, regardless of the content of any particular law or
instruction, that makes it so important to the coordination of
people's behaviour in all spheres of social life.
The legitimacy or rightfulness of power, then, provides an
explanation for obedience through the obligation it imposes on
people to obey, and through the grounds or reasons it gives for
their obedience. I emphasise 'grounds or reasons', because there
are numerous psychological studies of obedience which explain
it in terms of acquired attitudes and characteristics, such as the
internalisation of a respect for authority, which is confirmed by
continuous symbolic reinforcement (e.g. Merelman, 1966; Mil-
gram, 1974). The problem with such explanations, which reduce
obedience to the sum of attitudes and characteristics developed
and internalised over time, is that they fail to explain why people
stop obeying, whenever they do so; or else they have recourse
to considerations of the psychological 'stress' or 'dissonance' to
which people are subjected. Yet people only disobey because
they have sufficient and compelling reasons for disobedience in
particular contexts; and such reasons will turn out to be precisely
the counterpart to the reasons they have for not stepping out of
line in normal times or situations. Psychological explanations for
obedience only carry plausibility when they are in fact disguised
accounts of people's reasons; or when we are driven to appeal
to non-rational factors because rational ones will not suffice on
their own.
However, normative grounds or reasons are not the only
reasons people have for obedience. As I shall set out more fully
Social Science and Legitimacy 27

in the next chapter, power relations are almost always constituted

by a framework of incentives and sanctions, implicit if not always
explicit, which align the behaviour of the subordinate with the
wishes of the powerful. They do so by giving people good reasons
of a different kind, those of self-interest or prudence, for not
stepping out of line. Obedience is therefore to be explained by
a complex of reasons, moral as well as prudential, normative as
well as self-interested, that legitimate power provides for those
who are subject to it. This complexity may make it difficult to
determine the precise balance of reasons in any one situation;
but it is important to distinguish them analytically, since each
makes a very different kind of contribution to obedience.
The fact that power relations typically involve a framework of
incentives and sanctions gives plausibilty to a 'realist' or 'organis-
ational' view of power, which holds that obedience is only a
matter of the resources available to the powerful to ensure com-
pliance with their wishes, and that legitimacy is irrelevant: an
issue to be debated by moral philosophers, perhaps, but of no
importance to an explanatory account of obedience, and there-
fore of no interest to the social scientist (Skocpol, 1979; Mann,
1986). This sceptical view is also to be found in the currently
fashionable 'rational choice' approach, whose explanatory force
depends upon the assumption that social action is to be explained
by the agents' calculations of their own self-interest (e.g. Taylor,
1988). What such an account leaves out is obvious to all but the
most hardened exponents of the theory: that people are also
moral agents, who recognise the validity of rules, have some
notion of a common interest, and acknowledge the binding force
of promises they have made - all elements involved in legitimate
power. To explain all action conforming to rules as the product
of a self-interested calculation of the consequences of breaching
them, is to elevate the attributes of the criminal into the standard
for the whole of humankind, and to make a prison regime into
the paradigm case of power. People relate to the powerful as
moral agents as well as self-interested actors; they are coopera-
tive and obedient on grounds of legitimacy as well as for reasons
of prudence and advantage. It is the task of a theory of legitimacy
to identify as clearly as possible what its distinctive place in this
complex is.
One way of doing this is to see what happens to a power
28 The Legitimation of Power

relationship when legitimacy is eroded or absent. In such a case,

power does not necessarily collapse, or obedience cease, since it
can continue to be kept in place by incentives and sanctions.
However, coercion has to be much more extensive and omnipres-
ent, and that is costly to maintain. Moreover, the system of
power now has only one line of defence, that of force; and it
can therefore collapse very rapidly if coercion is insufficient or
people believe that those in power have lost the will to use it.
Once Gorbachev made clear that the USSR would no longer
intervene militarily in Eastern Europe, the writing was on the
wall for the communist regimes there, since they were only kept
in place by the ultimate threat of Soviet invasion. Only the timing
and manner of their demise was unpredictable. To admit the
role of force in this situation is not to acknowledge the validity
of the 'realist' theory of power. What has to be explained is the
prior loss of legitimacy of communist rule in Eastern Europe,
which made it so reliant on coercion, and therefore so vulnerable
once sufficient coercion could no longer be guaranteed.
The collapse of authority where legitimacy is eroded, and
coercive force is insufficient to maintain power on its own, pro-
vides only the most dramatic evidence for the significance of
legitimacy to the obedience of subordinates. Less dramatic, but
equally important, is the effect a lack of legitimacy has on the
degree of cooperation, and the quality of performance, that can
be secured from them, and therefore on the ability of the power-
ful to achieve goals other than simply the maintenance of their
position. Where the powerful have to concentrate most of their
efforts on maintaining order, they are less able to achieve other
goals; their power is to that extent less effective. The classroom
teacher provides a typical example. If pupils do not share a belief
in the value of education, on which the justification for the
teacher's power is based, or have no respect for the individual
teacher, he or she will have to devote correspondingly greater
energies to maintaining order than to teaching. To that extent
the purposes for which power is held will not be achieved, and
this may lead in turn to a further erosion of legitimacy.
The kind of vicious circle in which the powerful can become
trapped when their legitimacy is eroded is well illustrated from
the attempts at economic reform under communist rule in East-
ern Europe. These repeatedly failed, in part at least because
Social Science and Legitimacy 29

governments had insufficient legitimacy to demand the short term

sacrifices, or risk instituting the price rises, necessary to the
development of a more market-oriented system (Lewis, 1982,
pp.137-8; MacFarlane, 1984, pp.177 ff.). And the failure of eco-
nomic reform in turn further eroded the system's legitimacy. The
point was well made by President Jaruzelski in an interview in
May 1990:

'The party could not mobilise Polish forces into constructive

channels ... We tried economic reforms time and time again.
But they always met with public resistance and explosions. It
is very different now. Now with a government that enjoys
public confidence, it is possible to demand sacrifices.' (The
Guardian, 7 May 1990)

Without the legitimacy to demand sacrifices, the ruling party's

power over society became a largely negative one: able to control
the population in the sense of preventing them doing what they
wanted, but not in the sense of securing the cooperation neces-
sary to the achievement of the government's policies.
These examples of the loss of moral authority, from the class-
room to the state, indicate that legitimacy is significant not only
for the maintenance of order, but also for the degree of cooper-
ation and quality of performance that the powerful can secure
from the subordinate; it is important not only for whether they
remain 'in power', but for what their power can be used to
achieve. Passive non-cooperation, work to rule, feigned incom-
petence, 'looting', and so on: these attributes of the Good Soldier
Schweik and the 'worker in a workers' state' are typical of subor-
dinates where legitimacy has become eroded, and will to that
extent reduce the capacity of those in power to achieve their
goals (Hasek, 1973; Haraszti, 1977). The effectiveness of the
powerful, in other words, is not just a matter of resources and
organisation, as the 'realists' would contend, but also of their
legitimacy. The realists are at this point simply not realistic
enough; they do not take people seriously as moral agents, or
recognise that what the powerful can get others to do depends
upon normative considerations as well as upon I
the resources and
organisational capacities at their command.
Wherever the goals of the powerful are dependent upon the
30 The Legitimation of Power

degree of cooperation and the quality of performance on the

part of subordinates, therefore, to that extent is legitimacy impor-
tant for what they can achieve as well as for the maintenance of
their power. It follows that legitimacy is more crucial to some
kinds of power relationship than to others. It is worth exploring
in a preliminary way what these might be. We could distinguish
two different kinds of situation where the legitimacy of a power
relationship is unnecessary to the goals of the powerful.
The first is a labour regime where continuous work is required
from subordinates, but the quality of their performance is unim-
portant, and they can be treated as dispensable because there is
a ready supply of replacements available. Such was the position
in most historical examples of slavery, where the supply of slaves
was repeatedly replenished through conquest and trade. A close
parallel was early industrial capitalism, where the work required
little skill from the worker, and an unlimited supply of new
recruits was available from the labour market to expand pro-
duction or replace those incapacitated by illness, accident or
death. Both these examples of coercive labour regimes lacked
legitimacy from the outset, though the source of their coercion
differed (physical force, economic duress). In the case of capital-
ism, the increasing need to secure quality of performance, on the
one side, and the pressure of self-organisation by the workers, on
the other, necessitated the development over time of a more
legitimate relationship, based upon the employers' recognition
of collective bargaining rights and the negotiation of collective
agreements over the terms and conditions of work (Fox, 1985,
ch.4).
In the examples of slavery and early industrial capitalism the
absence of legitimacy in the relationship was irrelevant to the
performance of subordinates. At this point, however, we need
to observe an important distinction. The fact that the relationship
depended on coercion did not mean that the powerful did not
have a legal basis for their power, or that they did not seek to
justify it to themselves. Indeed, both slavery and early capitalism
witnessed the most elaborate justifications, derived, in the one
case, from Aristotelian notions of a slave 'nature' (Aristotle,
1962, pp.32-4) which was elaborated by later racial theories,
and, in the other, from the doctrines of classical political econ-
omy. Such 'legitimations', however, were addressed to the con-
Social Science and Legitimacy 31

science of the powerful, not at all to their subordinates. To

address the latter was not only practically difficult, given that
they often shared no common language. It was also logically
impossible, since the justification for treating slaves and early
industrial workers in the way they were treated was that they
belonged to a category of objects, and were therefore by defi-
nition incapable of being addressed 'as persons'. Slaves were
chattels, wholly owned by the slaveowner (Aristotle, 1954,
p.212). Industrial workers were 'hands', whose labour power was
owned as a commodity, and, like any other commodity, obeyed
the laws of supply and demand in the marketplace; if it was
overproduced, the numbers would simply be cut by the necessary
forces of want, disease and starvation (Smith, 1976, pp.89-90;
J.S. Mill, 1909, pp.343-60). In other words, we must distinguish
between the 'legitimations' that the powerful develop to reconcile
their consciences to the treatment (and maltreatment) of their
subordinates, and a legitimate relationship, justified in terms of
shared beliefs, regulated according to understood conventions
and confirmed through the expression of consent.
If the characteristic of coercive labour regimes was that they
did not require legitimacy for the level of performance needed
in the context, we could contrast them with other labour regimes
where quality of performance is crucial. An example from the
other end of the spectrum is that of gender relations in most
historical periods, where women's domestic work has demanded
qualities of skill, dedication and independent initiative that could
only be obtained from subordinates within a legitimate relation-
ship. Here the beliefs that provide the justification for the hier-
archical division of labour and the definition of a common
interest have been shared between men and women, and the
legitimacy of the individual relationship has been confirmed by
express consent. In the case of gender relations the distinction
between the legitimacy of the respective roles and the motivation
necessary to perform them is a fine one, and the account of
legitimacy will need supplementing with reference to the socio-
psychical processes whereby attitudes appropriate to given roles
become so deeply internalised that they appear as 'natural' (see
below, pp.78-9). Important to note here, however, is simply the
point that the quality of performance needed from the subordi-
32 The Legitimation of Power

nate party in a relationship, and the degree of legitimacy the

relationship requires, are closely connected.
One kind of situation, then, where the legitimacy of a power
relationship does not matter is where the quality of work perfor-
med by subordinates is unimportant. A second is where the
relationship between dominant and subordinate is so distant, or
indirect, that little is required by the one of the other. In many
pre-modern states it made little difference what the beliefs of
the vast majority of the population were, and their consent to
the state was irrelevant, since their obligations were to purely
local chiefs or superiors; it was the relationship between the
central state and local power-brokers that was all-important. The
contemporary state, in contrast, requires of its whole adult popu-
lation a general obligation to pay taxes and be available for
military service if needed; and there is a variety of situations in
which the cooperation of different sections of the population is
essential to the realisation of government policy. The legitimacy
of the relationship becomes even more critical where, as in a
command economy, the state is also the main employer of
labour, and economic performance itself is directly dependent
upon the state's legitimacy. One of the chief weaknesses of the
communist system has been that the state's requirement for legit-
imacy has been correspondingly greater than that of the capitalist
state, but at the same time it has been less able to sustain it
than states in at least the advanced capitalist societies, for reasons
that will be explored later in the book. Here it will be sufficient
to emphasise the qualitative difference between the erosion of
legitimacy where it matters, as in a communist regime, and the
absence of legitimacy where it doesn't, as in the slave systems
discussed above. One consequent difference is the repeated
attempt to reestablish a basis of legitimacy in the former; and
the danger of regime collapse, as opposed to merely revolt or
rebellion, in the event of failure to do so.
In considering the different levels of performance that are
required from those in different subordinate positions, an obvi-
ous distinction can be drawn between those who staff the admin-
istrative and coercive apparatuses of the state, and the population
as a whole. In view of the quality of performance and degree of
commitment required from the former if the state organisation
is to function effectively, considerable pains will be taken to
Social Science and Legitimacy 33

reinforce their support for the norms of the regime, and to bind
them to it with special contractual commitments and oaths of
allegiance. However, it is mistaken to conclude from this, as
some writers have done, that the legitimacy of government is
therefore chiefly of consequence for the members of the state
apparatus, or the political elite, and has little relevance for the
population as a whole (Therborn, 1980, p.109; Bialer, 1980,
pp.194-5). In the modern period, at least, it has been repeatedly
shown that the state apparatus cannot be immunised from an
erosion of regime legitimacy that has affected the attitudes and
behaviour of the rest of the population; and that the refusal of
the masses to do as they are told will provoke a corresponding
crisis of obedience among the armed forces that arc ordered to
discipline or crush them. The collapse of the Shah's regime in
Iran in 1979 is only the most spectacular recent example of the
erosion of an army's loyalty in the face of repeated mass protest.
It provides convincing evidence that the legitimacy of a regime
is as crucial to its effectiveness as the competence of its adminis-
trators or the firepower of its armies, and that the strength of
an organised power structure is dependent in the last analysis
upon the readiness of subordinates to obey orders under pres-
sure.
Enhanced order, stability, effectiveness - these are the typical
advantages that accrue to a legitimate system of power as a
result of the obligations upon subordinates that derive from its
legitimacy. 'Order' depends upon people obeying rather than
disobeying. 'Stability' is not mere longevity, but a system's ability
to withstand shock and failure because a solid level of support
from its subordinates can be guaranteed. 'Effectiveness' includes
the ability of the powerful to achieve their goals because of the
quality of performance they can secure from those subordinate
to them. Legitimacy is not the only factor contributing to the
order, stability and effectiveness of a system of power; organis-
ational capacities and resources are obviously crucial as well.
Moreover, as we have seen, it is possible in certain situations
for the goals of the powerful to be realised on the basis of
coercion alone. Yet legitimacy makes its distinctive contribution
to achieving these other qualities through the effects it has on
the attitudes and behaviour of the subordinate as moral agents,
not just as self-interested actors.
34 The Legitimation of Power

I have spent a good deal of time identifying the effects of

legitimacy, because there are so many different elements
involved in a power relationship, and the interaction between
them is enormously complex and easy to misread. Many political
scientists confuse legitimacy with regime-stability, or define it as
simply a by-product of effective system-functioning (e.g.
Luhmann, 1969). This conceptual conflation, which equates legit-
imacy with the consequences it produces, can only be avoided if
we are able to give a clear account, not only of what legitimacy
is, but of how it produces the consequences it does through the
obligations that subordinates derive from it. The accompanying
diagram (Figure 1.1) sets out in summary form the different
steps I have distinguished in my account of the consequences of
legitimacy, first for the behaviour of subordinates, and then in
turn for other characteristics of the system of power.

Figure 1.1 Characteristics of a power system or relationship

legitimacy enhanced order,

t
(validity, justifiability, consent) stability, effectiveness

+
moral grounds particular quality
for compliance of compliance
subordinates

So far I have discussed the consequences of legitimacy for the

behaviour of the subordinate within a power relationship. I now
turn more briefly to consider its consequences for the powerful.
If legitimacy, as I have argued, enhances the order, stability and
effectiveness of a system of power, then we should expect that
the powerful will seek to secure and maintain the legitimacy of
their power, in view of its advantages to them. Here again,
however, we must be careful to avoid drawing the wrong con-
clusions from a mistaken definition of legitimacy. If we reduce
it to people's 'belief in legitimacy', then we are likely to conclude
that the way in which the powerful maintain their legitimacy is
primarily by means of ideological work, and through the influ-
ence they have over the beliefs and ideas of the subordinate.
'Every power seeks to establish and cultivate a belief in its
legitimacy,' wrote Weber; and many have drawn the conclusion
Social Science and Legitimacy 35

from this statement that it is precisely by cultivating beliefs that

legitimacy is maintained (Weber, 1956, p.157; 1968, p.213).
I do not wish to discount altogether the role of ideological
work, particularly in reinforcing the basic norms that underpin
a given system of power, though I shall want to argue later
that the processes involved are complex ones, and have been
oversimplified in much of the relevant literature. What I would
emphasise at this point, however, is that we need to look quite
elsewhere for the effect of legitimacy on the behaviour of the
powerful. If legitimate power is, as I have argued, power that is
valid according to rules, and where the rules themselves are
justifiable by and in conformity with underlying norms and
beliefs, then the main way in which the powerful will maintain
their legitimacy is by respecting the intrinsic limits set to their
power by the rules and the underlying principles on which they
are grounded. Legitimate power, that is to say, is limited power;
and one of the ways in which it loses legitimacy is when the
powerful fail to observe its inherent limits.
What are these limits? I would draw attention to two different
kinds. One kind of limit is set by the rules which determine what
the powers of the powerful are, and what they can rightly expect
those subordinate to them to do- which specify, in other words,
the respective duties and obligations of those involved in a power
relationship. These rules may be largely conventional, or they
may be legally defined. A feature of the modern world is the
increasingly precise legal specification of the respective powers,
or 'sphere of competence', of each powerholder. Even today,
however, there is still considerable room left for 'custom and
practice', for conventional understandings built up over time
through processes of struggle and compromise, which govern the
expectations of the powerful and the subordinate about what is,
and is not, required of them; what can, and cannot, legitimately
be demanded.
For the powerful to breach these rules in a substantial way, say
by imposing some new or additional obligation on subordinates
without warning or consultation, is either to invite action for
legal redress, or, where the law is silent, to provoke informal
protests which may develop into a more widespread crisis of
legitimacy for the system of power. Unless they are arrogant or
stupid, powerholders will only take such action when it is essen-
36 The Legitimation of Power

tial to some important purpose, or if they are driven to it by a

serious predicament of their own. The fact that mostly they do
not do so, and that they mostly respect the rules and conventions
governing their relations with those subordinate to them, makes
it easy to overlook an essential feature of legitimacy: that it sets
limits to the behaviour of the powerful as well as imposing
obligations on the subordinate. Because we more readily notice
what the powerful do than what they refrain from doing, this
essential feature of legitimacy tends to go unremarked.
The other kind of constraint which their need to maintain
legitimacy imposes on the powerful is a more fundamental one:
to respect the basic principles that underpin the rules or system
of power, and to protect them from challenge. Rulers who derive
their legitimacy from a divine source must respect religious tra-
ditions and defer to religious authorities; they will regard any
threat to religion or religious belief as among the most serious
they face. Those who derive their authority from the people will
ignore at their peril any insistent and widespread popular current
of opinion; to be seen to favour foreign interests at the expense
of national ones will do more damage to their standing than
almost anything else. Those who claim a monopoly of represen-
tation of the working class by virtue of a privileged knowledge
of their interests cannot afford to allow independent sources of
working-class opinion to find expression, or alternative insti-
tutions of representation to develop, which might challenge their
monopoly. The legitimating ideas and justificatory principles that
underpin the given institutions of power define which challenges
the ruler has to take most seriously, because they strike at the
basis of the system of rule itself (Rothschild, 1977, pp.490-1).
To understand the limits inherent in a system of legitimate
power, and the conditions necessary to its maintenance, is to
identify the distinctive character of the rules governing it, and
the justificatory principles which underpin them. Here we find
that the study of legitimacy is not only a study of the difference
that it makes to the behaviour of the people involved, by its
presence or absence, or according to its degree. It is also a study
of the consequences that the particular form of legitimacy - its
source or principle of justification, its criteria of consent, and so
on - has for the character of power relations. Whether access to
power is determined by birth or merit, by ascription or achieve-
Social Science and Legitimacy 37

ment, whether the ultimate source of rules is located in tradition

or the 'people', whether consent is defined in individualistic or
more collective terms: all these have important implications for
the character, the organisation and the institutions of power.
A basic assumption I shall make is that rules and institutions
of power embody legitimating ideas or justificatory principles
within them, whether this happens because they are consciously
created at a particular moment, or develop through a lengthy
process of historical evolution. This assumption does not imply
an 'idealist' theory about the logical or historical priority of ideas.
As I shall argue in the succeeding chapters, analysing power
involves understanding the interaction between three different
elements: the material and other means of power; the rules of
its social organisation; its justificatory principles or ideas. Rules
of power are shaped by both material means and legitimating
principles or purposes, albeit in a historically variable manner;
any opposition therefore between an 'idealist' and a 'materialist'
theory of power is in my view misconceived.
If it is correct that systems of power embody justificatory
principles or assumptions within them, then it is possible to
develop typologies of power systems according to the main differ-
ences of principle they embody. One axis of differentiation con-
cerns the distinctive social purposes they serve, which will differ
for gender, class and political power respectively. A second line
of differentiation concerns the different principles governing
access to property or political power, and the form of consent
appropriate to each, which will differ between different historical
periods and different types of economic and political system.
Besides an analysis of legitimacy in general, therefore, and of
its consequences for people's behaviour, this book will also con-
cern itself with a comparative study of the justificatory principles
and conventions of consent embodied in different rules or sys-
tems of power.

Conclusion: a social or political concept?

In this chapter I have argued for a social-scientific conception of

legitimacy that differs, on the one side, from the normative
philosopher's search for independent criteria of legitimacy or
38 The Legitimation of Power

ideal conditions for consent; and, on the other, from a mistaken

Weberian conception which reduces the legitimacy of power to
people's belief in its legitimacy. It differs from the latter in that
it involves a judgement about a given system of power, not a
report on what people believe about it; it differs from the former
in that the judgement is a judgement of legitimacy-in-context,
assessed against the relevant norms, principles and criteria for
consent pertaining in the given society. In identifying what is
relevant, and therefore what to look for in the particular society,
the social scientist is guided by the general criteria I have eluci-
dated, which together provide sufficient grounds for obedience.
Legitimacy is important, I have argued, because of its conse-
quences for behaviour, and for the character of power relations.
Against those who hold that the obedience of the subordinate
can be sufficiently procured by incentives and sanctions, I main-
tain that the normative commitments that derive from legitimacy
ensure a distinctive level and quality of compliance and cooper-
ation, though this level is more essential in some contexts than
others. Against those who equate legitimacy with stability or
efficiency, I argue that legitimacy should not be confused with the
effects it produces on a system of power through the enhanced
obedience of its subordinates. Finally I have shown that legit-
imate power sets limits to the powerful through the normative
expectations and principles it embodies, and that an analysis
of these norms and principles is essential to understanding the
distinctive character and institutions of a given system of power.
Throughout I have been critical of realist or sceptical theories
of power, which I have shown to be based on either a faulty
analysis of power, or an impoverished conception of human
nature. If power is one person's ability to achieve their purposes
through others, then it cannot be a matter of capacities and
resources alone, but also depends on the degree of the others'
willingness to cooperate. And that willingness cannot be suf-
ficiently created by incentives and sanctions on their own; it
depends on the normative status of the power holder, and on
normative considerations that engage us as moral agents. This
normative status derives from the character of a legitimate power
relationship as legally valid, justifiable according to shared norms
and beliefs, and confirmed through actions expressing consent.
Above all, therefore, I have been critical of accounts which
Social Science and Legitimacy 39

reduce legitimacy to the sum of 'legitimations' that the powerful

can get the subordinate to accept. Legitimacy is not the icing on
the cake of power, which is applied after baking is complete,
and leaves the cake itself essentially unchanged. It is more like
the yeast that permeates the dough, and makes the bread what
it is.
A final question remains to be answered in this introductory
chapter. Is legitimacy a specifically political concept, which
should properly only be applied to political power? So some
theorists have argued (e.g. Habermas, 1979, p.179), while others
have sought to narrow its application still further, to the sphere
of post-medieval politics, on the grounds that only in the modern
state have the issue of political obedience and the contestability
of legitimacy become persistent features of public life (e.g.
Hennis, 1976, pp.26-7). I would argue against both attempts at
narrowing the applicability of the concept. Political philosophy
may only flourish where legitimacy is contested, and social sci-
ence may only appreciate its importance from the experience of
situations where it is absent or under challenge. Disruptions of
order provide a necessary stimulus to both kinds of intellectual
activity. It does not follow, however, that legitimacy is only
significant where it is most noticeable and most contested. Indeed
it is precisely disorder and instability that demonstrate its neces-
sary contribution to order in more settled, not to say stagnant,
times.
If the concept of legitimacy cannot be restricted to the post-
medieval world, neither can it be limited to the sphere of politics
as such. It should be evident from everything that I have said
in this chapter that it is power itself that morally stands in need
of legitimation, though not every form of power requires it in
practice, and by no means all achieve it. All societies find it
necessary to regulate the access to and exercise of power; and
wherever power is organised and distributed in accordance with
social rules - in the spheres of production and reproduction, in
the family and the economy, as well as the polity - these rules
stand in need of legitimation. Legitimacy is an important aspect
of power in all these spheres. How we define what is 'political'
is a matter of convenience and focus of interest. In so far as the
power of gender and property are the subject of legitimation,
involving reference to law and convention, the public justification
40 The Legitimation of Power

of rules and the organisation of consent, these forms of power

are eminently 'political' in the broadest sense of the word.
It is undeniable, however, that legitimacy also has a special
significance for the more narrowly political sphere, the sphere of
the polity, for a number of reasons. The political domain is
responsible for the legal formulation, adjudication and enforce-
ment of society's rules, and thereby legitimates all other social
powers. At the same time there is no law-making authority
beyond itself to legitimate its own rules of power. Moreover, the
form of power which is distinctive to it - organised physical
coercion - is one that both supremely stands in need of legitim-
ation, yet is also uniquely able to breach all legitimacy. The
legitimation of the state's power is thus both specially urgent and
fateful in its consequences.
The fact that legitimacy is significant for power relations in
general, while also having a particular urgency for the state, has
determined a two-part structure to this book. The first part will
consider legitimacy across the three dimensions of gender, class
and political power together, while also identifying significant
differences between them. Chapter 2 will analyse what exactly
power is, and the respects in which it requires legitimation.
Chapter 3 will explore further the basic normative structure of
legitimacy, as already outlined. Chapter 4 will draw out the
implications of the fact that legitimacy in historical societies is
typically constructed and reproduced within established power
relations, rather than outside or beyond them. The second part
of the book will examine legitimacy in the contemporary state;
it will consider the internal dynamics and crisis tendencies of
different political systems, explore the different modes of non-
legitimate power, and explain why the contemporary state has
such difficulty in attaining the legitimacy it needs.
Underlying the substance of my analysis will be a deeper
methodological purpose: to arrive at a valid understanding of
the relation between social science and normative philosophy.
Although I have argued that the point of each activity is different
and should not be confused, by Chapter 4 it will have become
clear that an adequate explanatory account of legitimacy can
only be obtained by adopting the independent standpoint of the
normative philosopher. And by the end of the book the necessity
of unifying the two activities in a common enterprise will have
Social Science and Legitimacy 41

been reached, as the culminating point of the enquiry. What I

shall show is that a complete understandng of legitimacy can
only be attained through an adequate specification of the relation
in which social science and normative philosophy stand to each
other. For that reason the subject of legitimacy can claim to
constitute, not merely an important topic, but the central issue
in social and political theory.
2 Power and its Need of
Legitimation
In the first chapter I wrote about 'power' as if it were self-evident
what it meant. But what is power, and why does it stand in need
of legitimation? To answer these questions it will be necessary
to make a clear separation between the concept of power and
that of legitimacy. In practice such a separation will be artificial,
since the interplay between power, rules and legitimating norms
and actions typically constitutes a complex interrelationship, in
which each element is affected by the others. Yet it is only by
distinguishing them conceptually that we can come to understand
the connection between them. Where the first chapter
approached the subject of power through an analysis of legit-
imacy, the present one will provide an approach to legitimacy
through the analysis of power. In doing so I shall aim to identify
precisely what it is about power that calls for legitimation, which
in turn explains the underlying structure of legitimacy that I
outlined in the previous chapter.
Two caveats are in order before I proceed. First, although I
do not believe that the concept of power is an essentially or
necessarily contestable concept, as some have argued (e.g.
Lukes, 1974), there is no doubt that it has in fact been the
subject of considerable disagreement and contestation. In what
follows I shall simply set out what seems to me the most convinc-
ing way of analysing it, but it would require too much space to
defend my analysis against alternatives at every point.
Secondly, the account of power I give, like that of legitimacy
in the first chapter, is not intended to be historically limited, in
the sense of applying to only one society or historical period. It
is therefore couched at a level of generality which may require
qualification, and certainly elaboration, in particular contexts,

42
Power and its Need of Legitimation 43

but which is nevertheless appropriate to my purpose. The first

part of the chapter will seek to clarify what power is, and what
having power over someone entails. The second part will move
from the conceptual to the more sociological mode of enquiry,
by analysing the way power is socially organised into systematic
relations of dominance and subordination. The third part will
identify precisely what it is about power that requires legitimacy,
and why legitimations take the form they do.

The concept of power

In its widest sense the power a person has indicates their ability
to produce intended effects upon the world around them, to
realise their purposes within it, whatever these purposes happen
to be (Morriss, 1987). Power in this general sense depends upon
certain preconditions: the presence of personal capacities or
'powers' such as health, strength, knowledge and skill; the pos-
session of material resources; and space or scope, in the sense
of freedom from control, obstruction or subservience to the pur-
poses of others. Power and freedom are closely related, but
not identical, concepts. Without freedom, even the strongest
individual may be rendered powerless (Samson in chains); but
without resources of a personal or material kind even the most
free person will remain impotent (the physically incapacitated in
an open space, the penniless in a free market). Freedom is
necessary if we are to utilise our powers to achieve our purposes;
but without such powers in the first place, freedom will be worth-
less to us.
Power, in this most general sense of the ability to achieve our
purposes, is unequally distributed; some people have greater
power than others. We can thus speak of its relative distribution
within a society. But there is a familiar and more specific sense
of power which is also relational: the ability to influence or
control the actions of others, to get them to do what we want
them to, and what they would otherwise not have done (Wrong,
1979). This sense is linked to the first partly as means to end:
one of the ways in which we are able to achieve our purposes
is through influencing or controlling the actions of others. At the
same time one of the typical means of attaining such power is
44 The Legitimation of Power

through the possession of superior capacities or resources,

whether of strength, knowledge, material goods, or a combi-
nation of these. I can come to influence or control your actions
through greater strength (physical coercion or the threat of it),
through superior knowledge (persuasion or manipulation), or
through the promise to grant or the threat to withhold some
resource or service you desire or need (inducement or threatened
deprivation). Relative differences in capacity or resources
between people thus also tend to become relational, the means
to influence or control them.
Not all such relations involve an exercise of power by one
person over another. If I get you to change your behaviour by
means of persuasion, this does not affect your freedom, since
persuasion requires agreement. Nor does it affect your freedom
if I offer an inducement which you can readily refuse. In such
cases I cannot be said to have power over you. But if I can
threaten you with physical coercion if you do not comply, or
with the deprivation of some resource or service that is necessary
to you, or if I take advantage of your ignorance of psychological
vulnerability (manipulation rather than persuasion), then my
power is exercised at the expense of your freedom. I can be said
to have power over you.
Some philosophers would argue that only the actual application
of physical force inhibits a person's freedom; short of that he or
she is still free to act. Even if I point a loaded gun at your head,
they say, you can still refuse to sign the confession (Steiner,
1974; Parent, 1974). But this is a very heroic conception of
freedom, which will not serve for everyday purposes. Most legal
systems treat a confession obtained under such circumstances as
invalid, because it is extracted under duress. However, it is
precisely the actions of heroes, who place honour or ideals above
life, that can subvert a system of power through their example,
since they challenge the normal scale of values on which power
is based. It is because most of us for most of the time value life
above all else, that threats to deprive us of it limit our freedom,
and constitute an effective basis of power over us.
The above example suggests that the restriction of freedom is
not an aU-or-nothing affair, but is greater or less according to
the seriousness of the deprivation that can be inflicted. To speak
of 'having power over' someone usually implies a continuous
Power and its Need of Legitimation 45

relationship, in which a substantial sanction is always present.

One reason why sanctions are much more effective than induce-
ments as the basis for a continuous power relationship is that
they do not have to be continually used, or even explicitly
threatened, to be effective in modifying behaviour. It is sufficient
that people 'know their place'. Moreover, the possibility that
they may anticipate the wishes of a superior without having to
be explicitly instructed (the so-called 'law of anticipated reac-
tions') may make the nature of the relationship difficult to read at
first glance (Friedrich, 1963, pp.199-215; Nagel, 1975, pp.15-19).
This is particularly true of relations of dependency. These
derive from a situation of relative powerlessness which leads the
weak to seek the protection of the strong, the propertyless to
seek service with the propertied, and so on. The awareness of
their impotence outside the relationship can itself be sufficient
to keep the dependent party submissive to the wishes of the
superior, without any threats needing to be made. To possess
superior physical power or resources is not only to be able to
compel others; it is also to be able to offer protection against
physical coercion or destitution, and hence to establish relations
of dependency. Here an initial inducement can become the basis
of a continuous power relationship, through the vulnerability of
the dependent party to the withdrawal of the essential resource
on which they rely.
To say that having power over someone involves a restriction
of their freedom, a subordination of their purposes to one's own,
is not necessarily to say that these purposes are self-seeking, or
are always pursued at the expense of the subordinate. Otherwise
we could have no concept of paternalism. The restriction of the
freedom of children by their parents is often in their interests;
it enables their parents to realise their purposes for them. How-
ever, the subordination of children is a temporary condition, and
one whose end is the expansion of their own powers, of their
capacity to realise their purposes in the future. And it is precisely
because of the value that we place upon realising our purposes
that in general we regard the restriction involved in power
relations, and the subordination of one person's will to another,
in a negative light, and something that therefore requires justifi-
cation. It is just here, as I shall show later, that legitimacy, or
at least one key dimension of it, has its point.
46 The Legitimation of Power

The social organisation of power

So far, to assist conceptual clarification, I have treated power as

if it were simply an attribute of individuals. Yet the relative
differences of power between people are largely the result of
social arrangements. And the relations of power between them
are typically the product of collective organisation. Sociological
theories are often distinguished according to whether they
emphasise one or other of two different aspects of the collective
organisations of power, which are extrapolated into two different
definitions of power itself. One is the power of the dominant
over the subordinate, and its collective organisation to achieve
the purposes of the powerful. The other is the power of the
society or collectivity as a whole - over the forces of nature and
the environment, and in relation to other societies - and its
collective organisation to achieve common purposes (Parsons,
1960, pp.219-25; Wrong, 1979, pp.237--47).
To emphasise one of these aspects to the exclusion of the
other is erroneous, since they are interrelated. The manner in
which a society's internal power relations are organised has obvi-
ous consequences for the power of the society as a whole. And
the limitations of freedom and subordination of purposes
involved in power over others constitute an important means of
social coordination. Power is not the only means of coordinating
social activity; voluntary agreement, market arrangements and
socialisation into common values are all familiar ways in which
people's purposes are aligned with one another to make collec-
tive purposes attainable. But power relations constitute a crucial
element in this process, both in themselves and as an under-
pinning to the other means of coordination.
However, it is important not to confuse the analysis of power
with the question of its legitimacy. One of the most common
justifications for the power of any dominant group over a subordi-
nate one is that it enables the collective purposes or general
interests of the society as a whole to be realised. Such justifi-
cations are rarely all they seem. A characteristic attribute of the
powerful is their ability to define what the goals of the collective
as a whole should be, i.e. to achieve their purposes for the
colfective in preference to others'. And a recurrent feature of
the social organisation of power is that general interests can only
Power and its Need of Legitimation 47

be met in ways which also serve the particular interests of the

powerful, since the means of realising them are under their
control. To treat any collective as an undifferentiated whole, as
a single entity with definable purposes and interests, is to over-
look the way these purposes and interests are both constructed
by and mediated through its internal relations of power.
It is with these internal relations of power within societies
that this book is primarily concerned. It is not concerned with
differences of power between societies, except in so far as these
come to affect the legitimacy of power relations within them (see
Part II). It is not even concerned with differences or inequalities
of power between groups or individuals within societies, except
in so far as these become the means to exercise power over
others. The subject is large enough without extending it to cover
all forms of inequality. More importantly, it is precisely relations
of power over others that raise the most central question of
legitimacy for normative philosophy and social science alike: that
of cooperation with, or obedience to, the powerful.
In order to indicate that it is power relationships that I am
concerned with, I shall use the terms 'dominance' and 'subordi-
nation' to specify those systematic and continuous relations of
power that occur between groups of people, or between indi-
viduals in so far as they represent social categories. Two ques-
tions are crucial to understanding the social organisation of these
relations. First, what are the typical means of power whereby
one social group is able to dominate another? Secondly, what
are the basic groups or categories between which relations of
dominance and subordination recur in most historical societies,
and what explains this recurrence?
To begin with the first question, it is useful to distinguish a
number of different means of power, which can provide the basis
for relations of dominance and subordination. Although they are
often found together, mutually reinforcing one another, it is
helpful to separate them analytically for the sake of clarity. First
is the possession of material resources, of which the means of
production and subsistence on one side, and the means of physi-
cal force on the other, are the most significant. A common
feature of such resources is that they only become a means of
social power in so far as people can be systematically excluded
from access to them. If we all had access to the means of pro-
48 The Legitimation of Power

duction and subsistence (land, tools, the fruits of the earth) or

to the instruments of physical force (weapons, armour, etc.),
they could not constitute a basis for power relations. Central to
the social organisation of power, therefore, are processes of
exclusion, typically embodied in rules, which prohibit general
access to key resources, and which determine who may acquire
the use or possession of them, and by what means. These are
usually called rules of property; and it is through the possession
of, or the privileged access to, property that some people acquire
and maintain power over others who lack or are denied such
access (Parkin, 1974; Cohen, 1981).
As forms of property, the means of production or subsistence
and the means of physical force obviously differ in the way they
operate as agencies of power. The means of physical force give
those who possess them the ability to obstruct, constrain, hurt,
harm or destroy those who lack them, and to compel them to
action or inaction through the threat of such inflictions. The
possession of the means of production or subsistence, on the
other hand, works through the need of the excluded to have
access to the use of these resources, and through an exchange
relationship whereby their services (usually labour services) are
required in return for such use. The fact that power in this
context is exercised through an exchange relationship both com-
plicates and obscures its character, since exchange relations can
also occur between equals. By comparison, physical coercion
seems stark and uncomplicated, and for that reason is often used
as the paradigm example of power. Yet the means of physical
coercion are also a means of physical protection, and as such
can provide the basis for a power relation through exchange,
because of the need of the defenceless for protection against
third parties, and their consequent dependency upon those who
can provide them with it. In whatever form it occurs, however,
what is common to both the means of physical force and the
means of production and subsistence is that in each case relations
of dominance and subordination are constituted in the first place
through the possession of, and exclusion from, a key material
resource, and by the rules that determine these respective con-
ditions (property rules).
A second basis or means of power lies in the control of socially
necessary activities, and the possession of the skills associated
Power and its Need of Legitimation 49

with their performance. Here too, such activities and skills only
become a source of power through rules of exclusion and access,
which limit the number of those who may acquire the skills and
take part in the activities, and define who is entitled to do so.
These rules are usually referred to as the division of labour. At
first sight the division of labour looks like the basis for a mutual
dependence and reciprocal benefit through the specialisation of
social tasks and functions. However, in most societies the division
of labour is hierarchically ordered, in that some activities and
skills are defined as being more socially important and valuable
than others, and the majority are prevented from engaging in
them either by formal rules of exclusion, or because they are
unable in practice to acquire the skills necessary to their perform-
ance. As with the ownership of key resources, the degree of
power is determined by the extent of the exclusion, and by the
importance of the activity to the excluded. It is further enhanced
by the association of a given skill or practice with an esoteric
body of knowledge, couched in a language removed from every-
day discourse and comprehensible only to the initiated.
A third means of power consists in the occupancy of positions
which carry with them the power of command over others. Such
positions, in which the power of command is constituted, defined
and circumscribed by rules, are called positions of authority.
Authority is often said to be 'legitimate power', but this is an
inadequate formulation because of the ambiguity inherent in the
concept of legitimacy. All the means of social power mentioned
above possess a primary legitimacy in so far as they are deter-
mined by rules. What distinguishes authority from the power
over others that derives from the possession of property or the
control of some necessary social activity or skill is that, in the
case of authority, it is the relationship between the dominant and
subordinate itself that is specified by the rules. At this point the
distinction between the different means of power becomes at
best an analytical one, since most relations of dominance and
subordination embody an element of authority, i.e. the rule-
governed specification of the relationship. In this sense authority
constitutes both an aspect of power relations and a means of
power in its own right (power deriving from positions of com-
mand).
The same point could be expressed differently by saying that
50 The Legitimation of Power

most historical examples of authority have in practice been

associated with a hierarchical division of labour or the possession
of key resources, whether because an initial power relation based
upon the control of a resource such as land led to a formalisation
of the relationship with the landless in rule-governed terms
(authority); or because holding a position of authority enabled
its occupants to acquire such resources; or because their pos-
session was necessary to the enforcement of its commands (as is
true particularly of the means of physical coercion). In whichever
manner the connection may historically have occurred, authority
has typically comprised a part of most systematic relations of
dominance and subordination, even though an analysis of these
relations cannot by any means be exhausted by an account of
their formal authority rules.
Before leaving the discussion of authority, it is worth noting
its importance as a means of coordinating and organising the
power of the dominant in face of the subordinate. Military
power, for example, is as much dependent upon the coordinating
role of authority as it is upon the possession of the physical
means of coercion. It is rare for subordinate groups to lack any
means of power of their own, even if this only takes the defensive
form of a power of limitation or veto in face of the power of
initiative exercised by the dominant. The subordinate are often
more numerous, and their own contribution within the social
division of labour makes them indispensable, as a group, to the
dominant. However, even a defensive power requires coordi-
nation for it to be effective, and it is precisely one of the purposes
of the organisation of power on the part of the dominant, e.g.
through their authority over an administrative staff, to ensure
the disorganisation among, as well as their control over, the
subordinate.
The main bases or means of power, then, underlying relations
of dominance and subordination, involve exclusion from and
access to necessary material resources, activities or skills and
positions of command, i.e. property, the division of labour and
authority respectively. A distinctive feature of the account I have
just given is the emphasis I have placed upon rules (whether
conventional or legal in form), and in particular upon rules of
exclusion, as both the source of social power and the means of
maintaining it. Here we should distinguish between two different
Power and its Need of Legitimation 51

types of exclusion rule. Primary rules of exclusion are those

which turn key resources, activities and skills into a means of
power, and protect the authority of those who hold positions of
command. These are the basic 'keep out' signs, which exclude the
majority, and determine their relative powerlessness. Secondary
exclusion rules are those which determine who can come to hold
the means of power thus constituted and protected, and which
may exclude whole categories of people from attaining them, as,
for example, women have at various times and places been
excluded from owning property, engaging in socially important
activities or holding positions of authority. Such rules may oper-
ate indirectly as well as directly, in that a person's position
within the division of labour may prevent them in practice from
obtaining access to key resources or positions of authority to
which they are formally eligible. Taken together, I shall call the
rules of exclusion and access discussed above the 'rules of power',
in view of their significance for the way that power is socially
organised.
Now that the main bases of social power have been identified
in the possession of key resources, in the division of labour and
in authority positions, and in the rules of exclusion and access
underpinning them, we need to explain why it is that the division
of societies into dominant and subordinate groups takes histori-
cally recurrent forms. In theory, the different means of power
could be distributed or combined in an almost infinite variety of
ways. In practice, three lines of division are historically recurrent:
those of class, gender and political power. The reason for their
recurrence is that relations of dominance and subordination are
organised around the most fundamental social activities - of
production, reproduction, physical protection and rule determi-
nation respectively. While each of the three dimensions involves
all the different means of power identified above, one of these
has historically had a central place in each.
The dominance and subordination of classes is constituted in
the process of material production. The primary basis of class
power has historically lain in the private ownership of the means
of production, and in the compulsion of the propertyless to
work for the propertied as a condition of obtaining their own
subsistence. Marx's theory draws attention to the different types
of ownership - of the person of the worker, of land, of capital
52 The Legitimation of Power

- which form the basis of different systems of class dominance

(Marx and Engels, 1968, pp.181-5). His theory of surplus appro-
priation also shows how the labour of the direct producers has
freed dominant classes to engage in a whole variety of activities
from which the direct producers have been excluded. The owner-
ship of property can thus provide the basis for a social division
of labour, as well as access to the means of physical coercion,
whether formally through the linkage of property to political
functions, or informally through the purchase of weapons and
the acquisition of the skills necessary to use them (military
aristocracies).
The significance of Marx's analysis of class is that it emphasises
the control over labour and its product as the central feature of
class relations. Other theories of class look at class relations in
a different way. Weberian theory, for instance, defines classes
by the different kinds of resource they possess (e.g. property,
skill, collective organisation) in competing for differential shares
of the social product. Its concern, in other words, is more to
analyse the inequality between classes than the dominance of
one over another (Parkin, 1979, pt.1). From the standpoint of
analysing power relations, however, Marx's approach is prefer-
able (even if one does not accept his labour theory of value),
because it draws attention to the way in which the power of
ownership secures control over labour.
Under contemporary capitalism the power of property is
increasingly exercised through impersonal institutions, rather
than as a direct personal relationship between owner and worker;
and the typical capitalist is as much the company director who
acquires a personal share stake in the business by virtue of
position, as the owner of a family firm or the wealthy rentier
whose claim on the product is only indirect (J. Scott, 1982,
ch.6). In command economies of the Soviet type (now almost a
historical phenomenon), in which private ownership was largely
divorced from power over the means of production, control over
the labour of others was determined wholly by authority position
within state-owned institutions. What these economies exempli-
fied was not so much the end of control over the means of
production as the basis of class dominance and subordination, as
its reconstitution under a different form, i.e. under different rules
Power and its Need of Legitimation 53

of power, with correspondingly different principles of legitimacy

(Djilas, 1958).
Where the dominance and subordination of classes is consti-
tuted in the process of material production and by the power
deriving from control over key productive resources, that of
gender is organised around the activities of reproduction, and
determined by the social division of labour. Although only
women are capable of giving birth and suckling children, the
sexual division of labour has historically gone well beyond these
physiological differences, in two respects. First, the conventional
assignment to women of key domestic tasks such as child-rearing,
care of the home, servicing the personal needs of men and
women, has in practice discouraged, if not always prevented,
them from engaging in activities accessible to males who do not
have these responsibilities. Secondly, these disabilities have been
reinforced by formal exclusions, debarring women from certain
activities and positions, and creating a sharp separation between
'public' and 'private' spheres. It is these exclusions established
by the division of labour that have been primarily responsible
for women's physical and economic dependency upon men, albeit
varying in degree according to historical circumstance (Rosaldo
and Lamphere, 1974; Friedl, 1975).
The above account, which derives the dominance and subordi-
nation of gender from the division of labour, is at odds with
accounts which explain women's subordination in terms of male
physical power or sexual aggression. It is true that physical force
has had a part to play in gender relations, whether as direct
coercion or as the protection of women from the coercion of
other men. Yet anthropological evidence suggests that male
aggression against women is not universal. And it tends to be
associated with a division of labour that removes men from any
involvement in child-rearing or caring roles, and assigns them
exclusive control of activities involving the means of violence,
such as hunting dangerous animals or fighting against other
societies. It is this joint division of labour which creates a particu-
lar type of male physical power on the one side, and women's
vulnerability to it on the other (Harris, 1977; Sanday, 1981).
If the sexual division of labour is responsible for women's
vulnerability to male coercion and their consequent physical
dependency, the same division of labour also produces women's
54 The Legitimation of Power

dependency on men for the means of subsistence. The extent of

this dependency has varied historically according to the role that
women have played in production, and this in turn has depended
on whether production has been organised within the domestic
economy or outside it. The same process that 'freed' the worker
from the soil in modern Europe, and created wage labour, also
'freed' married women from any substantial role in production
and left them economically dependent upon men. Although the
legal rules excluding women as a whole from many spheres of
paid employment have now largely been abolished in Western
societies, the persistence of the sexual division of labour in the
family, combined with a dual labour market in the economy,
tends to restrict married women to lower-paid, part-time, less
secure types of employment. What we see is the power of con-
ventional rules to sustain informal processes of exclusion and
inequality, and so to perpetuate women's economic dependency
within the home and their subordination to men outside it (Bar-
rett, 1980; Whitelegg et a!., 1982).
A third dimension of dominance and subordination, the politi-
cal, is organised around two interconnected social activities, and
two complementary means of power. The first activity is that of
physical defence, and the power associated with it is that which
derives from control over the means of physical force. The
second activity is that involved in deciding and giving legal
expression to the rules of the society, which provides a crucial
means of power for those who control it, since it gives them the
ability to determine the 'rules of power' for the society as a
whole. If, as I have argued, the social organisation of dominance
and subordination in its different dimensions is effected through
rules of exclusion and access, then the ability to shape or deter-
mine such rules is a particularly important power. What links
this distinctively 'political' activity of rule-determination with that
of physical defence is the key part that the organisation and
control of the means of force plays in both. Organised physical
coercion provides the most effective means of enforcing society's
legal rules, as well as protecting it against attack. As a conse-
quence the two activities have usually been connected, and the
two means of power combined in the same group or institution
to form a potent basis of dominance and subordination.
In conclusion, it is the organisation of power relations around
Power and its Need of Legitimation 55

the most fundamental social activities that makes the dimensions

of class, gender and political dominance and subordination recur-
rent features of societies throughout history. There are of course
other recurring bases of inequality, of which race, ethnicity and
religion are the most persistent historically. However, these only
become structured as forms of dominance and subordination
when associated either with political power, through conquest,
or with the ownership of property, or with both, as in slave or
caste societies. Of course the political or economic dimension
does not exhaust the significance of such relations; but the ideas
of superiority and inferiority associated with them belong as
much to the sphere of legitimacy as to power itself, and will be
considered at the appropriate point in the final section of the
chapter.
In this section I have sought to identify the main dimensions
of dominance and subordination, to distinguish the typical means
of power associated with each, and to explain their historical
recurrence. The fact that they are recurrent does not of course
mean that the forms they take may not also vary considerably
from one society to another, e.g. according to the environment,
technology and historical legacy available to each. Variable also
are the ways in which the different dimensions of class, gender
and political dominance interrelate with one another, and in
particular whether they are mutually reinforcing or cross-cutting.
Marxist theories, as is well known, assign primacy among them
to the dimension of class. If this priority is based upon the
assumption that the activity of production is socially more impor-
tant than the activities of reproduction, defence or rule-determi-
nation, or that the power deriving from control over the means
of production is inherently superior to other means of power,
then such a conclusion is simply untenable. More plausible is a
different argument to the effect that the history of the expansion
of human powers can most convincingly be written in terms of
the expansion of productive forces, though even such a history
is only explicable as a response to predicaments arising from
military competition between societies, as well as to problems of
material production within them. In any case such an argument
could provide no grounds for assigning priority to the dominance
of class within any particular society, the pattern of each of
56 The Legitimation of Power

whose power relations has to be analysed separately, and on its

own terms.
If the dimensions of dominance and subordination vary in form
between societies, they are not immutable within them either.
As I have already indicated, the subordinate are rarely ever
powerless in face of the dominant and relations between them
contain elements of tension and struggle: between classes over
the rights of property and the extent of the surplus product;
between men and women over the sexual division of labour;
between the holders of political power and their subordinates
over the power to influence rule-determination and the use of
the means of coercion. Such struggles rarely challenge the
relationship itself, as opposed to the terms on which it is con-
ducted; and the outcome is necessarily a matter of context and
circumstance, to be investigated in each particular case. My con-
cern here has simply been to identify the main bases on which
such relations of dominance and subordination have historically
been constructed.

Power and legitimacy

In the previous section I argued that relations of dominance and

subordination are determined by rules of exclusion and access,
which make key resources, activities and positions into a means
of power, and regulate who may gain access to them and upon
what conditions. Yet in the first chapter I also argued that a
primary ground for the legitimacy of power was that it was
obtained and exercised according to established rules. Why does
legality confer a prima facie legitimacy, and does it follow there-
fore that all power is by definition legitimate?
The basic point about social rules, whether conventional or
legal, is that they have a normative force, not only in prohibiting
us from certain actions, but in defining our duties and obligations
towards others, and in conferring rights and entitlements that we
can require others to respect in turn. In as much as power is
itself constituted by rules of exclusion, therefore, it is also legit-
imated by them, since they confer the right on the powerful to
require others to respect the exclusiveness which is the basis of
their power. In other words, the very rules that constitute
Power and its Need of Legitimation 57

resources, activities and positions as a means of power in the

first place also serve to legitimate them as forms of power.
Is this process circular? Does it follow that all power is necess-
arily legitimate? This is not so, for two reasons. One is that the
occupancy of property, the development of a hierarchical division
of labour, or the establishment of a command structure can
occur, and historically frequently has occurred, through acts of
forcible appropriation, exclusion or subjection which take place
in violation of existing rules or outside of them. Such usurpations
cannot be legitimate. However, the resulting power relations
typically become consolidated and perpetuated through the
establishment of rules which underpin and give legal form to the
original usurpation. From that time on, subsequent positions of
power come to be derived from the rules, in the same way as
those whose rules originate through custom or agreement.
Although the memory of the original usurpation may be kept
alive, with the passage of time the issue of the rules' legitimacy
becomes less a question of their origin than one of the ongoing
character of the relationships they embody, and the nature of
the requirements they impose.
Here lies the second reason why power is not necessarily legit-
imate, even when it is legally valid, and that is, as I argued in
the first chapter, that legality constitutes only one dimension of
legitimacy; it is a necessary but not a sufficient condition for it.
The rules of power themselves require justification, however they
may have originated. There is an important distinction to be
observed, in other words, between the legitimacy of an individual
power-holder, which is a matter of validity according to the rules,
and the legitimacy of the power system as such, with its particular
relations of dominance and subordination, its characteristic mode
of access to power, and so on. Where the legitimacy of the
individual derives from the rules, the rules themselves, and the
power arrangements they define, can only be justified by moral
considerations that go beyond them. Why is such justification
necessary at all? The simple answer is that power relations
involve negative features - of exclusion, restriction, compulsion,
etc. - which stand in need of justification if the powerful are to
enjoy moral authority as opposed to merely de facto power, or
validity under a given system of law.
An examination of these negative features present in any
58 The Legitimation of Power

system of dominance and subordination will reveal why the justi-

fications for power relations take the historically recurrent form
that they do. Such relations involve, first, a marked difference
of fates between dominant and subordinate, in the character and
quality of their lives, in their respective powers, to use the lan-
guage of the first section; i.e. their capacity to achieve their
purposes, and the kinds of purposes they can achieve. They
are distanced from one another by various forms of inequality.
Secondly, these differences or inequalities are systematically con-
nected, in that the expanded powers of the dominant are depen-
dent upon the limitation of powers of the subordinate, and are
achieved primarily at their expense, for instance because they
are able to appropriate the product of the others' labour, or
because the labour of others frees them to engage in activities
that expand their own powers. The respective inequalities of
condition are not a disconnected, but a causally related, phenom-
enon. Thirdly, the relationship involves the subordinate in a
restriction of their freedom, through their direct subordination
to the purposes of the powerful in a relation of command and
obedience, or of submission to the wishes of another. Each of
these three components of dominance and subordination carries
a prima facie negative connotation, as evidenced by the terms in
which they can be described. Thus the difference of fates of
life chances can be termed inequality. The transfer of powers
(resources or opportunities) from subordinate to dominant can be
called exploitation or parasitism. The loss of freedom is termed
subjection, subservience, or compulsion.
Now it is possible to avoid such negative connotations in the
first place by simply removing the rules of pvwer from the sphere
of human agency and treating them as part of the natural world,
whose allocations, like the weather, are morally blind and beyond
the possibility of human control. The idea of fate, chance, for-
tune, is a commonplace one. However, these concepts are much
more readily applicable to the unpredictabilities of life, the
sudden shifts in fortune, than to the regularities of social relations
and their rules of power. Here, appeals to the idea of nature
take the form much more of nature as 'natural order' or 'natural
law'. While it may be conceived as an order that is unchanging
and beyond human control, nevertheless it is one that is also
Power and its Need of Legitimation 59

seen to make moral sense, either because it is intrinsically pur-

posive, or because it is the creation of a purposive divine will.
Whether the social order and its power relations are conceived
as having a natural, divine or human origin, then, there is a
common expectation that they should satisfy certain moral
requirements, and that the negative features of dominance and
subordination should be amenable to moral justification.
Although the content of such justifications is historically variable,
the recurrent form they take is logically related to the negative
features distinguished above.
Thus, first, the inequality of circumstance between dominant
and subordinate is justified by a principle of differentiation,
which reveals the dominant as specially qualified, suited or
deserving to possess the resource, pursue the activity or hold the
position which forms the basis of their power, and the subordi-
nate as correspondingly unsuited or unfitted to do so, and hence
rightly excluded from it. The inequality, in other words, is not
arbitrary or fortuitous, but is based upon a normative distinction
of superiority and inferiority, whether determined by birth, or
developed and demonstrated over time. One feature of such
normative distinctions is that they tend to carry over into all
social relations between dominant and subordinate groups,
beyond the site where the relationship is primarily constituted
(e.g. class relations outside the workplace, gender relations out-
side the family); and also that they have a capacity for historical
survival beyond the collapse of the power system which they
served to justify (e.g. racial stereotypes after the end of slavery
and colonialism).
Secondly, the systematic transfer of powers (of resources and
opportunities) from subordinate to dominant is justified when it
can be shown to serve not merely the interests of the powerful,
but those of the subordinate also, or else to make possible the
realisation of larger social purposes in which they have a concern.
Legitimacy requires the demonstration of a common interest
which unites, as well as a principle of differentiation which div-
ides, dominant and subordinate. Where the subordinate are not
conceived as having any interests of their own meriting attention,
independent of their utility to the dominant, as a slave is equated
with a domestic animal, there can be no moral community
between them, and no legitimacy to the relationship.
60 The Legitimation of Power

Thirdly, the limitations of freedom involved in subservience to

the requirements or purposes of the dominant are given moral
acceptability by evidence of consent on the part of those qualified
to give it. The free acceptance of subordination cancels out the
restrictions on freedom involved in the relationship. Who counts
as qualified to give consent varies according to the conventions
of the given society, but, as I have already suggested, it is
typically related to the category of those who count as free or
independent. To be able to give consent a person has to be
treated as free or independent in the first place, or at least as
capable of freedom or independence. In the case of political
power, consent is related to membership of the political com-
munity, which can be a more restricted category than that of the
free.
To each of the negative characteristics of power relations,
therefore, there stands a corresponding form of justification or
rectification, which renders it morally acceptable. The inequality
of powers and life chances is justified when it can be seen to
correspond to the respective qualities of the dominant and subor-
dinate; the difference is not one of power, merely, but of superi-
ority and inferiority in some relevant respect. The transfer of
resources and opportunities from subordinate to dominant is
justified when it can be seen to serve not merely the interests of
the powerful, but those of the subordinate also, or larger social
purposes in which they have a stake. The restrictions on freedom
are rectified when those who qualify as independent agents freely
consent to them. The different dimensions of legitimacy, in other
words, are not themselves arbitrary, but derive from the negative
characteristics intrinsic to power relations as such, which arouse
our concern as moral agents.
To say that, for a power relationship to be legitimate, it has
to be justified in the ways outlined, is not to say that convincing
justifications will always be provided, or that they will find accept-
ance among the subordinate. However, it is a notable feature of
power relations that they are themselves capable of generating
the evidence needed for their own legitimation. Thus the evi-
dence of superiority and inferiority which justifies the inequality
of condition between dominant and subordinate is itself largely
the product of that condition. Those who are excluded from key
positions, activities or resources are thereby denied the oppor-
Power and its Need of Legitimation 61

tunity to acquire or demonstrate the capacities and characteristics

appropriate to their occupation or exercise, so justifying their
subordinate position. This is true even where relatively open
processes of selection are at work, once the selection is perfor-
med by an education system which is given the task of preparing
children differentially for their respective future roles. Evidence
about the fitness or appropriateness of people to exercise power
thus tends to be structured by the relations of power themselves,
and therefore to have a self-fulfilling quality about it.
The same holds true for demonstrations of the general interest.
Once some necessary social resource or activity comes to be
controlled by a particular group, it follows that the interests of
society at large can only be met through satisfying the interests
of that group, and on terms acceptable to them. Those who have
historically controlled the means of production or subsistence, of
violence or administration, have been in a position to ensure
that general needs for welfare, employment or security could
only be met through the power relations that simultaneously
secured their own privileges. The link between these relations
and the performance of necessary social tasks, already described,
ensures this legitimating connection. Marxist accounts which
insist that power serves the particular interests of the powerful
alone in a zero-sum manner, and that all claims to satisfy general
needs are spurious, overlook the way that social relations are so
arranged that general interests can in practice only be secured
through satisfying the particular interests of the powerful. Func-
tionalist accounts, on the other hand, which hold that any given
distribution of power has evolved as the one most suited to
carrying out necessary social functions, overlook the role that
power itself plays in the construction of social relations, and the
alternative patterns of distribution that could be equally 'func-
tional' to the satisfaction of given social needs. That society at
large should need the capacities and resources of the powerful
is plain evidence of their public utility; that the utility might be
realised without their power requires a perspective which goes
beyond, and stands outside, the existing relations of power.
If we look, finally, at consent, then we find that it is precisely
the lack of some key resource or skill deriving from the rules of
power themselves that leads the subordinate to voluntary accept-
ance of their dependency upon the powerful. Arrangements
62 The Legitimation of Power

made under the direct threat of physical coercion are usually

held not to be morally binding, because they are extracted under
duress. So too are those made from a condition of need or
insufficiency, which is the intended outcome of actions by those
who will benefit from any consequent agreement (e.g. driving
people off their land so that they will be forced into dependency).
But those conditions of need or insufficiency which are the sys-
tematic product of established rules of power have all the appear-
ance of 'naturalness', since they are not the result of intentional
human agency; and agreements arising from them are commonly
held to be morally binding. The consent of the subordinate thus
both derives from the rules which have determined their initial
powerlessness, and also serves to reinforce them.
The capacity of power structures to generate the evidence
necessary to their own justification, and to reproduce the con-
ditions of dependency from which consent to subordination is
freely given, helps to explain how it is that their legitimacy can
come to be widely acknowledged by those involved in them, the
subordinate included. 'Dominant ideology' theories tend to put
far too much emphasis upon the determining influence exercised
by the powerful over the ideas of the subordinate, through their
preferential access to the means of ideological construction and
dissemination (Abercrombie and Turner, 1978). The account
offered here suggests a different kind of explanation: that both
the evidence and the interests of the subordinate are so struc-
tured that the justifications advanced for the rules of power
prove plausible to them within the given social context. Their
plausibility can only be challenged from a position or standpoint
outside that context, e.g. by comparison with alternative rules
of power, or when social changes have come to undermine from
within the evidence on which they are based. It is a characteristic
of the modern period that such comparisons are more readily
available, and substantial social change is more continuous, than
in the pre-modern world. It follows that the legitimacy of any
society's power rules is now more open to question, and its
legitimating principles and procedures have to be more capable
of withstanding comparison and challenge, if they are to provide
a credible support for the rules of power.
In conclusion, I have attempted in this chapter to define what
power is, to analyse what is involved in power relations, and to
Power and its Need of Legitimation 63

identify the negative features which require justification, and

which in turn provide the normative structure of legitimacy. I
have argued that the basis for systematic relations of dominance
and subordination lies in what I have called the 'rules of power',
the rules governing exclusion from and access to key resources,
activities and positions of command. Such rules also provide the
first dimension of legitimacy, though they may come to be
infringed, and in any case stand in need of justification them-
selves. What is required for their justification can be demon-
strated from an analysis of the negative aspects of power: the
inequality of fates or life chances has to be shown to be appropri-
ate to the differential characteristics of dominant and subordi-
nate; the expanded powers of the dominant have to be seen to
serve a general and not merely a particular interest; the limi-
tations of freedom of the subordinate have to be made good by
actions expressive of consent. As I have also suggested, a
society's rules of power are typically capable of generating the
evidence, and of structuring the interests of the subordinate, in
a manner conducive to their own legitimation. The task of the
next chapter will be to explore the underlying normative struc-
ture of legitimacy more fully, and the different forms it can take
in different societies and periods; and to elucidate further what
I have called the self-legitimating character of power rules.
3 The Normative
Structure of
Legitimacy
This chapter has a number of purposes which it will be as well to
distinguish at the outset. Most basic is to elaborate the threefold
structure of legitimacy as rule-derived validity, the justifiability
of power rules, and expressed consent; and to consider some
problems relating to them that I have not so far addressed.
Within this structure, secondly, I propose to explore some of the
main differences between legitimating principles, beliefs and
forms of consent that characterise different historical epochs and
types of legitimate power relations. If my treatment here appears
overly schematic, it is because this is the only way to achieve a
manageable grasp of such a diverse range of material. My final
purpose will be to exemplify at appropriate points what I have
called the self-fulfilling character of legitimate power: the way in
which systems of power themselves structure many of the beliefs,
interests and conditions of consent that provide for their legitim-
ation; and to identify the kinds of circumstance in which this
self-reinforcing cycle tends to break down. Since this last theme
forms the subject of Chapter 4, my treatment of it here will not
be systematic.

Legitimacy derived from rules

That any form of power should be acquired and exercised in

accordance with established rules, whether the power be derived
from property, position or social function, is the first condition
of its legitimacy. 'I obtained it by the rules, and am therefore

64
The Normative Structure of Legitimacy 65

entitled to it,' is a standard claim in all societies. Rules form a

basic component of social life. It is only through their presence
that we are able to predict the behaviour of others, and introduce
any settled expectations into our lives. Without rules we could
make no plans for the future, or entertain any projects beyond
the merest hand-to-mouth survival. Such settled expectations
have a very different basis from the predictabilities of the natural
order, with its circulation of the seasons, and succession of night
and day. Social rules ensure predictability through their norma-
tive or prescriptive force; they impose obligations and create
corresponding entitlements, which are publicly acknowledged
and collectively enforced. As such they both serve to regulate
behaviour in a predictable fashion, and provide the reference
point for entitlement claims which people can expect to have
recognised by others.
It follows from the nature of social rules that in any rule-
governed social order the existence and acquisition of power
cannot be separated from the normative expectations and entitle-
ments by reference to which its possession is also justified. The
rules of exclusion which make necessary resources into property,
and therefore a means of power, are rules which lay obligations
on the excluded to keep off or keep out; the social recognition
of the obligation is essential to the power of those who possess
such resources. So too, the rules which determine who shall
come to acquire the power of property, position or function,
and by what means, confer the right to its exercise and the
corresponding duty to acknowledge and respect it on the part of
others. In a rule-governed social order we cannot separate p9wer
from legitimacy, since both occur simultaneously; in acquiring
power according to the rules a person also acquires the right to
exercise it.
Social rules may be customary and conventional in form, or
be part of a legal order. The drawback to custom and convention,
particularly with regard to important and potentially contested
issues, is that they carry with them no means of adjudicating
disputes about their precise reference and scope, and they rely
on diffuse means of enforcement applied by society as a whole.
The distinctive features of a legal order, as H.L.A. Hart has
shown, are the existence of second-order rules prescribing the
means whereby primary rules are to be recognised and adjudi-
66 The Legitimation of Power

cated, and the presence of a specialised agency to enforce them

(Hart, 1961, pp.77-96). In most societies the basic rules deter-
mining access to the means of power come to be defined in
legal form, even where they may have originated in convention,
because of the disputes that arise among the powerful themselves
over women, property and position. Among the characteristic
features of the modern world are the extent to which its power
structures are subject to precise legal regulation; and the flexi-
bility of its procedures of legal enactment which enable the disad-
vantage of customary rules in the context of a rapidly changing
environment to be overcome (Weber, 1968, pp.217-19).
However, even in the contemporary world the force of conven-
tion still persists, whether to qualify and subvert a status of
formal legal equality, or to limit the power of the powerful in
areas where the law itself is silent. The ability of conventional
rules sustaining a hierarchical division of labour between the
sexes to survive when no longer legally enforced is not only the
result of the deep-seated assumptions and structured incentives
supporting them. It is also related to distinctive characteristics
of the rules in question which make them eminently sustainable
in conventional form: the lack of contestability about which sex
a person belongs to, and the effectiveness of dispersed means of
enforcement, such as disapproval within the family and the peer
group. On the other side, modern power relations of all kinds
retain substantial scope for convention in the regulation of
relations between dominant and subordinate. The almost
absolute formal power of a UK Prime Minister with a guaranteed
parliamentary majority constitutes a particularly striking example
where the limitations on power are conventional rather than
legally prescribed.
Throughout history, in fact, the limitations on power which
the subordinate have been able to secure, and which they under-
stand as constituting rights for themselves, have usually been
conventional rather than legal in form, in contrast to the rules
securing the power of the dominant. Such rights as those of
peasants to a portion of the crop sufficient for subsistence, of
industrial workers to control aspects of the work process, of
women to a sphere of activity from which men are excluded, are
typically the product of historical struggles between dominant
and subordinate, and represent the crystallisation of a particular
The Normative Structure of Legitimacy 67

balance of forces at a given moment of time. Like rights of

way, they become confirmed by repeated use, and established as
customary rules governing the relations between the respective
parties. The breach of such rules by the dominant constitutes
one of the most frequent sources of grievance on the part of the
subordinate, though one which does not necessarily undermine
the legitimacy of the power structure as a whole (J.C. Scott,
1976, ch.6; Moore, 1978, pp.18-31). Where the powerful are
able to secure legal validation for their action, the grievance
becomes a dispute over which type of rule should have priority:
that of law or custom; statutory rights or the 'rights of free-born
Englishmen' (Thompson, 1975).
The much readier access of the powerful to the law, and the
fact that it provides both the source and protection of their
power, makes appeal to the law as the ground of legitimacy a
particularly favoured strategy for dominant groups. Indeed,
respect for the law is insisted on as the first duty of the subordi-
nate, and legal validity is made to appear not only as the neces-
sary, but also the sufficient, condition of legitimacy: its ultimate,
rather than merely its proximate, source. Moreover, there are
features inherent in most legal systems that serve not only to
encourage respect for the law in general, but to put the particular
content of existing law beyond question, and make it difficult to
challenge. The sanctity that surrounds the law itself comes to
impart its aura to the contingency of the law's content.
What are these features? Most deeply embedded are those
terms used in everyday language which serve to distinguish the
lawful from the unlawful in the achievement and exercise of
power, and which demarcate, for example, theft, violence and
murder from legally permitted forms of acquisition, compulsion
and deprivation of life or livelihood. Such linguistic distinctions
are reinforced by the solemn rituals and imposing figures associ-
ated with the conduct of the law, and by the weighty anathemas
pronounced against those who infringe it. Associated with the
presentation of the law as a morally edifying spectacle or theatre
is the idea of the law as a seamless web, such that to challenge
one element is to threaten the whole. In similar vein, the rules
of power, which could well be other than they happen to be,
are accorded the same moral status as those rules without which
social life would be impossible at all. While societies could not
68 The Legitimation of Power

exist without some property, division of labour, or authority

position, the particular rules governing these could well be differ-
ently arranged. The necessity of the one serves to conceal the
contingency of the other, and gives any challenge to them the
character of a moral transgression.
Such features combine to set the existing rules of power
beyond question, and turn legal validity into the ultimate, rather
than simply the proximate, criterion of legitimacy, to the advan-
tage of those who derive their power from established law.
However, the more that the powerful appeal to the law as the
self-sufficient justification for their power, the more they have to
respect it themselves for their legitimacy to be sustained. The
absolutist idea that those who make the law are themselves above
it is historically exceptional. Much more usual is some version,
however embryonic, of the idea of the 'rule of law': the idea
that the powerful and their agents, whatever influence they may
exercise over the formulation of the law, are themselves subject
to it, and have to conform to recognised procedures if they wish
to change it. What is distinctive about the modern concept of
the rule of law is the development of institutional arrangements,
such as the strict separation of powers between legislative, execu-
tive and judiciary, to give it practical effect. Although such a
separation cannot of itself guarantee to the subordinate any con-
trol over the content of law, it can at least secure them against
the arbitrary exercise of power, and provide the protection of
due process when they fall foul of the powerful. The idea of the
rule of law thus serves as a limitation on power as well as a
guarantor of it, and as a resource available to the subordinate
as well as the dominant, albeit to a lesser degree (Thompson,
1975, pp.258-69).
On its own, however, legality cannot provide a fully adequate
or self-sufficient criterion of legitimacy. Despite all the features
mentioned above which conspire to set the law beyond question,
circumstances will always occur which expose a more fundamen-
tal issue: why these particular laws, and what gives them their
legitimacy? Such occasions arise, for example, when there is a
conflict over the interpretation of existing law, which can only
be resolved by reference to some basic principle. Or social
changes take place which provoke demands for reform of the
law, thereby exposing its contingency. Or there occurs some
The Normative Structure of Legitimacy 69

infringement of the law by the powerful, which they seek to

justify by reference to norms or an authoritative source that lies
beyond existing rules. Such occasions expose a general truth,
that appeal to the law can never provide more than a primary,
and therefore provisional, ground for legitimacy. That such an
appeal is a necessary first step is ensured by the fact that estab-
lished rules provide the recognised source of entitlements, and
because a generalised respect for rules is the condition for any
social order or settled expectations. That it is no more than a
first step follows from the fact that rules cannot justify themselves
simply by being rules, but require justification by reference to
considerations which lie beyond them. These considerations will
be explored in the following section.

Legitimacy as justifiability of rules

The stipulation that power is legitimate to the extent that its

rules are justifiable in terms of shared beliefs raises a number
of questions. Justifiable to whom? What are the criteria for
justifiability? What kinds of belief are relevant? The answer to
the first question is already contained in the specification that the
relevant beliefs be shared between dominant and subordinate.
Without a common framework of belief, the rules from which
the powerful derive their power cannot be justifiable to the subor-
dinate; the powerful can enjoy no moral authority for the exercise
of their power, whatever its legal validity; and their requirements
cannot be normatively binding, though they may be successfully
enforced. If the point of legitimacy is its significance for the
character of power relationships, and its effect on the behaviour
of the subordinate, then it follows that the rules of power must
be justifiable to them.
The answer to the second question, about the criteria for
justifiability, is largely contained in the answer to the third.
Although such criteria are not as tight as those for legal validity,
once we know what the relevant beliefs for the justification of
power rules are, these will set limits to the range of possible
disputes about them, and will identify what sort of disputes they
will be. It is with the third question, therefore, about the kinds of
justificatory belief, that this section will be primarily concerned.
70 The Legitimation of Power

In the previous chapter I distinguished a number of consider-

ations relevant to the justification of power rules - for instance
that they should be based upon a principle of normative differen-
tiation between dominant and subordinate, and that they should
satisfy some general interest or social purpose in which the sub-
ordinate have a concern. However, there is a different, and
arguably prior, way in which rules must be justified, and that is
by identifying an authoritative source from which they stem. It
is the impressiveness of the source from which they derive as
well as the moral persuasiveness of their content that gives social
rules their justifiability. This distinction between source and con-
tent is an important one, and I shall treat these considerations
separately in what follows, although in practice they are comp-
lementary.

i) legitimacy deriving from an authoritative source

What is the ultimate source of law and social rules, from whence
do they derive their authority, what provides the guarantee of
their authenticity or validity - these are questions that concern
the most fundamental of a society's beliefs, its metaphysical
basis, if you like, which cannot itself be questioned (except by
philosophers!), since it provides the terms in which all other
questions are settled. Whether it be tradition, divine command,
scientific doctrine, popular will, or whatever, this constitutes the
ultimate source which validates society's rules and system of law.
It also has a special significance for the legitimacy of political
power, for an important but simple reason. Where all other
power relations in society are validated in the first instance by
the law, the political sphere is itself the in~titutional source and
guardian of that law; and there is no positive law beyond it to
which it can appeal for its own validation. As institutional creator
and guardian of the law, therefore, the political sphere is
especially bound up with the ultimate source of normative
validity that is acknowledged within the society. And since there
is no positive law beyond it, it is uniquely dependent upon that
source for its own legitimacy.
In identifying the ultimate source of authority and validating
basis acknowledged within a society, therefore, we shall find the
legitimating principle for its system of political power as well as
The Normative Structure of Legitimacy 71

for its body of law in general. These sources will differ in differ-
ent types of society, and each will have its respective guardians
and interpreters who enjoy a special prestige because of their
connection with it. At the expense of seeming unduly schematic,
I shall distinguish between sources that are external to the
society, and those that are internal to it. I should emphasise that
these different sources are not necessarily mutually exclusive,
since some can coexist within the same society; however, it is
useful to distinguish them for analytical purposes (see Figure 3.1,
page 72).

a) external sources
For most of history, societies have derived the legitimacy of their
rules of power from a transcendent source: a divine being or
beings, belief in whose wisdom and power makes all rules that
can be shown to emanate from them binding on behaviour. The
divine will, however, requires definitive human interpretation;
and most systems of religious belief are mediated through priests,
prophets, or jurists, whose authority provides an essential under-
pinning to secular forms of power. Religious ideas are well suited
to the legitimation of hierarchical social relations, since the rule
of one God or many over the world they have created provides
a definitive model for earthly authority, whether of rulers over
subjects, masters over servants and employers, or husbands over
wives (the German word for authority, Herrschaft, links all of
these with the rule of a masculine deity). Although more egali-
tarian interpretations of religious doctrine are possible, their
exponents have usually found themselves pitted against the
religious as well as secular authorities, and been forced into
schism. Today the processes of secularisation on the one side,
and the multiplication of faiths on the other have made religious
belief largely incapable of providing a single coherent source of
legitimacy for a society, though there are notable exceptions
within the Islamic world (see below, pp.191-204).
An external source of legitimacy of a secular kind has been
provided at particular periods of history by natural law doctrines,
which derive society's rules of power from universal normative
principles grounded in nature and appropriate to an unchanging
human condition. Although such doctrines can be linked with
religious belief, such linkage is not necessary, and in the Euro-
 --J
Figure 3.1 Legitimacy as justifiability of power rules N

~ divine command
~ (priests/prophets/theologians

external to the society natural law

~ (philosophers)

~ scientific doctrine
(scientists/technocrats)
i) authoritative source of rules
(with respective interpreters)

society in the past: tradition

~ internal to the society ----------- (elders/cultural leaders)
--------------society in the present: the people
(representatives)

a) principle of differentiation
~ between dominant and
~ subordinate
ii) justifiable content of
rules ------
- - - - - - b) demonstration of common
interest uniting dominant and
subordinate
The Normative Structure of Legitimacy 73

pean Enlightenment they took a strongly anti-religious form,

involving the replacement of the moral and intellectual influence
of the priest or theologian by that of the secular philosopher.
Natural law doctrines played an important part in the assault on
the feudal order in modern Europe, but they are not necessarily
egalitarian; that depends upon how human nature itself is
defined, and in particular whether differentiation is made
between different types of nature. The Aristotelian version of
natural law justified slavery as well as aristocracy, while most
Enlightenment versions postulated a radical difference between
the nature of men and women. Theories of human nature, its
needs and capacities, still have an important part to play in
philosophical debates about legitimacy, but the idea of an eternal
moral law independent of human creation is difficult to sustain
in a world in which a multiplicity of different beliefs and practices
is so evident.
A further possible external source of legitimacy for a society's
rules of power is provided by the laws of science. Scientific
theories are logically incapable of generating normative principles
on their own yet the prestige of science in the modern world is
such that it is often appealed to as an authoritative source for
rules of social organisation. So laissez-faire was justified in the
first half of the nineteenth century by the science of political
economy, and in the second half by the theory of Social Darwin-
ism, though the latter was also appealed to by social imperialists
to justify collectivist organisation at home and the 'survival of
the fittest' abroad (Semmel, 1960, ch.2). In the twentieth cen-
tury, science has provided Soviet-type societies with a legitimat-
ing basis in the form of Marxism-Leninism, with its claim to be
able to guide society according to a scientific understanding of
the historical process, knowledge of which is vested in the party
as the representative of the most historically advanced class.
More generally, a belief in the validity of science sustains the
authority of experts in many fields, and justifies rules of organis-
ation for a whole range of activities where technical decisions
have been made paramount, to the exclusion of normative or
political choices. Wherever science serves as a source of legit-
imacy, it works in an anti-democratic direction, by assigning the
power of decision-making to the expert at the expense of the
citizen, and to the professional at the expense of the lay person.
74 The Legitimation of Power

The evolution of intellectual epochs from religion through

philosophy or metaphysics to science, first charted by Saint-
Simon, represents a major shift in the belief systems of societies,
in the character of their leading intellectuals, and in the mode
of discourse in which social rules have been legitimated: from
myth and cosmogeny, through philosophical argument to scien-
tific proof. However, the Saint-Simonian belief that science could
provide a definitive and non-controversial basis to social rules of
power was premised on the assumption that all issues of social
organisation in the industrial age could be reduced to technical
questions of production. For all the prestige of science today,
this conception of a wholly de-politicised social and legal order
is in principle unattainable, despite the frequent recurrence of
technocratic illusions to this effect (Saint-Simon, 1975).

b) internal sources
Whatever the main intellectual forms of reasoning and criteria
of validity in different historical societies, none have relied on
these external sources alone for their ultimate basis of authority.
All have been complemented by sources internal to the society.
Of these there are two major variants capable of providing a
source of legitimacy: society in the past (tradition) and society
in the present (the people). The first derives the rules of power
from the society's own past, through the sanctification of tra-
dition. There is a fine but important distinction to be drawn
between accepting rules because they have always been there
(habit) and the idea that those rules are the best for society
which have been tried and tested over time, and embody the
accumulated wisdom of successive generations (belief in tra-
dition). Tradition by its nature conveys authority upon 'elders',
and those whose role is to study and perpetuate a society's
cultural legacy. It also favours rules of succession, such as her-
edity, which embody continuity with the past. The traditionalist
principle makes changing rules difficult though not impossible;
traditions are usually rich and complex enough to allow precede-
nts for most things to be discovered. Yet the traditionalist prin-
ciple tends to be incompatible with continuous rule change, such
as is required by modern industrial economies. And its way of
privileging society's past is inconsistent with a conception of
history as one of unilinear development towards a more progress-
The Normative Structure of Legitimacy 75

ive future. For these, a source of legitimacy in the present society

is needed.
The most common source of legitimacy in contemporary societ-
ies is the 'people'. This source offers a highly generalised basis
of legitimacy for the political domain, from which the rest of
society's rules can in turn derive their legitimacy via the legislat-
ive process, freed from the limitations of tradition. Although the
characteristic institution of popular origin is the representative
assembly, the actual form of political arrangements can vary
widely according to the precise definition of the people, and the
type of representation adopted. Until the twentieth century the
concept of the 'people' usually excluded propertyless males and
all women. In Marxist-Leninist regimes the people have been
defined primarily as the working class, to be represented through
the monopolistic class party. And in all societies the definition
of the people as the 'nation' has carried both inclusive and
exclusive connotations, which have had important implications
for the territorial boundaries of the state and for different citizen-
ship statuses within it.
These various sources of legitimacy, both external and internal
to societies, are rooted in clearly distinguishable types of belief
system, each with its own respective interpreters and mode of
discourse. Among the most profound social changes are those
marked by a shift in these belief systems that determine the
source of legitimacy for a society's rules of power: from religious
to secular; from external and universalistic to internal and par-
ticularistic; from society as past, to society as its people in the
present. Such shifts in belief, as gradual as they are also funda-
mental, will leave a society's established power rules intellectu-
ally unsupported, like a bridge whose foundations have been
weakened by the slow processes of erosion. Thus the power of
a hereditary monarchy becomes vulnerable to a decline in the
belief in tradition, or the power of a theocratic order to the
onset of secularisation. A similar gap between rules and beliefs
can develop, conversely, through the attempt to introduce or
impose new rules of power in a context where the appropriate
beliefs are lacking: a representative democracy in a society that
is highly traditionalist in orientation, or a Marxist-Leninist
regime in a deeply religious society. All these are examples of
what can be termed a legitimacy gap or deficit, in contrast to
76 The Legitimation of Power

illegitimacy, i.e. a breach in the rules themselves. Whereas the

latter is relatively clear-cut, the former requires careful analysis
to elucidate, though one of its distinctive symptoms will be funda-
mental contestation between different groups of intellectuals.
Now although such disjunctions can develop between rules of
power and the beliefs necessary to support them, it is also impor-
tant to insist that any particular legitimating source does not
deliver an unequivocal set of power rules. The authority of
religion, for example, has historically been called in aid of a
wide variety of social and political arrangements, from individual
to collective property ownership, from absolute monarchy to a
Calvinist republic. Such a belief system will only exclude certain
possibilities, rather than unequivocally determine their form. By
the same token, it is possible for similar power rules to derive
their legitimacy from widely differing intellectual sources, as, for
example, gender rules have claimed validation successively from
religious doctrine, natural law and scientific argument. The belief
systems which provide the source of legitimacy thus constitute
only the most general intellectual framework within which legi-
timation of rules of power, and also challenges to them, may
take place. They serve to define the boundaries of possible belief
and debate, and the terms in which any debate must be con-
ducted. It is important therefore to underline that the belief
systems we have been considering at most locate an ultimate
source of authority for power rules, rather than determine the
particular form they should take. To justify the content of such
rules, i.e. who should have access to the means of power, on
what terms, etc., an appeal to more specific and substantive
principles is needed.

ii) legitimacy as justifiable rule-content

As I have indicated above, the demonstration of an authoritative

source for the rules of power does not suffice to answer the
question: why these rules? Why this particular arrangement of
the division of labour, of property, of political power? Behind
the variety of answers that can be given to this question, it is
possible to identify a common structure of argument. This struc-
ture comprises two distinct elements. The first is a principle of
differentiation or separation, which distinguishes the dominant
The Normative Structure of Legitimacy 77

from the subordinate, and justifies their respective access to and

exclusion from essential resources, activities and positions. The
second is a principle of community or common interest, which
links dominant and subordinate, and demonstrates the advantage
the latter derive along with the former from the rules of power
in question (see Figure 3.1, 72). As I shall argue, the first of
these principles, that of differentiation, may be sufficient for the
dominant to justify their power to themselves. But for their
power to be justified to the subordinate, the second principle is
required as well.

a) principles of differentiation
All social relations of power, and the rules of access and
exclusion on which they are based, presuppose the differentiation
or separation of the dominant from the subordinate. This separ-
ation is justified on the grounds that those who hold power
possess qualities lacking in those subordinate to them, and that
these qualities are appropriate to the particular form of power
that is exercised. Rules of power, in other words, are considered
rightful in so far as they select the qualified and exclude the
unqualified, and ensure the dominance of the superior and the
subordination of the inferior. How these qualities are defined
will vary according to the particular form of dominance and
subordination, but a broad distinction can be made between
those justifying principles which assume that the qualities are
assigned to and detectable in people at birth (so-called 'ascrip-
tive' theories), and those which hold that they are demonstrated
only through performance and achievement ('meritocratic' the-
ories).
Among theories of ascription, the principle of aristocracy has
been prevalent for much of human history. Its central idea is
that of noble birth: the superior qualities appropriate to the
possession of wealth and the exercise of power are supposedly
inherited at birth, and the more surely so, the longer the line of
noble ancestry that can be traced. It is the family into which a
person is born that determines his or her natural position in
society, and that confers the qualities appropriate to that pos-
ition. Such a belief is associated with land as the major source
of wealth, with traditionalism as the ultimate source of legit-
imacy, and, for all the tensions between the two, with a heredi-
78 The Legitimation of Power

tary monarchy deriving its justification from the same aristocratic

principle. Aristocratic societies generally provide scope for new
talent, whether in political service or mercantile activity; but the
achievements are only publicly recognised once they are trans-
lated into titles and estates through gift or purchase. Even then
it takes a generation or two for a new family to become accepted,
since it is inheritance, not achievement, that provides the guaran-
tee of genuine quality. As with fine wine, so with breeding: only
the passage of time can confer excellence (Powis, 1984).
While the principle of heredity has provided a historically
important justifying ground for the differentiation of power and
life-chances, that of gender has been even more universal. Here
the distinguishing element is not the family into which a person
is born, but the physical characteristics of sex that are supposed
to reveal the naturally determined qualities appropriate to given
social roles. The idea of women as the 'weaker sex' justifies their
exclusion from a range of activities that involve the exercise of
power, or lead to its acquisition. The definition of women as
natural nurturers justifies their confinement to caring and servic-
ing roles, for which men in turn are considered unsuited. The
concept of the 'feminine' represents women as a passive adorn-
ment, rather than endowed with active intellectual or deliberative
capacities, such as could be exercised in a public realm. These
supposedly innate characteristics are typically combined into an
ideal of womanhood, which is more positive than simply the
absence of masculine qualities. Such an ideal does more than
justify the sexual division of labour. It serves to obscure the
power relationship between the sexes under the rubric of 'differ-
ent but equal', and to reconcile women to the limitations of their
condition by offering them a positive image to which to aspire,
and a domain of their own, the domestic, within which to exercise
a carefully circumscribed power (Millett, 1971, pp.88-108; Okin,
1980, pt.3).
The example of gender demonstrates clearly, however, that
what is at issue in the legitimation of power rules is not only the
development and dissemination of an appropriate body of ideas,
or ideology, but the construction of a social identity by a complex
set of often unconscious processes, which make that identity
seem 'natural', and give the justifying ideas their plausibility.
The processes whereby a differentiated gender identity is con-
The Normative Structure of Legitimacy 79

structed in girls and boys take place through various stages from
infancy to adolescence. They are ones through which children
develop qualities, attitudes and capacities, learn patterns of
behaviour, form desires and expectations, that are appropriate
to their adult gender roles and relationships with the opposite
sex. Although different theories in the fields of sociology and
social psychology stress different mechanisms responsible for the
construction of such an identity, few see it as the product simply
of conscious instruction on the part of parents and other authority
figures. More important is the exemplary effect of relationships
between parents themselves; their less conscious assumptions and
expectations for their children; the role-models to which the
children are exposed both within the home and outside it; the
reinforcement of attitudes and behaviour by siblings and peer-
group pressure; the stereotypes presented in stories, books and
other media. The cumulative effect of these influences is the
construction of a social identity within the personality which,
while by no means uniform within each sex, is systematically
different between them (Oakley, 1972; Chodorow, 1978).
What has this to do with legitimacy? Central to the justification
of gender roles, as of a hereditary aristocracy, is the demon-
stration that those assigned to their respective roles and hier-
archically ordered positions are justly so assigned, because they
are naturally endowed with the qualities and capacities appropri-
ate to them. Yet the very 'naturalness' of this endowment is
itself socially constructed, not only in the realm of intellectual
doctrine or ideology, but by the complex set of processes
whereby different groups from birth onwards develop the charac-
teristics appropriate to their future roles in such a way that
they appear as naturally given. An ascriptive principle of social
differentiation, in other words, by requiring from birth the devel-
opment of differential qualities in different groups, and excluding
the development of other characteristics, itself generates the evi-
dence necessary to its own justification.
The significance of such evidence can be grasped by asking
what would be required for people to make the crucial but
difficult distinction between naturally determined and socially
acquired attributes and capacities. Either there has to be external
evidence provided from other societies in which social roles are
organised differently. Or there has to be a process of social
80 The Legitimation of Power

change at work internal to the society, which allows or requires

the supposedly ill-endowed by nature to assume roles involving
the acquisition and exercise of power. The erosion of the aristo-
cratic principle in Europe resulted from the expansion of new
roles of this kind in economy and state, to the point where the
capacities of the ignobly born could no longer be regarded as
exceptions to be accommodated within the aristocratic principle.
Such a process was a necessary condition for political philos-
ophers from Hobbes onwards to make the critical distinction
between 'natural' and 'political' (i.e. socially constructed)
inequalities. Even though gender differences were still assigned
to the 'natural' sphere, the distinction, once made, was available
as a weapon to be used by later advocates of women's equality
(Wollstonecraft, 1970; Mill, 1970). But it has taken the evidence
of women successfully occupying supposedly 'masculine' roles for
an effective challenge to be mounted upon the assumed natural-
ness of gender characteristics. The same social changes, in other
words, which open up roles of power to the previously excluded,
are also necessary to weaken the evidence that justifies their
exclusion.
It is one of the basic contentions of a society which makes
access to key resources, activities and authority positions subject
to open competition, that it avoids the self-fulfilling cycle of
closed societies and structures of power, whereby characters are
moulded to fit the future roles they must necessarily occupy. The
superiority claimed for meritocratic over ascriptive rules is that
the processes of open competition, whether operated through
the market, the education system, or appointment procedures,
ensure that those who gain access to the means of power in any
sphere are indeed the most suited to it, and have demonstrated
evidence of achievement as a condition for attaining it. More-
over, for these processes to be effective, there must be equality
of opportunity, and a principle of non-discrimination on grounds
of irrelevant characteristics; i.e. a basic presumption of equality.
In practice, however, the meritocratic principle is substantially
restricted in its operation by the disposal rights of property own-
ership (doing what you want with your own) and the kinship
claims of the family (doing what you can for your own). Together
these ensure that the material and cultural resources enjoyed by
the powerful in one generation are transmitted to the next, so
The Normative Structure of Legitimacy 81

that the acquisition of the means of power is as much the result

of good fortune as it is of personal effort, of social advantage as
of relevant ability.
There is a considerable divergence, therefore, between the
principle of meritocracy and the social arrangements which it is
required to justify, though the extent of that divergence will vary
from one society to the next (Payne, 1987). Such discrepancy is
obscured by the fact of competition itself, which serves to vali-
date the qualities of the successful, however these have been
acquired. And the celebration of impressive examples of ascent
from base to summit of the social order provides evidence that
access to power is in principle open to anyone, though not every-
one, whatever the odds that may be stacked against them in
practice. Furthermore, as far as the losers in the competition for
power are concerned, the transmission of appropriate expec-
tations and attitudes along with an impoverished cultural and
material legacy suffices to reconcile them to their subordinate
position. In this respect the open or meritocratic society is not
altogether as different as it appears from the closed or ascriptive
one in the manner in which children develop the characteristics
and expectations appropriate to their future social roles, or in
the agencies through which they do so. Studies of socialisation
among the manual working class closely parallel those of gender
in their demonstration of the way in which attitudes and ideas
appropriate to their more limited future expectations are repro-
duced within the family and peer group (Willis, 1977). To this
extent 'meritocratic' societies are unable, in practice, wholly to
break out of the self-fulfilling cycle of legitimation, characteristic
of ascriptive societies, whereby natural-seeming qualities are soci-
ally created.
Such a restriction of the meritocratic principle in practice is
the result of an inability to prevent the intergenerational trans-
mission of social advantage. To overcome this limitation would
require either the abolition of the family, or the removal of the
very inequalities whose allocation to particular individuals it is
the purpose of the principle to justify. Yet even if, per impossible,
a pure meritocracy could be attained, it would still operate as a
principle of differentiation or selection, justifying the access of
certain individuals to power, and the exclusion of others from it
(Young, 1961). Like any principle of differentiation, what it
82 The Legitimation of Power

could not justify would be why this particular distribution of

power was necessary in the first place. Why this particular distri-
bution of resources, division of labour, or position of authority?
A principle of differentiation, in other words, can serve to justify
why individuals of a particular kind should have access to power
or be excluded from it, but not why the division itself should
exist. For that, some argument from social necessity or social
utility is required.

b) ideas of the common interest

One part of the justification of power-rules lies in the principle
of differentiation, which distinguishes the qualified from the
unqualified. It emphasises the difference in personal capacities,
so as to justify the separation between the lives and powers
of the dominant and those of the subordinate. Its necessary
counterpart, however, is the belief that the dominant and the
subordinate, however much they may differ, are also linked by
a community of interest; and that the distribution of power serves
the interests of the subordinate, and not those of the powerful
alone. A sense of justified superiority may well suffice to rec-
oncile the dominant to their power, its advantages and its impo-
sitions on others, with an easy conscience. But in the absence
of any idea of community linking themselves with the subordi-
nate, or of the latter as having any interests of their own worth
protecting, even if those of inferiors, then that power can have
no justification for the subordinate; the latter will be excluded
not only from power itself, but from participation in a shared
moral order within which alone, any justification could make
sense to them.
A characteristic feature of power which seeks to justify itself
to the subordinate, therefore, is the claim to serve a common
interest. Feudal power justified itself in terms of the provision
of law and protection to the peasantry subject to it, and the
exactions on their labour as necessary to the fulfilment of these
functions. In a different age, capitalist property and profits, and
the accompanying authority of management, justify themselves
in terms of their necessity to expanded production and employ-
ment, and the satisfaction of consumer wants. In state-socialist
societies, the power of the bureaucracy has been justified as
necessary to the effective organisation of production, distribution
The Normative Structure of Legitimacy 83

and most other essential social activities. Sceptics may regard

such claims as spurious, serving merely as a cloak for the particu-
lar interests of the powerful. In doing so they pose a mistaken
dichotomy between general and particular interests. As I have
already argued, it is typical of the way social power is organised
that general interests, including those of the subordinate, can
only be met through satisfying the particular interests of the
powerful; and therefore their claim to serve general interests is
rarely an empty one. It only becomes so when a dominant group
loses its social function, or fails to satisfy vital interests of the
subordinate.
As I pointed out in the previous chapter, once some necessary
social resource or activity is controlled by a particular group, it
follows that the wider needs of society can only be met by
satisfying the interests of that group, and on terms acceptable to
them. Provided the resource or skill they command has been
acquired according to the rules, making it available toothers for
a consideration will seem less like an extortion racket than a
public service, and the price demanded as their legitimate reward
for performing it. Where landowners rent out their land to a
peasantry, the return they can command in kind or labour is
the due reward for providing the peasants with the means of
subsistence. 'If the landlords did not let us rent the land, we
would starve,' said the peasants in William Hinton's account of
discussions in the village of Fanshen during the Chinese revo-
lution (Hinton, 1966, pp.128-31). As the relationship presented
itself to the peasantry, their existence depended upon the land-
lords, rather than the existence of the landlords upon their
labour, and therefore the rent exacted was acknowledged as
morally justified.
From such a perspective, it is the failure to guarantee subsist-
ence and the means of livelihood that is destructive to legitimacy,
rather than the extent of any exactions made, since it infringes
the basic interests that are presupposed in the relationship. This
is a point made in many studies of peasant economies. James
Scott, for example, argues that it is not the proportion of the
crop, but its effect on subsistence that is crucial to the perception
of exploitation:

'If the balance of exchange is deteriorating but the material

84 The Legitimation of Power

situation of the cultivator's family is stable or even improving,

discontent may be evident, but it is unlikely to provoke mass-
ive unrest. It is when a worsening balance of exchange menaces
crucial elements of subsistence routines, when it stretches exist-
ing subsistence patterns to breaking point, that we expect
explosions of rage or anger.' (1976, p.177)

As he goes on to show, the basic peasant interest in subsistence

becomes part of a settled pattern of expectations, a 'norm of
reciprocity', which imposes obligations on those who monopolise
the scarce resource of land, and which is infringed when exac-
tions are made, or continued, to the point where they threaten
subsistence. The historical circumstances in which this can
happen, and in which the legitimacy of landed property become
eroded have been enormously varied.
The distinctive characteristic of industrial capitalism in its early
stages in Europe, especially in comparison with systems of rural
tenancy, was that from the standpoint of the workers it was a
power relationship in which their interests were systematically
disregarded. The capitalist undertook no responsibility for the
livelihood of the workers, who could be employed at a wage
insufficient for subsistence, and could be dismissed at will into a
situation where they had no means of livelihood at all. The
worker was simply an instrument for serving capitalist profit, on
terms and conditions dictated by the capitalist, and without any
security of subsistence. The argument that capitalist property and
profit served a general interest could only be made at the level
of the market economy as a whole, and in terms of the abstract
constructions of theoretical political economy: viz., that the
worker's insecurity was necessary to ensuring efficiency of pro-
duction and its responsiveness to consumer needs in a competi-
tive market; and that the accumulation of profit ensured
expanded investment and production, and enabled an ever-
increasing population to be employed (Smith, 1976, bk.2, ch.3).
None of these arguments addressed the immediate interests of
the workers as producers. At the point where the power relations
of capitalism were experienced directly, in the individual factory,
their interests went unattended. The public benefits of the
system, such as they were, could only be demonstrated through
the working of the economy as a whole, and in the theoretical
The Normative Structure of Legitimacy 85

terms of a political economy whose purpose was to prove the

intuitively implausible proposition, that the unrestrained pursuit
of private advantage by the capitalist served the public interest
by way of its unintended consequences via the 'hidden hand' of
the market.
In the light of this disjunction between the official justification
for capitalist profit and the actual experience of the workers it
is hardly surprising that many of them were attracted to an
alternative, socialist-oriented political economy, which directly
addressed their experience as producers, and promised the
humane working conditions, the security of employment and the
right to the product of their labour that capitalism denied them.
Capitalism itself was only able to attain a measure of legitimacy
as, on the one hand, the workers were able to secure a share of
expanded profits in improved wages and working conditions
through the pressure of collective organisation at the workplace;
and, on the other, as the state came to guarantee the means of
subsistence to them when unemployed, and in general compen-
sated for the inadequacies of the market. It has been Marxism's
insistence that these necessary conditions for capitalism's legitim-
ation could not be achieved without a chronic crisis of profit-
ability that has proved to be the central weakness of its theory,
for all its impressive analysis of the condition of labour under
the capitalist mode of production.
Whereas capitalist legitimacy has proved weakest in the early
stages of industrialisation, and stronger in the later stages as
workers have secured a share of expanded profits, and the state
has alleviated the conditions of insecurity inseparable from the
market economy, the record of state socialism has proved to be
the reverse. Its claim to be able to satisfy the public interest
directly, through the planned organisation of production, rather
than abandoning it to the vicissitudes of the market, seemed
initially vindicated in the USSR by its impressive achievements
in industrialisation and the development of the necessary human
and material infrastructure. For this reason state socialism has
offered an attractive model to other developing societies. How-
ever, the central planning system has proved far too insensitive
for the more intensive stages of economic development, and its
producer-goods bias has left basic consumer wants chronically
unsatisfied, to the point where bureaucratic power and privilege
86 The Legitimation of Power

have become divorced from the effective performance of the

economic purpose that has provided their rationale (Bahro, 1978;
Feher et al., 1983). What we have witnessed is the classic erosion
of legitimacy of a power structure through its inability to satisfy
the basic interests it claims to serve. The comparison offered by
capitalism at its most successful stage of stimulating and satisfying
consumer wants has no doubt contributed to this erosion.
The upshot of these examples drawn from different systems of
production is that the control exercised by dominant classes over
essential resources and activities, which ensures that general
social needs or interests, including those of the subordinate, can
only be met on terms acceptable to them, achieves justification
provided that these interests continue to be satisfied, and that
the claim to this effect is validated by the experience of the
subordinate themselves. However, there is an important distinc-
tion to be observed, as I remarked earlier, between the failure
of a system of power to serve the interests it claims, and the
absence of any such claim in the first place. The distinctive
feature of most slave systems is that they did not accord any
place to the interests of the subordinate whatsoever, nor see
them as part of a common moral order. Slave-owners certainly
justified the institution of slavery to themselves, and to third
parties whose cooperation was needed for its survival. In this
sense slavery constitutes a paradox for legitimacy. If we are
interested in explaining the rise and fall of slavery as an insti-
tution, then we will want to talk about the 'legitimacy' of slave-
owning within a given society, and examine the history of its
justifications. But if we are concerned to analyse the character
of the power relationship between slave-owner and slave, then
the important thing is that the form these justifications took
posited such a radical differentiation of status between the two
that no considerations of common interest between them were
possible; nor could there be any basis of right or morality in the
relationship. As the eighteenth-century historian Bryan Edwards,
himself a slave-owner, wrote:

'In countries where slavery is established, the leading principle

on which the government is supported is fear: or a sense
of that absolute coercive necessity which . . . supersedes all
questions of right.' (Genovese, 1970, p.162).
The Normative Structure of Legitimacy 87

For the analysis of legitimate power relations, then, slavery con-

stitutes the limiting case: the example of purely coercive, unlimi-
ted, arbitrary power of one group over another, where legitimacy
is absent. 'All slaves are enemies' went the Roman proverb; they
were indeed recruited through conquest, and the relationship
formed a perpetuation of the conditions of warfare within the
society (Hopkins, 1978, pp.118-23). As with all coercive regimes,
power would dissolve when coercion slackened, as was demon-
s~rated by the spectacular examples of mass slave desertions in
the ancient world when their owners were away fighting (De
Sainte Croix, 1981, p.147). However, even such a regime would
need incentives as well as sanctions as a means of getting tasks
done, and to lighten the weight borne by physical coercion in
maintaining discipline. Yet it is important to distinguish between
the use of incentives to encourage cooperation, within a predomi-
nantly coercive regime, and the claim that a power structure is
justified because it serves the interests of the subordinate. The
first is a purely instrumental means to promote the interests
of the powerful; the second involves the recognition that the
subordinate have interests of their own which merit consider-
ation.
In this context the differences that Eugene Genovese identifies
between different historical types of slavery are relevant. He
draws attention, for instance, to a shift that occurred in the
pattern of slavery in the southern states of the USA after the
end of the slave trade, when the supply of slaves could only be
maintained through reproduction. This entailed allowing a cer-
tain independence of family life to slaves, usually on estates
where the slave-owner was also resident, and acknowledging that
they had interests of their own meriting attention. Here the
justification of slavery took a distinctively paternalist turn: slaves
occupied the status of 'permanent children' within an extended
family, where they were said to enjoy a greater degree of both
physical and economic security than outside the institution of
slavery, under a system of supposedly 'free' labour. In other
words, slavery was in their own best interests! Genovese con-
trasts this situation with the slave regimes of the West Indies,
with absentee landlords and dependence upon the slave trade
for new recruitment, where the relationship was the historically
88 The Legitimation of Power

more typical one of coercion rather than dependency (Genovese,

1970; 1975, pt.l).
Genovese's distinction between different types of slave regime,
the paternalistic and the purely coercive, is important for mark-
ing one of the boundaries between a legitimate and a non-legit-
imate relationship: between one based on a recognition of the
interests of the subordinate (and, therefore, with some corre-
sponding limitation on the exercise of power) and one where no
such recognition exists. In the practice of slavery in the southern
states of the USA no doubt this distinction was more blurred
than I have presented it, but the distinction is an important one,
analytically. Genovese's example of paternalism also raises a final
question to be considered here, and that concerns the different
ways in which the interests of the subordinate are treated in a
power relationship, and the correspondingly different types of
relationship this treatment implies.
A paternalist form of power involves a relationship in which
the subordinate are defined as wholly or partially, temporarily
or permanently, incapable of recognising and defending their
own interests, and these therefore have to be defined and
advanced by the powerful on their behalf. It can be contrasted
with a form of power in which the subordinate are defined as fully
capable of identifying their interests, and therefore as entitled to
be consulted in any decision that may affect them; in which,
therefore, processes of consultation are formally or informally
built into the relationship. The paradigm case of a paternalist
relation is of course that between parents and children. Here the
principle of differentiation that justifies the power dimension of
the relationship (adult-child) specifies a relative incapacity in the
child that requires the parent or guardian to make provision,
take decisions and secure obedience in the child's own interests.
Yet the incapacity is only a temporary one. In the case of women,
the principle of differentiation between the sexes that has histori-
cally justified male supremacy has defined them as permanently
weaker, less capable, less rational, and so on, and their interests
as therefore having to be determined and protected on their
behalf by men. The logic of racial differentiation has been simi-
lar. In each case it is some presumed weakness or incapacity
inscribed in the principle of differentiation that defines the subor-
The Normative Structure of Legitimacy 89

dinate as necessarily dependent, and their interests as requiring

interpretation and protection by a superior.
There is a different form of paternalism which is to be found
in the political domain, and which should be distinguished from
the above, because it is not necessarily based upon any differen-
tiation of inherent status between categories of people. This is
where the fundamental belief system that specifies the ultimate
source of authority for the political domain has implied in it the
premise that decisions about the public interest, the common
good, etc., must be matters of special knowledge, and that those
who have attained this knowledge are thereby entitled to deter-
mine policy, decide laws, etc., on behalf of those who haven't.
If the ultimate source lies in tradition, then it will be the elders,
or those whose families have privileged access to traditional
wisdom, who will acquire the right to decide. If it lies in divine
command, then it will be spiritual leaders, or those qualified in
religious law, as in a theocracy. If it lies in science, then it will
be those who have access to scientifically determined principles
or expertise about the proper organisation of society, or the
necessary course of history (as in Marxist-Leninism). In each
case the underlying principle of authority entails that some
people are entitled to take public decisions on behalf of others
by virtue of some special knowledge they have acquired about
the public good, and which others do not have. All such prin-
ciples by definition entail a paternalist relationship between gov-
ernors and governed: 'trust us, because we must know best'.
The only principle of government that is non-paternalist is one
where the ultimate source of authority is located wholly in the
people, because only here is the criterion for the public good
to be found not in some special wisdom, revelation or expert
knowledge, but in what the people, freely organised, determine
it to be, whether directly or through their representatives. Only
here are those subordinate to government recognised as the ulti-
mate judge of what their interests are. I say 'wholly' in the
people, because the distinctions I have made are typological
only, and there are mixed sources of authority where, say, a
representative assembly has limited rights of decision or consul-
tation in relation to a hereditary monarch or religious ruler. The
principle of popular sovereignty can be qualified formally, in this
and other ways, by a traditionalist, religious or scientific source
90 The Legitimation of Power

of authority. It can also be qualified informally, when those who

are popularly elected claim access to a privileged source of truth
that others do not have, and where consultation with the subordi-
nate becomes thereby irrelevant or a charade. There are, finally,
other kinds of mixed system where government is constituted on
paternalist lines with one section of the society, and on non-
paternalist lines with another, as when voting rights are limited
on grounds of property qualification, gender or race. (Here it is
the differentiation of inherent characteristics that is the determi-
nant).
The problem intrinsic to all paternalism is that, in denying the
subordinate all independent means of expressing or defending
their interests, there is nothing to stop it degenerating into the
exercise of power in the interests merely of the powerful, except
the latter's own integrity; and even integrity is no guarantee
against self-deception or ignorance. Moreover, for its legitimacy,
paternalist power depends upon the continued belief on the part
of the subordinate in the principle of differentiation, or the
source of authority, that provides its justification. Among the
complex of factors contributing to the erosion of such beliefs, not
the least is the experience of interests systematically neglected or
impaired because they are subject to paternalistic definition by
the powerful.
In this section I have briefly considered the way in which
different systems of production meet the interests of subordinate
classes, or fail to do so; the limiting case of slavery which makes
no such claim; and the link between different types of paternalist
power and the different sources of authority and principles of
differentiation that define the subordinate, for one reason or
another, as incapable either of determining their own interests
or of having a valid conception of the public interest for them-
selves. The final section will discuss problems relating to consent.

Legitimacy through expressed consent

The subject of consent is a confusing one, and its relationship

with legitimacy makes it doubly so. If by consent to power we
mean a condition of voluntary agreement to it, then what counts
as voluntary, and what sort of evidence is needed to demonstrate
The Normative Structure of Legitimacy 91

such agreement? Simply obeying the requirements of the power-

ful, or keeping to the law is not enough, since obedience can be
maintained by coercion. The view that people consent to power
if they are not openly protesting against it is therefore quite
inadequate. What sort of evidence, then, do we need?
These questions become even more difficult when we try and
relate consent to legitimacy. If we equate legitimacy with the
'belief in legitimacy', then do we say that the belief in legitimacy
is something different from consent, because more normative?
Or that it is part of consent, because it helps produce it? Or
that consent provides us with evidence of people's belief in legit-
imacy? If the last of these, then again what sort of evidence will
count?
Most of these problems disappear once we free ourselves of
the mistaken view that legitimacy is to be equated with people's
belief in it. Then we are able to see that what is important about
consent is not the condition of voluntary agreement, but the
specific actions that publicly express it; and that these are impor-
tant because they confer legitimacy on the powerful, not because
they provide evidence about people's beliefs. They confer legit-
imacy because they constitute public expressions by the sub-
ordinate of their consent to the power relationship and their
subordinate position within it; of their voluntary agreement to
the limitation of their freedom by the requirements of a superior.
Such actions are not the only source of legitimacy for the power-
ful; but they make a distinctive contribution to it, along with
validity according to justifiable rules, by demonstrating the con-
sent of the subordinate to the relationship.
What sort of actions, then, will carry this significance? The
answer of the liberal, individualist tradition, which has become
embodied in the practice of Western societies, has typically been
that consent is given to a condition of subordination only by a
specific and voluntary promise, agreement or contract on the
part of each individual in person: e.g. of the worker to the
employer at the time of employment, of the woman to her
husband at the point of marriage (I am referring here to the
promise of obedience contained in the traditional form of mar-
riage ceremony). Necessary conditions for the voluntariness of
such an agreement are not only that it should not be coerced,
but that there should be a choice available between potential
92 The Legitimation of Power

employers and husbands respectively. One of the classical liberal

objections to feudal society was that the worker had no choice
between employers, and to arranged marriages that the woman
had no choice of husband.
As regards the political sphere, the early liberal approach to
consent was to develop the contract model into the idea of a
social contract, whether historical or hypothetical, made by all
individuals at the original establishment of government (Lessnoff,
1986, ch.4). However, the problematical status of this fictional
social contract has led to a different approach today, which
focuses instead on elections as the source of legitimacy. The
convention within contemporary liberal democracies is that it is
the act of taking part in elections that legitimates government
and secures the obligation of citizens in principle to obey it.
Here again, it is the existence of choice that is crucial - between
candidates, programmes and parties. In other words, in the
sphere of government as in that of employment or marriage, it
is making an agreement to subordination under conditions of
choice between alternatives that confers legitimacy on the exer-
cise of power, and a corresponding obligation to obey.
I have begun with the liberal or individualist model of consent
because it is familiar and paradigmatic for our society. However,
it does not follow that this is the only form of action that can
demonstrate consent or confer legitimacy on the powerful, or
that there are not other types of action which, while failing to
meet the liberal criteria for voluntary agreement, may not have
or have had a legitimating force within the conventions of differ-
ent social systems. Let me mention three such types of action
that have historically been seen as conferring or confirming legit-
imacy, and which continue even within a liberal society, albeit
in a subordinate role.
First is swearing an oath of allegiance. This of course is a form
of promise, but differs from the liberal model in that typically
there is no choice about whom a person is to obey. However,
in societies where such oaths are required of subordinates, the
lack of choice has not been regarded as making the promise any
the less binding (Bloch, 1961, ch.ll); nor has it in the case of a
bride in circumstances of an arranged marriage. Even today
when members of armed forces swear an oath of loyalty to the
commander in chief the oath is not associated with any choice
The Normative Structure of Legitimacy 93

about whom the commander is to be. In traditional systems of

rule such an oath given by the most important figures within the
society was a significant element in the legitimation of the leader,
chief or monarch. It carried an exemplary force for others, as
well as a binding one for those who took part; for the ruler to
show that he or she had the consent of those who carried weight
in the society or in its different localities was an important and
necessarily public aspect of legitimacy. By the same token, in
countries subject to conquest, the ceremonies in which the estab-
lished leaders of the conquered paid homage and swore
allegiance to the conqueror, were an important element in the
legitimation of alien rule (see, for example, the institution of the
durbar or the indaba under the British Empire) (Cohn, 1983).
A second type of action expressive of consent is that of taking
part in consultations or negotiations with the powerful, either
about aspects of policy or about the terms on which a particular
service is rendered, which culminate in agreement. In some
respects this resembles a contract, but it does not entail any
choice about who the superior will be, does not take place at
the outset of the relationship, and may only affect certain aspects
of it. Yet it carries with it, not only a commitment on the part
of the subordinate to support the policy or observe the terms
agreed, but an implication that they acknowledge the authority
of the powerful more generally, and this act of public recognition
itself confers legitimacy. In traditional societies the regular con-
sultation between a ruler and other powerful figures provided a
reinforcement to an original oath or agreement of allegiance. In
early modern political systems the point of parliamentary elec-
tions was not to choose the head of government, but to select
representatives who would negotiate with the monarch the terms
on which taxation was to be levied and for what purposes. In
present-day society, collective bargaining by trade unions with
an employer or employers over the terms of the labour contract
confirms the latter's legitimacy much more than any individual
contract of employment (which is one reason why trade union
leaders are so much the butt of attack by critics of capitalism).
A third type of action expressive of consent, which differs
from the others in that it does not carry any connotations of
promising, is the public acclamation whether of a monarch upon
coronation or of a popular leader at a rally or other notable
94 The Legitimation of Power

event. The history of the twentieth century has made us deeply

suspicious of such occasions, especially where the organisers have
offered inducements to the masses to attend, or sanctions for
non-attendance. However, there is no denying that demon-
strations of popular support and mass mobilisation confer a dis-
tinctive legitimacy in the age of popular sovereignty; and that a
party or movement that can harness the enthusiasm and commit-
ment of a mass following thereby enhances the prestige and
authority of those who lead it. The most impressive example is
a revolutionary movement in which the people are ready to
sacrifice their lives to overthrow a discredited government, or to
support the one that replaces it. Such commitment confers enor-
mous moral authority on a new regime; and the continued mobil-
isation of the people to help carry out its policies can prove an
effective alternative to the electoral process as a means of popu-
lar legitimation. This is the expression of consent through mass
participation in activity manifestly supportive of a regime.
As I shall argue in Part II, the general acceptance of the
principle of popular sovereignty in the modern world has meant
that, whatever the precise form of political system, and however
qualified the principle may be in the rules of appointment to
office, political legitimation has to be mass legitimation; but it
does not necessarily have to take the form of the liberal contract,
or be exercised through the electoral choice of government. It
is this requirement of mass legitimation that forms the decisive
point of difference from a traditional order, in which the right
to express consent (equivalent to membership of the political
community) was limited to the privileged or propertied section
of society, whose consent was both exemplary and binding for
everyone else. For political legitimation to be effective in the
modern world, the expression of consent has in principle to be
available to all, whether they take advantage of it or not, and
whether it takes place in the electoral mode or the mobilisation
mode, as I shall call it.
There is, however, a crucial difference to be observed between
the electoral and the mobilisation modes of consent, or the con-
tractual and expressive modes, to give them a more general
characterisation, in respect of the way they confer moral author-
ity upon the powerful. A contract confers moral authority
through the promise that is made and the commitment that is
The Normative Structure of Legitimacy 95

undertaken for the future. It matters not at all for what reason
or motive the promise is made. Indeed, it can be made purely
out of self-interest. If I judge it in my interest to enter into a
particular contract, then I may do so regardless of my attitude
in general to the authority concerned; but my action carries its
obligatory force and legitimating effect regardless, because of the
kind of action it is. In other words, contracts can harness people's
self-interest to the cause of legitimacy through the normative
force of promising. (This still leaves intact the distinction
between the prudential and moral grounds for obeying a particu-
lar requirement of an authority.)
The expressive mode of consent, on the other hand, whereby
the subordinate demonstrate their commitment to a regime by
voluntary actions supportive of its policies, carries no obligation
with it, since it involves no undertaking in respect of the future.
This means not only that it has continually to be demonstrated,
but that its normative or legitimating effect depends entirely
upon the quality of the action undertaken. If performed purely
for self-advancement, because, say, it is the means to further
one's career, rather than for the promotion of a cause or a
collective project, then its expressive or normative impact will
be diminished. This distinction may be difficult to judge from
outside, but it will be evident enough to people within the
society. The expressive or mobilisation form of consent, there-
fore, is typically effective when linked to an ideological pro-
gramme or cause; and it tends to degenerate once enthusiasm
for the cause wanes, unless it is revived through some form of
cultural revolution or revivalist campaign. A degeneration into
mere self-interest does not undermine legitimation in the contrac-
tual mode, in contrast, since its normative force stems from the
action itself, not from the quality of the motives, or the degree of
enthusiasm, with which it is undertaken. For this reason popular
legitimation in the contractual mode can survive in societies
where self-interest is paramount, and where the population is
widely de-politicised, except at election time.
The fact that the expressive or mobilisation mode of consent
is open to manipulation by the powerful, and to the abuse of
people's commitment, when it is associated, as it typically is,
with paternalist and unaccountable forms of power - all that is
evident enough from the history of the communist systems. Less
96 The Legitimation of Power

obvious, because more subtle, is the way in which power relations

structure the expression of consent in the liberal or contractual
mode. To explain this, let me return for a moment to the liberal
criterion for the voluntariness of an agreement to accept a pos-
ition involving subordination: that people should have a choice
between those to whom they are to be subordinate. What they
have much less of a choice about, however, is whether to make
that choice in the first place, because the situation is such that
their interests and therefore their choices are already pre-struc-
tured.
This is most obviously true of the wage-labourer. Once
workers were 'freed' from any control over their own means of
production, they had the Hobson's choice of submitting to given
conditions of subordination under whatever employer, or star-
vation. The fact that today's alternative to subordination is the
dole does not alter the position that the choice is already struc-
tured by the existing rules of power determining access to, and
the use of, property. For most of the modern period women
have had the choice of being subordinate to their fathers or their
husbands; or, if they achieved independence through remaining
unmarried, of forgoing the possibility of having children. And
today the choice of who undertakes childcare is structured by
the relative opportunities and wage levels in the labour market,
as well as by the socially developed desires and capacities of
men and women respectively. For this reason women within a
supposedly equal relationship may have a strong incentive to
agree to a strategy that perpetuates their disadvantage and
dependency. In the political sphere, finally, citizens who do not
vote may forfeit any influence on the outcome, and in any case
will not avoid subordination to government by abstaining. The
alternative to expressing consent by participating in the ballot,
and thereby legitimating the government, is hardly a serious one
(Pateman, 1980; 1985, ch.S).
There is a further point to be made. In agreeing to a particular
condition of subordination, people are also thereby confirming
the rules of power that underpin it. In strict logic this does
not follow. For a person to agree to a particular condition of
subordination is not to agree to the rules that determine the
necessity of their subordination in the first place. Yet in practice
and by social convention it is exceedingly difficult to separate
The Normative Structure of Legitimacy 97

the two. Is there any doubt that, when regularly in a society,

women agree to obey their husbands in marriage, or workers
their employers, or citizens their government, they thereby con-
firm the relations of dominance and subordination in general, of
which their agreement is a part? We thus see once more the
self-confirmatory circle at work between rules of power and the
process of their legitimation. The existing rules of power them-
selves structure the condition of relative powerlessness, disadvan-
tage or dependency that gives people the incentive to make
agreements of subordination which in turn confirm the existing
rules of power.
At this point, defenders of the liberal democratic system will
insist that, in a society where freedom of speech and association
pertain, it is open to anyone to campaign with others for a
change in the rules; and the fact that they persist can only be
because people in general do not want to change them. Such a
claim points to an important principle, as I shall argue later. Yet
what it overlooks in practice is the way in which the structures
of gender and class power are themselves reproduced within the
political domain, and the political domain is itself structured, in
a manner that is weighted against such change. The self-con-
firmatory circle is less closed, to be sure, in the liberal democratic
system than in others, just as it is less closed in an avowedly
meritocratic than in an ascriptive order; and this reveals some-
thing significant about the principles that underpin these forms
of power. For the moment, however, what I wish to draw atten-
tion to is the way in which the processes that reproduce legit-
imacy in practice occur within these types of power relation also.

Conclusion

In this chapter I have explored the threefold normative structure

of legitimacy as validity according to rules, the justifiability of
rules in terms of shared beliefs, and expressed consent on the
part of those qualified to give it; and I have elucidated the inner
rationale of each, as well as the different forms each can take.
Of the three, the second is the most complex, involving in turn
an authoritative source for the rules, a principle of differentiation
between dominant and subordinate, and a common interest that
98 The Legitimation of Power

the system of power serves. All these are cumulative and comp-
lementary, rather than alternative, requirements for legitimacy.
By reference to them slavery through conquest can be seen as
the limiting case of a non-legitimate order, because no common
interest between dominant and subordinate is allowed by the
principle of differentiation between them. Caste systems and
serfdom constitute historical examples of orders that were legit-
imate without being legitimated, since the subordinate belonged
to the category of the unfree, and were hence unqualified to give
consent (though in the case of serfdom the fiction was observed
that consent to subordination had been given by the ancestors
of those enserfed.) Even in these examples the character of
their power relation can be elucidated by means of the threefold
structure of legitimacy I have outlined. Although I have pres-
ented the different components of legitimacy as separable
elements for purposes of analysis, for any given structure of
power it is the ensemble that is important, and the internal
connections between them, as I shall demonstrate in the account
of political systems offered in Part II.
In the process of explicating the normative structure of legit-
imacy, I have uncovered the features for two different kinds of
'story'. One is a story of developmental stages within each of
the components of legitimacy: from conventional to legal regu-
lation of power relations, especially in securing rights for the
subordinate; from the traditional to the popular source of author-
ity, and with it the expansion of those qualified to give consent
to include the whole adult population; from the ascriptive to the
meritocratic principle of access to the means and positions of
power; from the paternalist to the consultative determination of
subordinate or common interests; from the expressive to the
contractual mode of consent. These shifts under each aspect of
legitimacy do not coincide neatly with historical epochs, since
the shifts occur at different times, cut across one another, and
coexist in the same period within different societies. They cannot
therefore constitute a supposedly value-neutral or unequivocal
set of criteria for 'modernisation'. Yet in each case they represent
what can be seen as a developmental shift, in the sense that once
the later stage or level is reached in each case it is impossible
to regress to the earlier. It is this criterion of irreversibility that
entitles us to consider them as a hierarchy of higher and lower
The Normative Structure of Legitimacy 99

stages, not as the product of some unanchored value judgement.

This does not mean that legitimacy cannot be infringed or for-
feited once a higher level has been reached; only that it cannot
be restored on the basis of an earlier or lower principle. Working
out the implications of this 'story' is premature at the present
point of my enquiry, though it will concern me towards the end
of the next chapter, and in the second part of the book.
The second kind of 'story' concerns the way in which, within
any settled or established power relations, self-confirming pro-
cesses are at work to reproduce and consolidate their legitimacy.
This is because the legitimation of power takes place within
and not outside established power relationships - in given social
contexts, not in the sphere of abstract argument or ideal con-
ditions for consent. That is to say, the impressiveness of the
law, the evidence necessary for belief, the interests conducive
to consent are all structured by existing rules of power, not
independently of them. This self-confirming cycle is never perfect
or complete. But we need to analyse how it is constructed in the
first place before we can understand how and why it breaks
down where it does. No social-scientific account of the normative
structure of legitimacy can therefore be complete without an
account of how that legitimacy is socially constructed. Drawing
together the different strands of that account will be the task of
a brief concluding chapter to this part of the work.
4 The Social
Construction of
Legitimacy
Let me begin by recalling that the social scientist's primary
interest or purpose in the analysis of legitimacy is an explanatory
one. In the first instance this purpose is, through understanding
the distinctive rules, the justificatory principles and beliefs, and
the conventions about consent that underpin a given system of
power, to explain the behaviour of people within it, and the
kinds of relationship that it involves. Such an understanding also
provides the basis for a judgement about the degree of legitimacy
of a particular authority figure, power relationship or system of
power as a whole, when measured against its own criteria; this
judgement will help identify its potential points of vulnerability,
and explain any erosion of its ability to secure cooperation from
the subordinate when under pressure. In making such a judge-
ment the social scientist is not imposing extraneous or ahistorical
criteria, but employing those internal to the society or system of
power itself, against which it requires to be judged; he or she
is, as it were, reproducing the reasoning of people within that
society, and reconstructing the logic of their own judgements.
However, in doing so, the social scientist is also guided by a
general understanding of what legitimacy involves, as in the
threefold structure I have outlined, which provides an explora-
tory or heuristic framework for identifying the kinds of consider-
ation that are relevant. Such a framework is made possible by
the existence of recurrent features of power in all societies, and
a common structure of moral argument that is universal however
diverse or historically variable its actual content. It is only
because of these recurrent features and common structure that

100
The Social Construction of Legitimacy 101

it is possible for a person from one culture to understand to any

degree what is going on within a very different one.
The above account conforms closely to the anti-behaviouralist
programme for social science outlined by Peter Winch in his
book The Idea of a Social Science (1958), in which he character-
ises social activity as rule-governed activity, and the task of the
social scientist as being to recover the meaning and normative
force of rules for the agents involved (cf. Winch, 1964). Such a
programme corresponds closely to the analysis I have provided
of legitimate power as power that provides grounds for obedience
on the part of those subordinate to it, because of the normative
force that derives from rules, from justificatory principles, and
from actions expressing consent. However, if we were to stop
there, with the consequences of legitimacy for behaviour, we
should only have completed part of our explanatory task. We
also need to explain where legitimacy originates, how the prin-
ciples and beliefs that comprise it are maintained and repro-
duced, why consent is continually renewed, and what social
forces are responsible for the erosion of legitimacy whenever
such erosion takes place. In other words, we need an explanation
not only for the behaviour that is affected by legitimacy, but
also for legitimacy itself. And to accomplish this we need a
level of understanding that is not limited to that of the agents
themselves, however much that has to be our starting point.
When it comes to the explanation of legitimacy itself, we need
a level of understanding that goes beyond that of the agents
involved, and therefore transcends the limits of a Winchian pro-
gramme for social science (cf. Macintyre, 1970).
One way of approaching this explanatory task is a historical
one: to explore the origins of legitimacy in the origins of the
particular system of power in which we are interested. Such an
approach can only offer very limited returns, however. One
reason is that the origins of power relations tend to be so complex
that they are difficult to unravel; this is especially true of systems
of property or gender relations, stretching into the distant past
without any single moment of institution. Even if we consider
the political sphere and argue that constitutional rules which are
collectively agreed in a founding convention must carry much
greater legitimacy than those that are forcibly imposed from
above by conquest or coup d'etat, this is not decisive for the
102 The Legitimation of Power

legitimacy of a political system for all subsequent time; and the

less is it so, the longer the time that has elapsed since the
moment of institution. Rules of power that are imposed may
over time achieve acceptance (compare the history of the two
Germanies since the end of the second world war); those insti-
tuted by agreement may break down (see the constitutions
adopted at independence by the ex-colonial states during the
1950s and 1960s). It is history as a process of evolution rather
than as a determinate starting point that is usually important for
explaining legitimacy.
To underline this point, let us make the entirely imaginary
and fanciful supposition that all social rules - about property, the
division of labour, political authority- originated in a deliberate
agreement at some historical moment in the past. It could readily
be shown that such rules would quickly develop into something
quite different from that intended, by well known processes of
inequality-formation and power-accretion. Property rules specify-
ing an equal distribution of land between families produce
inequality over time, through differences of family size, of health
or strength, of soil fertility, and so on, to the point where the
only survival strategy for some is to hire out their labour to
others. Rules defining a non-hierarchical division of labour
between the sexes can readily produce a situation where males
come to control a resource that is necessary for subsistence or
security, and develop differential capacities and attitudes that
have consequences beyond those necessary to the original div-
ision of labour. Rules specifying the minimum necessary coordin-
ating and decisional powers for a political authority provide the
opportunity for acquiring and accumulating powers unanticipated
in the original agreement. None of this requires the assumption
of an innate striving for power in human nature; it can be
sufficiently explained by the respective strategies of groups
differently situated, acting under the pressure of changing
circumstances. Where the extremity of force and usurpation
tends to moderate over time as power becomes routinised and
reproduced within a rule-governed order, so original rules insti-
tuted by collective agreement, if unattended to, provide the basis
for the development of power relationships that are quite unin-
tended. Over time, therefore, an original difference between
usurpation and agreement tends to become blurred.
The Social Construction of Legitimacy 103

Much more important for the explanation of legitimacy, there-

fore, than the study of origins, is the study of power relations
and their legitimation as an ongoing process, whether in the
present or in some historical society with which the social scien-
tist is concerned. To say that the study of origins is less impor-
tant, however, is not to say that stories about origins may not
have a crucial part to play in legitimation (Hobsbawm and
Ranger, 1983). Consider this famous denunciation by Marx of
political economists' accounts of the origins of capitalist class
relations:

'In times long gone by there were two sorts of people: one,
the diligent, intelligent, and above all, frugal elite; the other,
lazy rascals, spending their substance, and more, in riotous
living ... Thus it came to pass that the former sort accumu-
lated wealth, and the latter sort had at last nothing to sell
except their own skins. And from this original sin dates the
poverty of the great majority that, despite all its labour, has
up to now nothing to sell but itself, and the wealth of the few
that increases constantly although they have long ceased to
work. Such insipid childishness is every day preached to us in
the defence of property . . . In actual history it is notorious
that conquest, enslavement, robbery, murder, briefly force,
play the great part. As a matter of fact, the methods of primi-
tive accumulation are anything but idyllic.' (1966, vol.l,
pp.667-8)

Whatever combination of force, fortune and effort may actually

have contributed to the early development of capitalist class
relations, it is certain that stories about such origins are impor-
tant; and therefore who tells them, or who controls their telling,
is of great consequence. This is why the content of history syllabi
is so contentious. But historical accounts are significant and con-
tentious precisely because of their relationship to the legitimacy
of power in the present, and because of their contribution to
disputes about it.
Once more, then, we come back to power relations as an
ongoing process, and to the central question of how legitimacy
is maintained and reproduced within given societies, not how it
may have 'originated', if it can ever be said to have done so.
104 The Legitimation of Power

Now the important point to stress about the maintenance and

reproduction of legitimacy- the maintenance of rules, the repro-
duction of beliefs, the continued expression of consent - is that
these do not take place independently of the structures of power
that they legitimate. There is simply no regular process of belief-
formation or consent-giving within historically placed societies
that exists independently of their power structures, or that can
therefore remain uninfluenced by them, though the degree of
this influence, or self-closure, will vary between societies and
between different power structures within them. It is because it
does not recognise any causal influence at all between power and
the process of its legitimation that the so-called 'consensus' model
of society falls down, not because it cannot offer any account of
power or social conflict in the first place. It is pre missed on the
unsociological assumption that 'consensus' is somehow estab-
lished and maintained by processes that are completely indepen-
dent of the existing relations of power within the society. Power
exists in one corner, as it were, and legitimacy is conferred,
and legitimating ideas developed, in quite another (cf. Giddens,
1968).
By what means, then, do ongoing power structures themselves
influence their own legitimacy, or condition their own processes
of legitimation? Two quite different accounts of this can be given.
The first account concentrates on the activity of the powerful in
influencing the beliefs of the subordinate, through their preferen-
tial access to the means of cultural development and the dissemi-
nation of ideas within society. In other words, among the powers
any dominant group possesses will be the ability to influence the
beliefs of others; and among the most important of such beliefs
will be those that relate to the justification of their own power.
The origin of such beliefs may be found in the first instance in
the need of the powerful for self-justification; but their privileged
access to the means of culture and ideological dissemination
ensures that their ideas become widespread throughout society,
whether as the result of conscious policy or not (Marx and
Engels, 1970, pt.l).
The above account is of course the familiar Marxist theory of
ideology, which Marxists employ in the context of class relations,
but other theorists have also used it in relation to political power
and gender. Thus the elite theory of Mosca, Pareto and their
The Social Construction of Legitimacy 105

followers within political science put considerable emphasis on

the 'political formulae' or 'myths' through which political elites
justify their rule to the non-elite, and which serve to consolidate
their power the more deeply they penetrate the consciousness of
the masses (Mosca, 1939, ch.3; Pareto, 1966, pt.l). An important
strand of feminist theory, also, argues in relation to gender that
the most basic elements in culture, including language itself, are
male-constructed, and that it is through their influence that
women come to accept a definition of themselves that is support-
ive of male power and the sexual division of labour on which it
is based (e.g. Spender, 1980). In their different spheres these
theories embrace the same explanation for legitimacy, and ident-
ify the same self-confirming feature of power, that is contained
in the Marxist concept of ideology: dominant groups are able to
secure their own legitimacy through their influence or control
over the processes whereby the beliefs of the subordinate are
shaped and reproduced.
Now it must be said that this account of how legitimacy is
reproduced is inadequate. In the first place, any explanation for
the beliefs held by subordinate groups which confines itself to
the processes of influence controlled by the powerful is open to
a number of objections. Not only do the powerful not influence,
let alone control, all the means of disseminating and reproducing
ideas in any society; but even if they did, this could not of itself
guarantee the acceptance of their ideas or justifications by others.
If we consider the twentieth-century examples of regimes that
have come nearest to a total control over the means of infor-
mation and ideological dissemination within their societies, we
do not find that this degree of control necessarily ensures the
acceptance of the messages that are disseminated. Indeed just
the opposite can occur: such regimes can be debilitated by a
general scepticism of official sources of information, which can
render them ineffective when they are needed most.
There are a number of related reasons for this effect. One is
that the processes of developing and disseminating ideas and
information require an independence from the powerful to secure
their authenticity. They have their own internal criteria or
standards necessary to effective practice, which can best be
guaranteed by a degree of autonomy from other social powers;
otherwise they are liable to degenerate. Secondly, as much con-
106 The Legitimation of Power

temporary research indicates, people are never merely the pass-

ive recipients of ideas or messages to which they are exposed
(e.g. McQuail, 1984). They are more like a sieve than a sponge.
That is to say, they tend to be selective, assessing ideas and
information in the light of their existing assumptions, and against
their lived experience. A third reason, underlying the other two,
is that the power of ideas, unlike other forms of power, cannot
be measured in terms of the means of power available to those
who control their dissemination, but rather in terms of their
credibility to the recipient. Any explanation for the ideas or
beliefs that people hold, therefore, must be based upon an
internal analysis of their plausibility or credibility to them in the
context in which they are situated, rather than simply on an
account of the means of their dissemination. If all the power of
the medium cannot ensure the credibility of the message, we
need to understand what makes some messages more credible
than others.
It is at this point that a different kind of explanatory account
becomes necessary, of the sort that I have advanced in the course
of the previous chapters. This is one which shows how a system
of power relations itself indirectly shapes the experiences, the
capacities, the expectations, the interests of subordinate groups,
through a variety of social processes, so that justifications for the
rules of power become credible because they are confirmed by
their own experience. Take, for example, the evidence of differ-
ential capacities that justifies assigning individuals to different
positions within a hierarchical division of labour or power struc-
ture. Such differences are themselves in part the product of
expectations about future roles that people will occupy, and
are thus constructed indirectly by the power rules, though the
immediate agents of their formation may be found among the
subordinate themselves (mothers, working-class parents, peer
groups and so on).
Or consider the evidence which confirms the claim that the
power of dominant groups serves a general interest, including
the interests of the subordinate. This evidence is provided by
the fact that their interests, for economic sufficiency or physical
security, cannot indeed be satisfied outside the given power
relationship, where necessary resources or capacities are con-
trolled by the powerful. It is also provided by the fact that the
The Social Construction of Legitimacy 107

system of power serves to shape some of the desires that it is

organised to satisfy, as a gendered division of labour both shapes
and satisfies the respective desires of women and men with regard
to the private and public spheres, or to caring and non-caring
roles; or as the late capitalist system of production generates
many of the consumer wants that it proceeds to satisfy, or more
generally, defines the satisfaction of consumer wants as the most
fundamental need.
In all these different ways the justifications advanced for a
given system of power are vindicated by effects generated by the
power system itself, but which are not understood as its effects,
because they appear autonomous or independent of it. As Marx
himself understood well, though not all later Marxists have fol-
lowed him, or worked out the implications for other dimensions
such as gender, it is the appearance of the socially constructed
as natural that lies at the heart of all ideology. What is socially
constructed is not itself imaginary or illusory, and its evidence
gives credibility to the justifications advanced for a given system
of power. Yet the fact that it is constructed indirectly by that
same system of power is obscured by the complexity of the
processes involved; and by the fact that these processes, such as
those of socialisation, are not necessarily managed by the power-
ful, but often by the subordinate themselves. It is, for example,
within women's historically limited sphere of power, over
children in the home, that the differentiation of gender has been
primarily reproduced. Both the sphere of power itself, and the
evident facts of its limitation, are of course necessary to this
process.
One reason, then, why a simple dominant ideology theory
cannot provide an adequate explanation for legitimacy is that,
on its own, an account of the beliefs of subordinate groups as
the product of direct influence by the powerful cannot explain
their credibility to them in the context in which they are situated.
For that we need a different, albeit complementary account, in
terms of the social effects produced indirectly by the system of
power, which provide evidence necessary for its justification. A
further important reason is that power relations maintain and
reproduce their legitimacy through their effects on people's
actions in expressing consent, not only on the beliefs that are
relevant to their justification. Here we can observe a tendency
108 The Legitimation of Power

among some neo-Marxists and others who adopt a 'dominant

ideology' approach simply to reproduce the Weberian definition
of legitimacy as the belief in legitimacy, because it fits so well
with their own presuppositions. This Marxisation of Weber, if it
can be so termed, by adding the assumption that the powerful
directly shape the beliefs of the subordinate to the Weberian
definition of legitimacy as belief, produces an appealing, though
doubly erroneous, combination.
With regard to consent, it will be sufficient here to recall the
account given in the previous chapter of the way in which rules
or structures of power themselves set the context and define the
terms under which consent to a position of subordination is
given, so that the decision to give consent is in practice a highly
constrained one. Again, because the constraint is not the result
of personal intervention by the powerful, but the impersonal
consequence of the rules or structures, it appears as naturally
given or determined, and the consent to which it gives rise
as therefore entirely voluntary. And because it seems entirely
voluntary, it helps to legitimate the rules that condition it in the
first place.
Altogether, then, we see a number of processes at work
whereby an established structure of power helps maintain and
reproduce its own legitimacy, though this self-confirming circle
operates in different ways in different types of power relation,
and is more closed in some than in others. Alongside the influ-
ence of the dominant over the ideas of the subordinate, which
cannot be denied, though it is often exaggerated and on its own
as an explanation becomes erroneous, there are the indirect
effects that a system of power produces: to confirm the differen-
tiation between dominant and subordinate which justifies their
respective positions; to structure the common interest so that it
can only be met through satisfying the purposes of the powerful;
to help shape the desires of the subordinate in directions that
the system is capable of meeting; to limit the choices available
so that consent to a position of subordination, although con-
strained, is also at the same time voluntary. Together these
processes, relevant to the different dimensions of legitimacy that
I have distinguished, serve to maintain and reproduce the legit-
imacy of an established system of power.
If, therefore, there are these processes at work, confirming
The Social Construction of Legitimacy 109

the legitimacy of established power relations, how does it come

to be eroded? I can offer no more than the merest outline here,
since the subject will be taken up in the second part of the book.
In the light of the account of legitimacy I have given, we could
distinguish two different kinds of erosion. One is when a system
of power becomes chronically unable to meet the interests of the
subordinate, whether those that it defines itself, or those that
the subordinate have come to define independently. This may
happen through the impact of external forces, or through blocka-
ges that are generated internally within the system of power at
a particular stage of its development. A second type of erosion
takes place when the system cannot satisfy some basic principle
of differentiation or source of authority on which it rests; or else
the beliefs that underpin its principle of differentiation or source
of authority themselves decay. The latter will happen when social
changes taking place within the society, or the evidence available
from other societies, reveals that what had previously been
assumed to be a 'natural' form of social organisation, or one
based upon 'natural' differences, is in fact socially constructed.
Finally, both types of legitimacy erosion can occur together, and
often do so.
When such an erosion takes place it necessarily has conse-
quences for the attitudes and behaviour of the subordinate.
Through their everyday experience as subordinate they are
already involved in the negative aspects of power relations - the
difference of fates, the transfer of resources and opportunities,
the limitations of freedom - though these may have become
obscured and redefined by the process of legitimation. Once
shorn of their justification, however, these negative aspects are
starkly exposed, and experienced for what they are.
On its own this 'negative' awareness may produce a condition
of frustration, impotence or resignation on the part of the sub-
ordinate, and a general unwillingness in fulfilling the require-
ments of their position. For it to have a transforming potential,
however, something else is needed: the possibility of communi-
cation with others and an autonomous space relatively protected
from the influence of the powerful within which to do so; and
the imagination to conceive of a different set of rules and
relations for the fulfilment of basic social needs from the existing
ones. In other words, the subordinate have to acquire an insti-
110 The Legitimation of Power

tutional facility (formal or informal means of communication,

movement, organisation) that is independent of, and a level of
consciousness or conceptual position that transcends, the estab-
lished power relations, if they are to develop the impetus to
transform them. What we are talking of is the possibility of a
legitimacy, not that is entirely asocial or ahistorical, but that is
constructed independently of existing relations of power. And
when such a transformatory consciousness impels the subordinate
to action, to the active withdrawal of consent, to the delegitimat-
ion of power in the sense I have defined it, then the authority
system enters a period of crisis, which may be resolved by
reform, repression or revolution according to the circumstances
and the relative balance of forces.
Let me draw this stage of the enquiry to a conclusion, and
pose a final question to be answered. I have said that the purpose
of social science is primarily an explanatory one. With regard to
the subject of legitimacy I have distinguished two different levels
of explanation. The first is the explanation of behaviour within
power relations; and I have shown how legitimacy affects this
according to its form and degree, its presence or absence, across
the different dimensions that I have identified. The second level
is the explanation of legitimacy itself; and I have shown how it
is maintained and reproduced within established systems of
power, and what are the typical conditions for its erosion and/or
transformation. These two levels of explanatory question clearly
differ from one another, and it is important not to confuse them.
There is a third level of explanatory question that I wish now
to confront, of a more epistemological kind: how is the social
scientist able to answer these two different kinds of question
about legitimacy? What are the conditions necessary for answer-
ing them? Whereas answering the first question requires adopting
a position internal to the given society or power relationship, so
as to make intelligible the norms, conventions and legitimating
principles that govern it, and to reconstruct the reasoning of
the agents involved; answering the second question necessarily
requires adopting a position outside it. It is only from a stand-
point outside given power relations that it is possible to under-
stand the processes whereby their legitimacy is maintained and
reproduced, and what are the forces at work eroding it, where
such erosion is taking place.
The Social Construction of Legitimacy 111

Is not this standpoint precisely the same as that I have just

described subordinate groups adopting, who aspire to an alterna-
tive legitimacy, unconstrained by existing power relations, in a
situation where the latter have entered a condition of crisis? Is
there not a congruence, therefore, between the explanatory pur-
pose of the social scientist and the emancipatory purpose of
alternative movements of the subordinate, not in the sense that
the social scientist must share the particular ideals of the latter,
but that their epistemological position is the same? Do not they
both have to be able to conceive of a legitimacy that is uncon-
strained by established power relations, the one in order to
explain the normal processes at work reproducing legitimacy, the
other in order to be able to transcend and transform them? Is
the social scientist thereby committed, not by individual political
choice or ideological predisposition, but by the very conditions
necessary for knowledge, to share the standpoint and therefore
the emancipatory concerns of the subordinate? Moreover, is this
standpoint not also that of the normative philosopher, who
imagines (and invites us to imagine) a discussion between equals
in a pre-social 'state of nature', in an 'ideal speech situation', or
behind a 'veil of ignorance' about our position in any given
power structure, because in such a situation alone is it possible
to establish philosophically valid principles of legitimacy, or
arrive at an agreement about power rules that is truly voluntary
because it is unconstrained by them? Have I not begun this work
by defining the projects of the social scientist and the normative
philosopher as fundamentally different, only to discover that
they are, if not precisely the same, then at least congruent or
complementary, because their epistemological stance has to be
identical?
These are large and fundamental questions, and the reader
must wait till the end of the book for a fully considered answer
to them. There is, however, one social scientific standpoint that
produces the answer 'no' to all these questions, and which I
must address here because doing so leads to a conclusion that is
necessary for the second stage of my enquiry. This is the stand-
point of the knowledgeable sceptic, who, from a position above
all power relations, sees how they reproduce the conditions for
their own legitimation, and concludes that the subordinate can
never break out of this self-confirming cycle of power: either
112 The Legitimation of Power

because power structures will always consolidate themselves in

ways that escape their control, and will always succeed in justify-
ing themselves through mystificatory processes; or because the
conditions of discourse can never be equal, and those who con-
struct or control its terms will always be able to legitimate their
position, or conceal that it is a position of power at all. From
such a standpoint the activity of an emancipatory movement
must be judged Utopian and the project of the normative philos-
opher as fanciful. They may help to bring about social or political
change, to be sure, but not change that is progressive; either
because such change in power structures is an empirical impossi-
bility, or because there is no standpoint from which any change
could conceivably be judged as progressive at all. For the social
scientist to share the standpoint of the emancipatory movement
or the normative philosopher, therefore, would be merely to
indulge in illusion.
Such a position may sound an extreme one, though it is in
effect presupposed by the practice of a 'value-free' social science,
which defines its task as being to explore the differences, say,
between the legitimating principles of different societies or politi-
cal systems, but never to make a judgement between them. I
want to show that this position is mistaken, and mistaken in a
way that has crucial implications for the explanatory task of the
social scientist. It is mistaken because the different legitimating
principles from different historical societies I exemplified in
Chapter 3 are not simply different from one another, but qualitat-
ively different, and in a particularly significant way. They differ
in the extent to which they in principle encourage or discourage
the self-confirming processes analysed whereby power rules main-
tain the conditions for their own legitimacy by reproducing differ-
ential attitudes, characteristics and expectations that confirm
existing arrangements of power, and by constructing the circum-
stances from which consent to a position of subordination is
given. In other words, some legitimating principles have to be
judged as, not just different from, but more emancipatory than,
others.
Of the legitimating principles I have mentioned two possess
this qualitative difference. One is the meritocratic principle of
differentiation, specifying that positions and means of power
should be open to all to acquire, on the basis of an equality of
The Social Construction of Legitimacy 113

opportunity to attain them. However limited the operation of the

principle may be in practice by the social advantages transmitted
through the institution of the family, it postulates the ideal that
people's qualities, attributes, capacities and opportunities should
precisely not be defined or limited in advance because of some
determinate role they are bound or expected by birth to occupy.
It therefore challenges the mechanisms whereby established rules
of power reproduce the differences that serve to justify them;
and it provides the means for eroding the processes by which
the socially constructed appears as 'natural', because it embodies
the recognition of the distinction between the two (natural and
socially determined differences) as its central assumption.
The second principle that has the qualitative difference I have
mentioned is the principle of popular sovereignty, which lot:ates
the ultimate source of authority for political power, and therefore
for all power rules, in the people. This principle - again despite
all the limitations, qualifications and distortions to which it is in
practice subject - postulates an ideal in which all power rules
are open to revision by public debate and decision between equal
citizens. The institution of the democratic legislative assembly of
all citizens is the closest approximation possible in historical
practice to the 'ideal speech' situation or pre-social 'state of
nature' envisaged by political philosophy, since it embodies the
requirement that consent to power rules be given directly at the
point of their institution or revision, and from a position of
equality, rather than indirectly by consent to a particular power
relationship from a condition of inequality. As a principle, there-
fore, the idea of popular sovereignty embodies within itself the
potential for breaking the self-confirming cycle whereby consent
to the rules of power is conditioned by circumstances created by
the rules themselves, and definitions of the general interest are
structured by the interests of the powerful.
Together, then, the principle of equal opportunity to acquire
positions or means of power, combined with the equal right to
take part in determining the rules that define what the necessary
positions and inequalities in the general interest should be, and
what should be the limits of power associated with them - these
two principles allow the possibility of a legitimacy that is not
constructed or reproduced by the established rules of power,
through their conditioning effects on the beliefs and consent
114 The Legitimation of Power

necessary for legitimation. They could therefore be termed 'trans-

parent' or 'authentic' principles, in that they do not depend for
their acceptability upon social processes appearing other than
they are: the socially constructed appearing as 'natural'; consen-
sus appearing as spontaneous and consent as voluntary, when
both are conditioned by the rules of power.
This 'transparence' or 'authenticity' of the principles men-
tioned explains two further characteristics about them. The first
is that they can act as critical, and not only as apologetic or
'legitimating' principles; that is to say, however inadequately
they are realised in practice, they embody within themselves the
capacity to generate criticism of these inadequacies. The second
characteristic that is explained by their 'transparence', as I have
called it, is their irreversibility: once we have attained a form of
legitimacy that is not conditioned by power rules, it is inconceiv-
able that we could regress to one that was.
Have I now abandoned my social scientist's concern with
explanation for the normative philosopher's interest in an ideal
legitimacy? If so, it is only because, unless we do so, certain
important features of the contemporary world cannot be
adequately explained: for example, why ascriptive or non-demo-
cratic principles of legitimacy cannot be indefinitely sustained in
a world where they are exposed to meritocratic or democratic
ones; and why, when the legitimacy of democratic political sys-
tems is eroded, it is for very different reasons and with very
different consequences than those based on traditional, theocratic
or 'scientific' principles of authority, which are vulnerable to an
erosion of the central belief system on which they are based, as
a source of ultimate authority for power rules, in a way that
democratic political systems are not. To work out the impli-
cations of this difference will be part of my purpose in the second
half of the book.
 PART II
LEGITIMACY IN THE
CONTEMPORARY
STATE
5 Dimensions of State
Legitimacy
In view of the argument advanced in the first part of the book
about how rules of power come to be socially reproduced in a
way that reinforces their legitimacy, it is remarkable how com-
paratively insecure is the legitimacy of many states in the contem-
porary world. Legitimacy, it seems, is as much the exception as
the rule for contemporary states. Many are subject to military
dictatorship, originating in a breach of the constitutional rules
whose lawlessness is extended into the subsequent practice of
government. In others the political order finds only weak support
in popular beliefs and values, or there is widespread disagree-
ment about fundamental aspects of it. In yet others there is only
limited legitimation through consent. Those who live in countries
whose political legitimacy is secure are likely to take it for
granted. On a global scale, however, what is striking is the
difficulty that contemporary states experience in achieving such
a legitimacy, and their rulers in governing in a manner that
maintains it.
Why should this be so? Why should political legitimacy be so
difficult to attain in the contemporary world? This is the central
question which I shall address and seek to answer in the chapters
that follow. As I shall show, the answer is to be found at a
number of different levels: in the distinctive character and
dilemma of the modern state as a mode of concentrating and
organising power; in the impact upon it of a variety of external
forces which it is unable to control; finally, in the impossibility
of reversing the principle of popular sovereignty as the ultimate
source of political authority in the modern world, while at the
same time the institutional forms for giving effect to it prove so
problematic and difficult to sustain.

117
118 The Legitimation of Power

The interrelationship between these different levels provides a

thematic continuity underlying the diversity and detail of the
following chapters.
This fifth chapter explores the distinctiveness of the state in
comparison with the other power-structures considered in the
first half of the book; it identifies the main criteria that have to
be met if the contemporary state is to achieve legitimacy; and it
outlines in a preliminary way some of the key problems that
confront it in doing so. In Chapter 6 I examine the different
principles of legitimacy embodied in different types of political
system and their respective institutions. I argue that each can be
seen as offering different solutions to the common legitimation
problems already defined, and each as confronting its own dis-
tinctive problems of legitimacy in turn. Chapter 7 distinguishes
the different ways in which legitimacy breaks down, and the
divergent dynamics of revolution and coup d'etat respectively.
Its conclusion draws together the different strands for an expla-
nation of why such breakdown is so common in the contemporary
state.
Before proceeding with this agenda I should reaffirm my view
that the contemporary state is a form of power-structure that
requires legitimation, not so much to function, or even to survive
over a period of time, but to achieve those purposes that depend
upon the support of its population, and to maintain its political
system intact in the face of serious policy failure or challenge to
it. It is superficially attractive to argue that the only thing the
state requires from its subjects, their taxes, are mostly collected
via non-state intermediaries, and in the course of transactions
that have no overtly political significance; and that, provided the
loyalty of the administrative and coercive apparatus is secure,
the absence of any normative commitment or sense of obligation
on the part of the wider population is irrelevant to the effective
functioning of the state. Such a view rests on the mistaken
assumption that the state apparatus is entirely self-contained,
and can be immunised from the attitudes and actions of the
surrounding population, an assumption that will not withstand
examination in the light of the political history of the last two
decades.
What such a history demonstrates is that the subjection of
regimes weak in legitimacy to stress - the failure of some policy
Dimensions of State Legitimacy 119

they have defined as important, the imposition of some special

hardship upon the population, the experience of national humili-
ation, the death or disability of the ruler, which in more legit-
imate systems would be resolved by a routine change of govern-
ment or its personnel - develops into a crisis of the regime itself.
This happens because of the low tolerance level of the popu-
lation, and because manifestations of public opposition to a par-
ticular policy rapidly develop into opposition to the system of
government as such, and to its authority. Such opposition,
whether by key elites or popular masses, is typically replicated
in divisions within the military and administrative apparatus
itself, and causes serious dislocation in its capacity to rule. The
spectacular collapse witnessed over the past two decades of
regimes whose control over a powerful military and administrat-
ive apparatus looked secure- in Greece (1974), Portugal (1974),
Spain (1976), Iran (1979), Argentina (1982), the Philippines
(1986) - demonstrates the importance of legitimacy to the effec-
tiveness and continuity of a system of rule. However much the
institutions of the state may appear to be independent and self-
contained, therefore, their operation, as with all systems of
power, involves an interactive process between the dominant and
the subordinate, in which the responses of the latter exert their
own effects upon the character and effectiveness of rule.
All the examples mentioned above were varieties of personal
dictatorship, in which the failure to establish an accepted consti-
tutional basis for rule left open the possibility of a humiliating
collapse in the capacity to govern as its eventual termination. A
parallel fate overtook the communist regimes of Eastern Europe
in 1989, which had only been secured against an autonomous
evolution towards a different political order by Soviet inter-
vention or the threat of it, as in Hungary (1956), Czechoslovakia
(1968) and Poland (1981). The removal of this threat exposed
the regimes to the full consequences of their own lack of legit-
imacy, and to their ultimate disintegration in the revolutions of
1989.
In other kinds of political system, a weakness of legitimacy
will tend to produce a different kind of outcome. In a parliamen-
tary or presidential democracy, the commitment to the rules of
electoral competition may be too weak to withstand the loss of
office where the stakes are high, and result in a coup d'etat,
120 The Legitimation of Power

either on behalf of the losers, or to resolve the resulting social

conflict and disorder. In some cases a coup may be preemptive,
to prevent an expected loss of power. Examples from recent
history are too numerous to list, including most countries in
Latin America, many in Africa and Asia, and some in Europe
(Greece, 1967, Turkey, 1980). In all types of political system,
finally, where there exists basic disagreement about which state
a given population or section of it should belong to, or about
the nature of the regime type itself, the prospect of civil war
looms as a possible outcome, as currently in Ethiopia, Sudan,
Angola, Sri Lanka, Northern Ireland, and elsewhere.
The collapse of government, coup d'etat, invasion, civil war -
these are some of the characteristic outcomes, probable rather
than inevitable, of the dynamic processes of interaction between
governments and their subjects where legitimacy is in some
respect weak or absent. Such outcomes can hardly be judged
insignificant for anyone involved, governments and subjects alike.
However, we should avoid drawing the conclusion, as some pol-
itical scientists have done, that because these outcomes are not
only significant but usually also unpleasant, involving a threat to
human life, that stability must therefore be the supreme virtue
of political systems. A period of social conflict and political
upheaval may be a necessary, if costly, price to be paid for the
transition to a more just or more progressive political order.
Although we are likely to disagree about such judgements, we
can at least agree that the outcomes described above are signifi-
cant ones; and it is because of its relation to them that legitimacy
is so central an issue for political science.
Any adequate explanation for the erosion of legitimacy in
particular countries requires an analysis of the specific dynamic
of different political systems, of their respective principles and
modes of legitimation, and of their distinctive crisis tendencies.
As a preliminary to this comparative study of political systems
in Chapter 6, I shall explore here the main legitimation problems
confronting the contemporary state as such, and shall do so in
terms of the threefold criteria of legitimacy already established:
those of legal validity, rule justifiability and expressed consent.
An examination of each of these criteria in turn will serve to
define where the recurrent legitimation problems of the contem-
porary state in general are to be found.
Dimensions of State Legitimacy 121

The elusiveness of legality

If the first condition of legitimacy is that power should be

acquired and exercised according to established rules, then our
starting point must be to explain why breaches of the consti-
tutional rules by those who wield state power, or aspire to do
so, occur so comparatively frequently, and so often with
impunity. Most serious of these breaches are those involving the
acquisition of power, since they must call into question the legal
validity of a government in all it does. By comparison, infringe-
ments of the law in the exercise of power only affect one aspect
of a government's performance, not its whole basis. Even these,
however, if sufficiently serious or chronic, will erode its legit-
imacy to the extent that they are publicly known.
Why, then, is it so difficult to ensure that legality is respecteu
in the acquisition and exercise of state power? Any complete
explanation will have to explore the diverse circumstances under
which constitutional rules come to lack sufficient authority, and
the various pressures which lead power-holders or power-seekers
to break them. In this section, however, I shall confine the
discussion to identifying those problems that derive from the
nature of the state itself, and its unique legal position, since
these constitute the most immediate source of difficulty.
According to conventional definition, the state comprises those
specialised institutions that exercise a monopoly of law-making
and adjudication over a given territory, and of the organised
physical coercion necessary to enforce it. The state is the
supreme law-making and enforcing agency for society. Even this
minimalist definition suggests certain legitimation problems that
distinguish the political from the other dimensions of power. To
begin with, the state is responsible for determining the rules
which govern all other power relations in society, and for legitim-
ating these relations by according them legal validity. Whatever
the limits in practice to this power, its possession makes the state
the site of intense struggle to control it, or to influence those
who do. The struggle is all the more intense in that, unlike all
other power relations, which are dispersed throughout society in
families, units of production, etc., the rule-making power of the
state is typically concentrated in a central set of institutions, and
attempts to influence or control it are focused on a single point.
122 The Legitimation of Power

On the one hand, then, the state's responsibility for determin-

ing and legitimising the rules governing the other power relations
in society makes it the site of intense struggle. On the other
hand there is no law-making power higher than the state, to
which it is subject, and whose authority can validate and guaran-
tee its own rules of power. This is the essence of what is termed
sovereignty: that the state determines its own rules, as well as
those of all other powers within its territory. In the legal sense
it is self-validating. Sovereignty thus represents at once a con-
dition of power and of vulnerability: of power, because the state
is independent of any higher authority in the legal control over
its own domain; of vulnerability, because there is no superior
legal authority to which it can appeal to confirm its own legit-
imacy, and to enforce its own rules in face of an intense struggle
within society to gain control of and exercise that sovereign
power.
The above statement about there being no higher authority
than the state needs qualifying in two ways. First, the state is
part of a system of states, whose mutual recognition of sover-
eignty confirms the right of each to exercise a monopoly of law-
making power within its own territory. Such recognition by the
international system of states may be particularly important to a
new regime, especially one which is facing internal opposition.
Recognition by an established authority itself confers authority.
However, such recognition specifically does not involve the legal
validation of any particular political arrangements or regime
type, since this would constitute an infringement of sovereignty.
In so far as there is a recognised body of international law, this
is concerned primarily to regulate the relations between states,
rather than the conditions within them; and it suffers from the
well-known limitation that it possesses no effective means of
enforcement. At most, therefore, the international 'system' con-
firms the legitimacy of each state's sovereignty over its own
territory; in doing so, it explicitly acknowledges that there is no
higher rule-making authority from which the validity of the state's
own rules of power can be derived (Tilly, 1975; Poggi, 1978).
My second qualification is that sovereignty as the exclusive
right to law-making and law-enforcement within a given territory
is not to be equated with power tout court, or the state's ability
to achieve its various purposes. States which are economically
Dimensions of State Legitimacy 123

or militarily weak, or geopolitically vulnerable, may be subject

to all kinds of external interference, whether at the hands of
other states, of international agencies, or of other non-state
powers. Where such interventions are manifest and extensive
they must damage the standing of the regime that is compelled
to submit to them, even if the formality of sovereignty is pre-
served, since they infringe the recognised principle that states
should be self-determining in their internal affairs. It is partly to
compensate for such weakness that states join mutual-security
blocs, or regional schemes of economic cooperation, in which
they surrender an element of formal sovereignty in return for
the prospect of enhanced power in the international environment
(Kolinsky, 1981; Held, 1989, ch.8).
With these qualifications, then, about the meaning and limits
of sovereignty, we can return to the fundamental problem it
poses alike in all states: if there is no higher rule-making power
to validate and enforce the state's own rules, what is there to
ensure that in the struggle to acquire and exercise its sovereign
power, the rules will be observed? The question takes on a
particular urgency in the light of the fact that the means all states
employ to enforce the law - the institutions of organised physical
coercion - are themselves uniquely placed to break the law with
impunity, since there is no superior coercive power to which they
are subject. What is there, then, to ensure that the state's own
rules will be observed; that constitutional legality and the rule
of law, the first requirement of legitimacy, will be respected?
Part, but only part, of any answer involves the internal organis-
ation of the state institutions themselves. The maintenance of
constitutional legality depends upon two institutional precon-
ditions. The first is the effective independence of the judiciary
from the legislative and executive branches of the state, so that
the latter can both be rendered subject to the law. In other
words, the absence of any higher legal authority than the state,
to which state power is subject, can be remedied by a separation
of powers whereby one part of the state is made accountable to
another for its respect of the law.
However, such a condition is easier to specify in theory than
it is to realise in practice. It depends not only upon a formal
institutional independence for the judiciary, but upon the devel-
opment of an ethos of legal impartiality, which takes no account
124 The Legitimation of Power

of the status of the individual concerned, and which is capable

of withstanding the pressures and blandishments of the powerful,
even, on occasion, at personal risk to the judiciary themselves.
It also requires an independent press and other media, to ensure
that breaches of the law by state personnel, especially its law-
enforcement agencies, are brought to public attention. There is
a paradox here, however, in that the ability to conceal illegality
from the population at large will prevent any damage to a
regime's standing, since what is not publicly known cannot have
public consequences. Yet where violations of legality are wide-
spread, the cumulative experience of them will have a corrosive
effect over the longer term.
The second institutional prerequisite for constitutional legality
is the effective subordination of the military to civilian control.
Again, this is a condition much easier stated than realised. Two
different ways have been developed for realising it. One, charac-
teristic of single-party, and especially of communist, regimes, is
a thoroughly politicised mode, in which the military are system-
atically inculcated with the outlook and values of the regime,
and have a special role in the attainment, as well as the defence,
of its goals. While this model, with its integration of all state
personnel into a cohesive elite through the institution of the
party, cannot deliver a sufficient separation of powers to guaran-
tee the rule of law for its citizens, it has historically secured
the effective subordination of the military to civilian leadership
(Perlmutter and Leogrande, 1982).
A second mode is a thoroughly de-politicised one, character-
istic of liberal systems, in which the army is trained in values
specific to the military profession alone, and is expected to give
loyal service to the government of the day, whatever its political
complexion and whatever policies it happens to pursue, since
'politics is not the soldier's concern'. This model is fine in theory,
but it presupposes that the military, however professionalised,
can be completely isolated from the pressures and conflicts of
the society around them, and from every inadequacy in the
government or the constitutional order which they are supposed
to defend. It is precisely here that the model tends to break
down in practice (Nordlinger, 1977, pp.12-19).
To recognise this practical limitation is to acknowledge the
inadequacy of any purely legal or institutional approach to the
Dimensions of State Legitimacy 125

'rule of law'. At the end of the day, however thorough the

arrangements made to ensure the defence of a constitutional
order, it is only as secure as the degree of support it can com-
mand from society at large; and that, as with any system of law,
is a matter of its substantive content and justifiability, as much
as the formal procedures for its implementation. Here lies the
inadequacy of accounts of legitimacy in the tradition of legal
positivism, such as Weber's idea of 'rational-legal' authority,
which assume that legality is not merely a necessary, but also a
sufficient, condition of legitimacy; and that it can therefore be
guaranteed by bureaucratic norms of rule conformity alone. On
this view, if constitutional rules are infringed, the cause must lie
in the inadequate socialisation of the relevant state personnel in
procedural norms, or in some defect in the legal procedures
themselves.
Such a theory takes no account of the content of the consti-
tutional rules, or of the principles and beliefs that underpin them.
Major breaches of a constitutional order in the acquisition or
transfer of power only take place because that order is itself a
subject of disagreement or dispute, or finds only weak support
from society or a significant section of it. Institutional procedures
to protect legality may deal with minor challenges, and may
postpone the point of breakdown in the event of major ones,
but cannot by themselves maintain the integrity of a consti-
tutional system when people have lost confidence in it. For this
reason, the Weberian concept of 'rational-legal' authority, or
procedural correctness in the creation and application of legal
rules, may effectively characterise the distinctive mode and
temper of modern officialdom in contrast to traditional types of
administration, but it cannot provide us with a sufficient criterion
or account of political legitimacy in the modern world. For that
we need some understanding of the principles and beliefs that
give the rules their justification (Habermas, 1984, pp.261-2).
Indeed, it is precisely because of the characteristic dilemma of
political sovereignty - with its discrepancy between the power it
confers on the one hand, and the absence of a higher authority
to validate and enforce its own rules on the other - that the
constitutional order needs a firm anchoring within society:
through its justifiability in terms of established beliefs and values,
and through the evidence of expressed consent on the part of
126 The Legitimation of Power

those subordinate to it. It is the very vulnerability of sovereignty

that makes these broader dimensions of legitimacy so crucial for
the contemporary state.
In conclusion, then, we can say that the first line of defence
for state legitimacy is the set of institutional arrangements and
practices designed to protect the 'rule of law' in the access to,
and exercise of, state power. Without such arrangements, and
the commitment of the relevant personnel to uphold them, there
is little likelihood that the first criterion of legitimacy - legal
validity- will be met. However, these arrangements, although a
necessary, are not in themselves a sufficient, condition of legit-
imacy. For that, the constitutional rules must also conform to
principles acknowledged as valid within the society. What these
principles must be, will be the subject of the following section.

Constitutional rules and their justification

It follows from what has been said above that, if, in assessing
the legitimacy of a particular government, our first consideration
must be its legal validity, the second must be the justifiability of
the constitutional order from which its power derives. Dis-
cussions of this second dimension in the literature of political
science have great difficulty in bridging the gap between the
underlying principles, or legitimating criteria, of a constitutional
system, on the one hand, and the practical, institutional arrange-
ments for their realisation, which may work more or less well,
on the other. It is essential to bridge this gap if the theory of
legitimacy is not to remain suspended between a wholly abstract
discussion of principles and a pragmatic analysis of institutions,
which are never able to make contact with one another. My aim
is to show how this can be done.
In Chapter 3 I identified a number of criteria that have to be
met if a system of power is to be legitimate. Of these, two are
crucial for a constitutional order. First is an authoritative source
for the supreme power or powers. Constitutional arrangements
must conform to established beliefs about the proper source from
which power should be derived. In doing so they will necessarily
embody procedures or rules that reflect those beliefs, such as
the principle of heredity in a traditional monarchy, of popular
Dimensions of State Legitimacy 127

election in a representative democracy, of clerical prerogative in

a theocracy, and so on. The fink between institutional arrange-
ments and legitimating criteria here is, simply, that institutions
embody accepted principles, and their rules reflect established
beliefs, about the rightful source of political authority (or, con-
versely, fail or cease to do so). In the section devoted to the
source of political authority that follows, I shall discuss the insti-
tutional consequences of the crucial change from belief in her-
edity to the principle of popular sovereignty marked by the
French revolution; consider what other principles are consistent
with it; and trace the relation of popular sovereignty to an idea
which is central to political legitimacy in the contemporary world
- that state institutions should express a national identity and
serve distinctively national purposes.
The idea of a 'national' purpose already suggests the second
basic requirement that a constitutional order must meet if it is
to be legitimate: it must facilitate rather than hinder the pursuit
of a general interest, particularly in respect of those purposes that
the state is expected to fulfil. Although meeting this general-
interest criterion might seem to be a matter for individual govern-
ments, on the one hand, or depend upon the social structure,
on the other, constitutional arrangements form the crucial inter-
mediary between the two. Thus they may serve to intensify or
moderate existing social divisions; they may ease or obstruct the
replacement of administrations that have failed in some necessary
purpose. And, for whatever reason, persistent failure on the part
of governments, or the pursuit of a divisive sectional advantage,
will bring discredit upon the constitutional arrangements them-
selves as incapable of guaranteeing the general interest. In the
second section, therefore, I shall consider some of the recurrent
problems contemporary governments face in satisfying the basic
security and welfare purposes of the state, and examine how
constitutional arrangements can resolve or exacerbate problems
of failure and social division respectively. I shall argue that con-
stitutional rules which satisfactorily embody an accepted source
of authority may be less effective at satisfying a general-interest
requirement, and vice-versa; in other words, the two main
requirements of legitimacy may in practice conflict with each
other.
128 The Legitimation of Power

i) an authoritative source of power

The decisive shift in legitimating ideas that marks the modern

world was the erosion of belief in the supreme value of birth and
historical succession, in the idea that the qualities appropriate to
the exercise of power are conferred by the status of one's family
and its pedigree. This belief underpinned the rules of property
and position throughout so-called 'traditional' societies, as well
as the rules of political office; its erosion had correspondingly
wide-ranging consequences, extending from the political to the
economic and social spheres. As the collapse of the ancien regime
in France at the end of the eighteenth century demonstrated, the
political and social revolutions were inextricably linked, because
power and position in all spheres had been based upon a common
set of beliefs about the primacy of birth and lineage.
However, in the political domain the basic ideological shift
involved more than an affirmation of the new principles of equal-
ity before the law, and the career open to talent (at least, to
male talent). If that were all, then the leading political roles in
the state could simply have been opened up for competition like
any other full-time post, and the most 'qualified' be appointed
by an appropriate committee, no doubt on submission of suitable
character references. For political legitimacy another shift proved
even more crucial: from regarding the state as the personal prop-
erty of the ruler, to seeing it as belonging to its people, and
political authority as deriving therefore from the people, and not
from the superior status and pedigree of the ruling family. So
the French Declaration of the Rights of Man, after enunciating
the principles of civil equality and the protection of natural
rights, declared that the 'nation is essentially the source of all
sovereignty' (1789 version) and 'sovereignty resides in the people'
(1793 version) (Stewart, 1951, pp.l14, 457). The universality of
the idea today was acknowledged in the United Nations Declar-
ation of Human Rights that 'the will of the people shall be
the basis of the authority of government' (Article 21). In the
contemporary world it is virtually impossible for a political
system to attain legitimacy without some acknowledgement of
this principle in the accountability of government to a representa-
tive assembly, elected on the basis of universal suffrage. (The
special conditions that explain the persistence of purely tra-
Dimensions of State Legitimacy 129

ditional systems in the Gulf states will be discussed in the follow-

ing chapter).
At the same time, although the principle of popular sover-
eignty enjoys such universality in the contemporary world, it
does not constitute the only legitimating principle, but in many
systems coexists more or less uneasily with other beliefs about
the rightful source of authority. In constitutional monarchies the
belief in traditionalism and heredity lives on, underpinning a
range of powers accorded the monarch, from largely ceremonial
ones in countries like Holland, Norway or the UK, to a much
more executive role in the states of Jordan or Morocco. In
communist states the principle of popular sovereignty is qualified
by assumptions deriving from Marxist-Leninist doctrine about
the leading role of the working-class party in the historical
scheme of social development. Most recently of all, in the Islamic
republic of Iran, the principle of popular sovereignty coexists
with a theocratic principle of political appointment, and the belief
in a divine rather than merely human source of law. In each of
these examples the legitimacy of the political system is only as
secure as the durability of the respective beliefs that complement
or qualify the idea of popular sovereignty among the population
at large.
It follows from the above that democratic republics, in which
the head of the executive is popularly chosen through electoral
competition, and which constitute a pure expression of the prin-
ciple of popular sovereignty, do not comprise the only legitimate
political systems in the contemporary world. It is possible for
'mixed' political systems to exist, in which the legitimacy of
government is derived from two different, even competing, sour-
ces of authority. How is such a mixture possible in practice
without contradiction? What sort of constitutional rules or proce-
dures could embrace such differences without incoherence?
One possibility is that the different principles should be
embodied in a separation of powers within the constitution, so
that the legislature is popularly elected, while the head of the
executive is selected according to different rules. This is the
position in those forms of constitutional monarchy where the
monarch retains the right to select ministers and approve policy,
but is subject to the laws passed by a popularly elected Parlia-
ment. Such a division was given classic formulation in the French
130 The Legitimation of Power

constitution of 1791, and was a typical arrangement in much

of nineteenth-century Europe. It could best be regarded as a
transitional form, in which the powers of the monarch come to
be progressively curtailed as support for the hereditary principle
declines in society at large. In revolutionary France the transition
lasted a mere two years, since the conflict between the two
principles was too acute to be contained within a single consti-
tution.
A different possibility is that the principle of popular sover-
eignty should be not so much complemented as qualified by
another source of authority. In communist systems of the 'classi-
cal' type (if those of the pre-Gorbachev era can be so called)
popular sovereignty was heavily qualified by the authority the
party derived from Marxist-Leninist doctrine, both as unique
representative of working-class interests, and as guardian of the
evolutionary path towards the future Communist society. It was
this authority that justified the one-party state, and the pre-
selection of candidates for popular election. In practice the differ-
ence between the two principles of authority tended to be
expressed in a divergence between formal and informal rules,
whereby the constitutional form of elections and a representative
assembly resembled that of a liberal democracy, but with actual
power over both residing in the ruling party. This difference of
levels explains how the system could become so rapidly trans-
formed in the Gorbachev era through a shift in the balance of
its legitimating principles, without initially any radical change in
the formal constitution.
According to the 1979 constitution of the Islamic Republic of
Iran, to take another example, the principle of popular sover-
eignty was both complemented by the executive powers accorded
Khomeini as vice-regent, and qualified by the power vested in
the religious Council of Guardians to approve candidates for
the presidency, and to scrutinise laws passed by the majlis, or
representative assembly. As one commentator put it, the consti-
tution 'left two concepts of sovereignty standing side by side ...
Iranians were still uncertain whether it was the people or God
and the clerics who ruled' (Bakhash, 1986, p.88). In the early
years of the republic the conjunction of the two legitimating
principles produced a situation of dual power, in which the
authority of Khomeini was repeatedly used to undermine that of
Dimensions of State Legitimacy 131

the first elected President, Bani-Sadr. Upon Khomeini's death,

however, and with no one of his standing to fill the position of
vice-regent, it seemed possible that the two principles might be
more effectively integrated in the position of the President, who
must be both approved by the Council of Guardians and elected
by popular vote; the precise balance of forces, however, remains
confused.
So there are mixed types of constitution in the contemporary
world, deriving legitimacy from other principles besides that of
popular sovereignty. However, even with mixed types the ulti-
mate authority of the people is recognised in the widespread
assumption that the constitution itself requires popular approval,
even when, or perhaps particularly when, that approval entails
the recognition of other sources of legitimacy for political office
than the people themselves. In other words, we should dis-
tinguish between the legitimacy of a constitutional order which
derives from the fact that its rules conform to established beliefs
about the rightful source of authority for political office; and the
legitimacy that derives from popular consent expressed at the
particularly decisive historical moment of its establishment. The
issue of expressed consent will be discussed more fully in the
final section of this chapter. It is worth noting here, however,
that one consequence of the universality of the principle of popu-
lar sovereignty in the contemporary world is that consent has to
be popular consent, even when the rules of office rest on a non-
democratic source of authority.
If that seems paradoxical, or contradictory, we should consider
the argument in favour of the French constitution of 1791, to
the effect that, since monarchical sentiment was so widespread
among the people, they would never approve of a republican
constitution; or likewise the possibility, demonstrated in Iran,
that religious beliefs be sufficiently strong among a people to
sustain a system of clerical power. Yet in each case it was felt
necessary that the people should have demonstrated their consent
to the principles of heredity and theocracy respectively.
How was such consent to be demonstrated? Most obviously
through the election of a constituent assembly, with or without
plebiscitary approval of its conclusions. However, besides the
electoral mode of expressing consent, there is also what I shall
call the mobilisation mode, whereby popular mobilisation dem-
132 The Legitimation of Power

onstrates effective support for a government or constitutional

system. The most impressive form of mobilisation is revolution
itself, in which the popular masses demonstrate their readiness
for sacrifice to overturn a discredited regime and defend a new
order in its place. It is just such actions that confer enormous
authority on a successor regime, and to which those in power
repeatedly return for confirmation of their legitimacy in face of
any challenge to it.
It is evident, then, that the principle of popular sovereignty -
the idea that the people are the ultimate source of authority for
government - has two different implications for the legitimacy
of a constitutional order in the modern world. First, that order
has to contain some element of popular representation by elec-
toral process, even if this is complemented or qualified by rules
embodying a different source of authority, whether hereditary,
theocratic or secular-doctrinal (as the Marxist-Leninist scheme
might for shorthand be described). Secondly, even the mixed
forms of constitution, and indeed especially those, require evi-
dence of popular approval for their legitimacy, whether through
a founding convention, plebiscitary confirmation, or the mass
mobilisation and suffering of a revolutionary process.
There is a third element, however, that the principle of popular
sovereignty contributes to the legitimacy of a constitutional
order, and that is the idea of national autonomy, or national
self-determination. Once it is accepted that the source of political
authority resides in the people, it becomes important to know
who exactly constitutes 'the people'. This has consequences in
particular for the spatial organisation of the state. Under the
anciens regimes, in which the state belonged to its ruling dynasty,
its boundaries were whatever happened to have been historically
determined by conquest, marriage or dynastic alliance; and it
was a matter of supreme indifference whether rulers came from
the same linguistic or cultural community as their subjects. Once
the state no longer belonged to the ruler, however, but to the
people, then the issues of national identity and national auton-
omy came to assume a wholly new political significance.
The importance of the idea of popular sovereignty in this
context was that it gave the cultural, ethnic and historical com-
munities with which people could identify, and around which
they could be mobilised, a political salience they had not pre-
Dimensions of State Legitimacy 133

viously enjoyed. To be sure, the features that gave people a

distinctive identity separate from others were historically rooted,
and developed by the slow accretions of time. Yet to become
aware of that distinctiveness, and to demand that it be recognised
and protected in a self-governing state, was something that could
happen very rapidly, though only as the principle of popular
sovereignty itself gained acceptance. Paradoxically, therefore,
because of the historical diversity of peoples, the universalistic
idea of popular sovereignty produced a highly particularistic out-
come in the form of nationalism, with profound consequences
for the spatial reorganisation of states. So at the end of the
eighteenth century the process of creating sovereign states inde-
pendent of colonial rule gathered momentum in the Americas.
The nineteenth century witnessed a wholesale reorganisation of
state boundaries in Europe, whether by secession or amalga-
mation, or a combination of the two, culminating in the dismem-
berment of the Hapsburg and Ottoman empires at the end of
the first world war. Finally, in the middle of the twentieth cen-
tury, the ideals of self-determination and national independence
were turned against the European powers by the peoples of the
Third World, and undermined whatever legitimacy their imperial
rule may have had (see e.g. Breuilly, 1982; Anderson, 1983).
What implications does the ideal of national autonomy have
for political legitimacy today, now that state boundaries are more
or less frozen by international agreement? Where a national
identity is securely established, and most people within a given
territory associate themselves with it, it provides the state with
a powerful source of support, especially in the face of external
threat or interference. Where, on the other hand, a sense of
nationhood is only weakly developed, and there is no single
ethnic or cultural community within the territory with which
people identify, its absence will be revealed in the degree of
loyalty the state can command when under pressure. Moreover,
if a particular community comes to believe that it is being dis-
criminated against or oppressed by another within the same state,
and demands political autonomy for itself, or incorporation in
another state, then nationalism, from being a unifying factor,
will become the most divisive political force, and a powerful
agent for the delegitimation of a system of rule.
The continued evidence of the internally divisive effect of the
134 The Legitimation of Power

idea of popular self-determination lies all around us, and hardly

needs elaboration. In many of the new states of the Third World
the boundaries that were inherited at independence were quite
artificial creations, bearing little relation to pre-colonial ethnic
or political groupings. In the worst examples (the Somalis or the
Kurds) historic nationalities were divided between many different
states. In Europe itself the post-colonial era has witnessed the
resurgence of older definitions of nationhood on the periphery
of many states, as regional economic disparities have intensified,
and the advantages of connection with the metropolitan centre
have declined. The example of Northern Ireland demonstrates
in extreme form the historical legacy left by past imperial conqu-
est and colonial settlement for contemporary state legitimacy. In
the USSR - last survivor of the great land empires that disinte-
grated at the end of the first world war - the upsurge of sup-
pressed nationalisms as the first fruits of 'Glasnost' confirms most
strikingly the link between popular sovereignty and the self-
determination of peoples, and the threat the latter poses to the
integrity of the state.
Now it could be argued that the problems posed by these sub-
state nationalisms involve only the geographical periphery of the
state, and do not affect the legitimacy of institutions at the
political centre. However, if they are severe enough, or intersect
with more general problems, they may readily do so. In Third
World states, in particular, managing these conflicts imposes an
enormous strain on limited resources, and provides a ready
occasion for external interference and destabilisation. For this
reason, a new constitutional settlement that redistributes power
between the centre and the regions, or involves new power-
sharing arrangements, may be the only way to reestablish the
legitimacy of the central institutions themselves, and prevent
demands for secession from bringing about a general weakening
of their authority.
In conclusion, then, the decisive shift in legitimating beliefs
definitive of the modern era, which locates the source of political
authority in the people rather than in the historical pedigree of
the ruling dynasty, has had profound consequences for the terri-
torial organisation of states and for the spatial distribution of
power within them, as well as for the rules governing access to
political office itself. Above all, it means that the ultimate court
Dimensions of State Legitimacy 135

of approval for constitutional arrangements rests with the people,

and that popular mobilisations of protest, whether against the
rules of political office or the spatial distribution of power, carry
a potent delegitimating force. However much it may be modified
by other principles of authority, and whatever complexities it
may assume in different contexts, it is with this basic principle
of popular sovereignty that we must begin if we are to understand
political legitimacy in the contemporary world.

ii) the satisfaction of a general interest

Besides being derived from a source acknowledged as authoritat-

ive within the society, state power has also to satisfy a general-
interest criterion. As I argued in Chapter 3, all systems of power
require justification in terms of meeting the interests of the subor-
dinate, as well as the dominant, and of fulfilling requirements
necessary to the society as a whole. The difference between the
traditional and the modern state in this regard is not that the
former, being the private preserve of the ruler, had no public
functions to serve, but that it fulfilled them largely from private
resources, and no clear distinction was made between the public
and the private in the finances, the domain or even the person
of the ruler. The modern state, by contrast, being based upon a
sharp distinction between the public and private spheres and
their respective concerns, exists to serve public needs by purely
public means; and all aspects of law, policy and finance have
therefore to satisfy a public interest criterion, and to be justifiable
on these grounds.
Now of course all governments will claim that their actions
serve the general interest, and their opponents will claim that
they fail to do so. Since there is no incontestable criterion of
how the general interest can best be realised, such claims would
seem to belong to the rhetoric, rather than the substance, of
legitimation. However, we should distinguish between the
normal, everyday disagreements over policy, and the settled con-
viction on the part of a substantial section of the population that
a government is seriously compromising the public interest.
There are two different ways in which it might do so. The first
is through a manifest failure of performance, such as presiding
over military defeat or rampant inflation, where the government
136 The Legitimation of Power

has clearly failed in some essential purpose that the state exists
to fulfil. The second is through manifest particularity, or partial-
ity, such as the abuse of public office for private gain, or the
pursuit of policies that advantage one section of society at a cost
to the vital interests of another: where government, in other
words, is seen to be pursuing a particular or sectional interest at
the expense of a general one.
Now all these are the presumed failings of governments in the
first instance, rather than of constitutional systems; and most
constitutions have some means of removing those officeholders
who have lost public confidence, whether through parliamentary
vote, popular election or some more arcane process of dismissal.
Such dismissals ensure that any loss of prestige is confined to
the individuals concerned, and does not damage the consti-
tutional rules themselves; indeed, confidence in them can be
enhanced by the promptness or effectiveness of dismissal. How-
ever, if government failings are chronic, or the prospect of
remedy through constitutional process is uncertain or remote,
then the legitimacy of the political system will be eroded also.
Moreover, it is characteristic of particular types of constitutional
arrangement that they may exacerbate government failings in the
first place, or hinder their resolution in the second. For this
reason the rules of the political system are involved as well as
individual governments, and we must pay attention to the inter-
action between the two.
In view of the complexity of these issues involved in a 'general
interest' criterion, I have divided the following discussion into
three parts. The first (section a) considers government perform-
ance in respect of the basic ends it exists to serve. The second
(b) discusses the question of government partiality or sectional-
ism, and the distinction between particular and general interests.
The third (c) distinguishes between the impact of failure or sec-
tionalism on individual governments and on a political system as
a whole, and considers the effect of different constitutional rules
in exacerbating or resolving problems of government perform-
ance.

a) performance and the ends of government

In a much quoted essay, S.M. Lipset drew a distinction between
the 'legitimacy' and the 'effectiveness' of regimes. 'Effectiveness,'
Dimensions of State Legitimacy 137

he wrote, 'means actual performance, the extent to which the

system satisfies the basic functions of government' (1958, p.86).
At one level this distinction is valid, since the two concepts are
not coterminous. Yet in practice there is a reciprocal connection
between the two. I have already drawn attention to the way in
which legitimacy makes a difference to the effectiveness of a
system of power, through the quality of performance it secures
from those subordinate to it. Equally importantly, the perform-
ance of government makes a significant contribution to, and
forms a necessary component of, its legitimacy (Rothschild, 1977,
p.488). After all, how can the enormous powers of the state be
at all justified, or people be obliged to obey it, unless it fulfil
requirements necessary to the society and their own well being,
and that it fulfil them effectively? Lipset's mistake here is to
regard performance as something entirely instrumental, and not
also as satisfying normative or moral criteria. And that in turn
betrays a conception of morality which confines it to an entitle-
ment ethic, based upon origins, to the exclusion of a utilitarian
or consequentialist ethic, based upon ends.
This point can be clarified by reference to the distinction drawn
by moral philosophers between the 'right' and the 'good', or to
the distinction made by political philosophers between justifi-
cations based upon rights and those based upon social welfare or
utility. 'Rights-based' arguments are typically backward-looking:
they make reference to how entitlements are acquired, from
whence they derive, or what people have done to merit them.
Arguments from social welfare or public interest are typically
forward-looking: they seek to show what contribution a given
institution or set of political arrangements makes to social utility.
The distinction between the two types of argument can be clearly
seen, for example, in justifications for private property, and the
rules underpinning it (Reeve, 1986, ch.5; Ryan, 1987, pt.2). Any
complete justification, however, needs reference to both types of
argument. And the same is true of political legitimacy: it requires
both a morally authoritative source for government, and an
ability to satisfy the ends which justify its enormous concentration
of power. While the logic of each needs to be distinguished, not
least because they have different institutional implications, both
are equally necessary to legitimacy (Schaar, 1969, pp.287-8).
What ends, then, does the state exist to fulfil? Unless we are
138 The Legitimation of Power

anarchists, most of us take it for granted that the justification

for the state lies in its provision of physical security and of the
conditions necessary to material welfare. Marxists may claim that
the state fails to make such provision, or in a way that satisfies
the general interest, but this claim presupposes that it ought to
do so. Neo-liberals may claim that material welfare is a matter
for individuals and the market, but even they cannot in practice
absolve governments from all responsibility for economic per-
formance. Others may claim different purposes for particular
states, such as the promotion of international revolution, the
defence of the 'free' world, or the salvation of people's souls;
yet these are additional rather than alternative to its security and
economic purposes.
The definitive statement of the view that the sovereign state
exists to guarantee the physical security of its inhabitants, of
their persons and property, is to be found in the work of Thomas
Hobbes (see Leviathan). According to Hobbes, the state's mon-
opoly of the means of organised physical coercion is necessary
to protect individuals against other members of society from
within, and collectively against the threat of invasion from with-
out. It is from the recognition of this necessity that people's
consent to the state is seen by Hobbes to derive, and their
corresponding obligation to obey it, whatever its form, provided
only that it continues to guarantee their security. Social order
and personal security depend upon a coercive framework; the
effective organisation of coercion to this end is the basis of the
state's rationale and hence of its legitimacy.
Hobbes's argument is an important one, not least in warning
against any oversimple antithesis between coercion and legit-
imacy. In so far as the state's purpose is to ensure the physical
protection of its citizens, then the effective application of
coercion to this end cannot necessarily be construed as either
illegitimate in itself, or as evidence of a deeper-seated weakness
in legitimacy. On the contrary, a persistent failure to guarantee
physical security will undermine confidence in the system of
government, since it will be seen to be failing in its most essential
purpose. It is precisely such a failure that provides one of the
most typical justifications for the intervention in politics of the
military, as the special guardians of 'order and security'; in so
Dimensions of State Legitimacy 139

intervening, they would seem to be conforming to a characteristi-

cally Hobbesian rationale.
However, if the effective threat or use of physical coercion is
a necessary condition for social order and hence for the state's
legitimacy, it cannot be a sufficient condition, as Hobbes's theory
supposes. One of the prime causes of disorder within any society
lies in the absence of agreement about some aspect of its power
rules (the rules of property, division of labour, political office,
citizenship, or whatever), and in the social conflict to which such
disagreement gives rise. Disorder and insecurity are as much the
product of inadequate legitimacy as they are its cause. To the
extent that this is so, any attempt to suppress discontent or
disagreement by force, rather than resolve it by political means,
may only weaken a state's legitimacy rather than enhance it.
Hobbes's theory of legitimacy is wrong in supposing that the
effective enforcement of rules is a sufficient ground for obligation
regardless of their content. And it follows that his explanation
of social order is mistaken in imagining that it can be guaranteed
in the absence of agreement about these rules.
The relation between coercion and legitimacy is thus a complex
rather than a simple one, at least as regards the state's function
in ensuring internal security. Here the use of coercion to suppress
deep-seated grievances can erode legitimacy in some contexts as
surely as the failure to use it to quell disorder can in others; the
state can be both too weak and too overbearing. The issue is
more straightforward when it comes to defence against external
threat, despite disagreements about how such defence is best
secured. Nothing is more damaging to a government's authority
than military defeat, or the widespread conviction that the
country's defence is being compromised through incompetence
or insufficient expenditure. At the same time countries vary enor-
mously in their geo-political position, and in the resources they
are capable of devoting to defence without crippling their econ-
omies. Although inherent military weakness can be offset by
alliance policy and collective security arrangements, these can
readily infringe national autonomy where much larger powers
are involved. In particular, the global contest between the super-
powers that has characterised the decades since the second world
war has had a decisive impact on the security of many states,
whether through the destabilisation of 'opponents', or the com-
140 The Legitimation of Power

promising of the autonomy of 'allies'. As a consequence, in many

regions of the Third World especially, the attainment of security
has been only partially under the control of the supposedly sover-
eign state.
The diminution of autonomy in face of an uncontrollable inter-
national environment that affects a country's physical security
applies even more to its economic well-being. Nowadays the
expectation that the state should guarantee the conditions for
the satisfaction of its citizens' material needs is as universal as
the demand for physical security. This expectation was given
particularly elevated formulation in the US Declaration of Inde-
pendence, with its claim that the 'pursuit of happiness' was a
universal right that states should uphold (Thorpe, 1909, pp.3-4).
A similar demand was expressed in more utilitarian terms in
Britain by Jeremy Bentham and James Mill, with their criterion
of good government as that which secured 'the greatest happiness
of the greatest number' (Mill, 1937; Bentham, 1960). At the
time they were made, these demands were not interpreted as
meaning that the state should itself provide for people's welfare;
rather, it should remove the obstacles to people pursuing it for
themselves. The demand was thoroughly laissez-faire in inspi-
ration. What was radical about it was the claim that everyone's
happiness should be given equal consideration.
However, as the political influence of the masses was extended
during the course of the nineteenth century, so the responsibility
of the state came to be interpreted in a more interventionist
manner: in the first place that it should protect people from the
most damaging effects of laissez-faire capitalism; then that it
should provide directly what the market itself could not; finally,
that it should take responsibility for the overall performance of
the economy. By the middle of the twentieth century the expec-
tation had become general that the state would secure a progress-
ive improvement in the satisfaction of material needs, and that
governments should be judged by their capacity to do so.
Although this presumption can be baldly stated, the actual
relationship between economic performance and government
legitimacy is a complex one, depending upon the pattern of
distribution of economic costs and benefits, as well as on the
overall level of performance. It also differs systematically accord-
ing to the type of politico-economic system or production regime
Dimensions of State Legitimacy 141

in force. As will be more fully explored in the following chapter,

governments in a centrally planned economy can much less read-
ily distance themselves from economic failure than in a market
economy, since by definition they are directly responsible for it.
However, this is a difference that can be easily overstated, and
is less important than differences in economic performance itself.
Here the divergence between developed and developing,
industrialised and industrialising economies in their ability to
satisfy the material needs of their population is most obvious
and glaring. Industrialisation in particular is a process that is
everywhere accompanied by enormous inequalities and dislo-
cations, which are rendered particularly acute for the developing
countries by the manner of their insertion into the international
economy. Although generalisation can be misleading, in view of
the variety of developing economies, many have been locked into
a perpetuation and extension of their colonial role of producing
primary commodities for export, a role which has reduced self-
sufficiency in food on one side, and made the economies particu-
larly vulnerable to the price fluctuation of a few commodities on
the international market on the other. The rise in the oil price
in the 1970s, and the collapse in the price of many other primary
commodities, as the Western economies went into recession,
proved catastrophic for many developing countries. The typical
consequence was acute indebtedness and subjection to conditions
imposed by the IMF, which bore most heavily on the impover-
ished majority of the population (George, 1988; Onimode, 1989).
Now, it is a condition of the present world economy that no
country is entirely autonomous in economic policy; the state of
each is dependent to a greater or lesser extent on what is happen-
ing outside it. However, the degree and character of that depen-
dence is what matters. Developed economies are precisely ones
that are sufficiently diversified and self-sustaining to be able to
protect their citizens from the most damaging effects of the world
market, and to adjust to its changes without undermining the
conditions for their further development. So-called 'dependency'
is a state in which neither of these is possible. Although blaming
external forces can provide a convenient scapegoat for the policy
failures of governments, there is no escaping the disadvantage
of being late-comers to economic development in a world already
well-advanced in that process (Wilber, 1984; Szentes, 1983).
142 The Legitimation of Power

In conclusion, the relative success of the sovereign states of

the developed world in meeting the security and welfare needs
of the majority of their populations has passed on to the peoples
of the developing world the expectation that the sovereign state
can and should achieve the same for them. Indeed the claim
frequently made at the time of independence from the colonial
powers was that sovereignty would of itself guarantee these ben-
efits. 'Seek ye first the political kingdom, and all else will be
added unto you,' said Nkrumah. In its simple form such a claim
overlooked the long history of struggle, even in the oldest states,
both to establish a viable state in the first place, and then to
extend its benefits from the few to the many. It also overlooked
the systematic dependencies on the international system that
serve to perpetuate the disadvantage of the economically and
militarily weak. There is a marked discrepancy, we might con-
clude, between the power the contemporary state possesses to
control and coerce those subordinate to it, and its ability to meet
the social needs that alone can justify its coercive and intrusive
power. It is in just this discrepancy that a recurrent problem of
governmental legitimacy resides.

b) particular and general interests

Governments may fail to satisfy a general interest requirement
because they are unable to fulfil some basic function of govern-
ment, whether of public security or economic management. They
may also fail because they are perceived as partial in the way
they carry out their functions: because they use public office to
satisfy merely particular or sectional interests. There are connec-
tions between these two aspects of performance, since, for exam-
ple, the condition of underdevelopment that makes the problems
of economic management so intractable is also one in which the
competition for scarce resources is most acute, and there is thus
increased pressure to use state power to favour certain groups
at the expense of others, whether legally or illegally. Partiality
and corruption exists in all kinds of political system, at all stages
of economic development; but they are more acute or chronic
in some situations than others.
From a purely legalistic standpoint, a sharp distinction can be
drawn between the use of the law to favour certain groups in
the population, and the illegal abuse of office for the personal
Dimensions of State Legitimacy 143

benefit of the office-holder and his or her connections. The

former, being a matter of policy, has to be framed in general
categories, and to be subject to open justification and defence in
terms of the public interest. The latter, being a matter of private
and particularistic advantage, is necessarily secretive, since it
cannot withstand public scrutiny, and it threatens the basic div-
ision between the public and private spheres that is essential to
the modern state. The characterisation of the latter as 'corrup-
tion' underlines the moral as well as legal distinction between
the two types of action.
From more substantive considerations of the general interest,
however, the difference is not so clear -cut. On the one hand a
government composed of super-tax payers which votes enormous
tax reductions for the rich will hardly avoid the charge of partial-
ity, simply because its beneficiaries comprise a general category
of the population, rather than the personal acquaintances of
those in office. On the other hand political systems that routinely
operate on a patronage basis, for instance by channelling govern-
ment expenditure to areas of electoral support, may have a
public opinion that is more tolerant of illegality than those which
operate by more universalistic criteria; and this tolerance may
be reflected in a different definition of what counts as 'corruption'
in the first place. There are many different ways of rewarding
one's own in politics; and the moral distinctions between them
are less clear-cut than can be captured by a simple contrast
between the legal and the illegal, between clean hands and the
finger in the till (J.C. Scott, 1969).
To say this is not to underestimate the discredit into which
the law can fall if governments are seen to flout it with impunity,
nor to overlook the political impact of particularly sensational
scandals, even in societies where a certain level of corruption is
routinely tolerated. However, what is decisive for crystallising
the impression of a government's partiality is not primarily
whether its members act legally or illegally. Nor is it even
whether its policies are designed to favour its own, since all
governments do that to a greater or lesser degree. What matters
is whether, and how far, in doing so, it damages the interests of
other definable groups in society; and whether these groups are
sufficiently organised and powerful to make a significant protest.
These criteria reflect two simple truths. The first is that unequal
144 The Legitimation of Power

treatment is more tolerable where it does not take the form of

a zero-sum contest. Thus governments can favour their own with
impunity if the cost is funded out of general taxation, and so is
spread across the population at large, rather than debited to a
particular section of it. But secondly, even where the cost is
borne by a particular section of society, it will not be politically
significant if the section affected is too isolated or disorganised
to make any visible protest. It is where the benefit to one group
is manifestly gained at the expense of another, and that other
group is capable of organised protest, that the impact will be
politically divisive, and the government will be convicted of par-
tiality, of serving particular, or as the utilitarians termed it,
'sinister' interests.
The point can be illustrated from the example of corruption.
Much corruption - for example, taking bribes on government
contracts, or favouring acquaintances in official appointments -
has the character of a victimless crime, in which the costs are
simply passed on to customers at large, or borne by society as
a whole in reduced efficiency. This presumably explains why it
is tolerated, where it is so. However, where government consti-
tutes a major source of privileged employment, or controls sub-
stantial access to business opportunities through its licensing and
contracting powers, then the cost to the excluded rises sharply,
and resentment at their exclusion is correspondingly intense. And
if the excluded comprise a section of the population identifiable
on ethnic, regional or religious grounds, then the government's
favouritism is bound to be socially and politically divisive. The
excluded will protest loudly at corruption and illegality; but what
they are really protesting about is their own exclusion, and that
the favouritism takes place at their particular expense (Szeftel,
1983).
What applies to corruption applies equally to public policies,
legally formulated and implemented, that advantage one section
of the population, where they do so at the manifest expense of
another. In racially or ethnically divided societies it is almost
impossible to favour one group without damaging the other,
especially where this involves the modification of existing
inequalities. In class-divided societies it is almost impossible to
enhance the powers or welfare of subordinate classes without
infringing the rights of property owners. The redistribution of
Dimensions of State Legitimacy 145

powers by political means - the only means available to the

subordinate and disadvantaged - is intrinsically more contentious
and conflict-laden than the reproduction of inequalities through
the persistence of natural-seeming social rules. The latter process
may become conflictual; the former is invariably so. Here an
appeal to general criteria of justice or welfare, however necessary
in matters of public policy, does not resolve the issue, when
competing criteria tend to be selected according to social position
in the first place. Nor will an appeal to the majoritarian principle
avail either. From the standpoint of the minority, if what they
define as their vital interests are threatened, it makes little differ-
ence by whose agency, or in the name of whom, they are so.
Just such a consideration inspired J.S. Mill's contention that
the majority could itself constitute a particular or 'sinister'
interest, and that the general interest would be more effectively
realised by taking equal account of majority and minority
interests, than through the suppression of one by the other.
Behind Mill's special pleading on behalf of the propertied there
lay the valid insight that, once contentious issues of redistri-
bution, driven by competing interests and theories of justice, are
at the forefront of the political agenda, a recognisably general
interest can only be attained through compromise between them
(Mill, 1964, pp.248-56). Whether such a compromise is in prac-
tice attainable depends upon the room for manoeuvre available
within the economy, upon the ability and willingness of the
respective parties to make concessions, and upon the extent to
which the constitutional system itself facilitates or discourages it.
This latter point in particular brings us to a consideration of the
relation between constitutional rules and the general interest.

c) government performance and political legitimacy

In discussing the relation of government performance to political
legitimacy it is important to stress once again that the various
failures to satisfy a general interest requirement, whether through
inability, incompetence or partiality, are failings in the first
instance of governments, not of political systems. Too many of
those who write about performance as a legitimating criterion
omit to observe this distinction. In doing so they overlook the
simple point that most constitutions are designed to allow for
the replacement of governments or leaders who have lost public
146 The Legitimation of Power

confidence; and that the process of replacement enables alterna-

tive policies to be pursued by a different administration enjoying
renewed authority. Such replacement can be seen as a means of
ensuring that the discredit associated with failure does not attach
to the constitutional rules themselves.
What we must ask, therefore, is under what circumstances
the political system itself becomes infected with the odium of
governmental failure, or the conviction becomes established that
the general interest cannot be secured under the existing consti-
tutional arrangements. One situation is where the procedures for
dismissal have ceased to be effective, because either they or the
personnel responsible for them are under the control of those
who need replacing. In extreme cases a society has to wait for
its rulers to die before a change of direction becomes possible,
however disastrous the policy failures with which they are associ-
ated. Single party systems, in which there is no electoral oppor-
tunity or incentive to remove a failed leader, who has all the
powers of patronage available to consolidate support, have a
marked tendency towards immobilism and gerontocracy. In these
contexts it is difficult for the political system itself to avoid the
odium of governmental failure; and change, when it eventually
comes, has all the character of a major convulsion rather than
an orderly transfer of power. The advantage of multi-party sys-
tems in this respect is that the regularity of the electoral process
gives party hierarchies a powerful incentive to remove leaders
once they have forfeited public confidence, so as to minimise the
prospects of electoral defeat.
If on one side, then, confidence in constitutional rules can be
eroded by the inability to replace those who have failed, on the
other it can be undermined by the conviction that a change in
government will bring no improvement. This may happen
because a particular section of society believes that the damage
inflicted on it, or in prospect, is irreversible. Or the electoral
rules may make it impossible for a given minority ever to achieve
a share of political office, so that it is the system rather than just
a particular government that is stacked against them. Or the
prospect of one sectional group pursuing its own interests in
government, only to be followed by another doing the same,
destroys confidence in the political system's ability to deliver
impartial government at all.
Dimensions of State Legitimacy 147

Here the part played by constitutional rules in exacerbating

social conflict and government partiality merits attention. A
system of electoral competition between parties may be effective
at removing those who have forfeited public confidence, but its
corresponding disadvantage is that it readily assumes the charac-
ter of an ali-or-nothing contest between different sections of
society. To the victors it affords the chance of supreme power,
but for a limited term, and with uncertain prospect of being able
to control the external circumstances on which performance so
much depends; they have a corresponding incentive to maximise
the advantages of power in the short term. To the losers it
guarantees powerlessness, and with uncertain prospect of recov-
ery in the future. Unlike the competition for position in other
areas of social life, where defeat does not necessarily signify
exclusion from power, and there are always opportunities to try
again elsewhere, the concentration of political power in a central
set of institutions makes defeat in electoral competition particu-
larly conclusive. If the damage consequent upon defeat is per-
ceived by the losers as irretrievable, whether on ethnic, religious,
economic or personal grounds, then the pressure not to accept
the outcome, or to cheat in the process, becomes enormous.
It is worth recalling the advantages of traditional systems in
this particular regard. With executive office at the disposal of
the ruling monarch or chief, competition for it was necessarily
limited by the ruler's own authority, which was itself, except in
rare cases of disputed paternity, beyond the scope of compe-
tition. The electoral mode of appointment, in contrast, renders
that competition in principle unlimited and unrestrained. This is
particularly true of winner-take-all political systems like the Brit-
ish, where first-past-the-post electoral rules are combined with a
lack of constitutional restraints upon the executive. It is not
surprising that the 'Westminster model' has proved largely
unworkable outside the special circumstances of mainland Brit-
ain. To say that this is because of the absence of a 'democratic
culture' or the insufficient integration of key political actors into
a common 'establishment' or political elite, as in Britain, is
simply to restate the problem, since these essentially cultural
conditions cannot be created overnight. Indeed they are them-
selves the product of an accumulated history of acceptance of
electoral outcomes on the part of the losers. And that requires
148 The Legitimation of Power

that not too much should be at stake in any one contest: that
the costs to the losers in any competition for office should not
be so high as to remove all incentive to abide by the outcome.
The weakness of the Westminster model is that meeting this
requirement depends on the self-restraint of the victors, and on
tacit conventions that can readily be overridden. The require-
ment can be more securely met by constitutional rules that mini-
mise the costs of defeat to the losers, e.g., by ensuring them
some power or influence so that they are not left completely
excluded from the political process by their electoral defeat.
Federal systems secure this by providing a local power base in
the event of defeat at the centre. Systems with a clear separation
of powers between a legislature and elected president may prod-
uce an outcome in which no one party controls both. Electoral
rules based upon proportional representation are more likely to
require compromise between political parties, while more formal
power-sharing arrangements may be needed in societies charac-
terised by deep ethnic or cultural division (so-called 'consoci-
ational democracy') (Nordlinger, 1972; Lijphart, 1977). All these
can be seen as devices that limit the damage of electoral defeat
in multi-party systems, and therefore encourage a tradition of
accepting electoral outcomes on the part of key political actors.
Alternatively, the cost of defeat may be minimised through a
limitation in the scope of electoral choice itself. This happens
where elections are restricted to the legislature, rather than the
executive, or where the choice is limited to candidates of a single
party. Single-party systems, whatever their drawbacks in other
respects, are much less politically divisive than multi-party ones,
because they typically embrace all the major social forces within
them, and because they minimise the scope, and therefore the
fatefulness, of electoral competition. The other side of the coin,
obviously, besides the tendency to immobilism already men-
tioned, is the reduction of public accountability. And from the
standpoint of government legitimacy, the formal restriction of
electoral choice necessarily reduces the force of popular legitim-
ation through expressed consent. This, together with the role of
political parties in legitimation, will be discussed more fully in
the final section of the chapter.
It remains here, by way of recapitulation, to underline the
differences between the legitimating criteria discussed in this
Dimensions of State Legitimacy 149

section, and to draw out some implications that will be developed

in later chapters. Legitimacy as the justifiability of constitutional
rules comprises two main elements, so I have argued, reflecting
the two normative criteria required to justify state power: it must
derive from a source that is acknowledged as authoritative within
the society; it must serve ends that are recognised as socially
necessary, and interests that are general. Constitutional rules in
turn will be legitimate to the extent that they conform to estab-
lished beliefs about the proper source of authority, and that they
facilitate rather than obstruct the attainment of a general interest
in respect of those purposes which governments exist to fulfil.
Their legitimacy will be deficient to the extent that they fail in
either of these respects.
Although this summary may seem highly general and abstract,
two very specific conclusions can be drawn from it for the con-
temporary age, in which popular sovereignty is recognised as the
main, though as we have seen not the only, source of authority
for government. The first concerns the different points of vulner-
ability of different types of political system. What I have called
'mixed' systems, in which the principle of popular sovereignty is
complemented or qualified by another source of authority, are
most vulnerable to an erosion in the beliefs that underpin that
other authority, and therefore in support for the constitutional
rules that qualify or limit the scope of the electoral process.
Liberal democracies, on the other hand, which most fully
embody the principle of popular sovereignty in the openness and
scope of their electoral process, are most vulnerable to the effects
of social division, in whatever form, and to an inability to sustain
a recognisably general interest. Military regimes, finally, that
typically emerge to 'rectify' the failings of democracy, are simply
incapable of establishing a credible source of authority for their
power, let alone rules of office that might embody it.
A second conclusion concerns the agencies through whom
legitimacy deficits and legitimacy crises are characteristically ter-
minated in the contemporary world, whether by a breach of the
constitutional order, or some other mode of resolution. Where
the legitimacy deficit or weakness occurs because the consti-
tutional rules do not conform sufficiently to accepted beliefs
about the proper source of authority - and that typically means
that they do not sufficiently embody the principle of popular
150 The Legitimation of Power

sovereignty - the main agent of political transformation is the

popular masses as chief 'bearers' of this principle. Where the
deficit occurs because of the social divisiveness of parties or
politicians, the chief agent of political transformation is the mili-
tary, as claiming to embody a national interest above the section-
alism of party, and representing the permanent interests of the
state beyond the transience and particularism of politicians. In
brief, in the contemporary world, deficits involving the basic
source of authority generate demonstrations of people power and
revolutions; deficits in performance alone and in realising a gen-
eral interest, without erosion of belief in the source of authority,
produce demonstrations of military power and coups d'etat.
These, necessarily oversimplified, formulations will be elabor-
ated and, where necessary, qualified in the chapters that follow.
It is time now to examine the final aspect of legitimacy relevant
to the contemporary state, that of expressed consent.

Modes of popular consent

Previous chapters have shown how the consent of the subordinate

makes its own distinctive contribution to the legitimacy of power,
through the symbolic and normative force of actions which are
conventionally recognised as expressing consent to the powerful,
and, by implication, to the rules of power or constitutional
system also. As I have argued, such actions confer legitimacy on
the powerful, both through the public acknowledgement that is
made of their position, and through the obligations that derive
from that acknowledgement. To have this effect, they must be
positive actions taking place in public, since inaction or privacy
can have no legitimating force. In relation to the legitimacy
that governments derive from their appointment according to
justifiable rules, that deriving from consent should be seen as
complementary, rather than alternative, to it. That is to say,
however secure the processes and procedures of their appoint-
ment may be, governments also need the confirmation and con-
solidation of their legitimacy that derives from the explicit con-
sent of the governed.
As I have also argued, it was a characteristic of 'traditional'
societies that consent could be given on behalf of others. In the
Dimensions of State Legitimacy 151

political sphere those who were entitled to give consent in person

- who comprised what one might call the political community -
were few in number and invariably belonged to the socially
and economically powerful. Their actions indicative of consent
(swearing an oath of allegiance, participating in a consultative
council, negotiating terms of support, etc.) were taken to include
their own subordinates and dependants along with them. This
was all changed by the idea of popular sovereignty, which not
only altered the rules of political appointment as discussed above,
but extended the political community - those who were entitled
to participate in politics - to the whole adult population, rather
than a privileged section of it. Consent in the contemporary era
now has to be popular consent if it is to have any legitimating
force.
There is, however, more than one form of political partici-
pation through which consent to government can be demon-
strated in contemporary political systems. We can distinguish
between the electoral mode, and what I have termed the mobilis-
ation mode; each has a different logic and different institutional
implications. In the electoral mode the act of voting, through
which governments are directly or indirectly appointed, is also
the means through which the consent of the governed is demon-
strated; elections thus perform two quite distinct functions
simultaneously. In the mobilisation mode the two processes are
separated, since the mass participation in political activity at the
grass roots which demonstrates the continuing popular support
for the regime is divorced from the process whereby office hold-
ers are appointed. As a consequence, political parties serve very
different functions in the two modes: in the one, to prepare
leaders and policies for electoral choice and approval; in the
other (typically single-party regimes), to mobilise the political
activity on which the system's legitimacy depends. Although
some political systems combine both modes of consent, in what
follows I shall treat the two as distinct types for analytical pur-
poses.
As I have demonstrated in Chapter 3, the electoral mode of
consent derives from two ideas that are central to the individual-
ism of the liberal tradition: that no one is entitled to give consent
on behalf of another unless specifically authorised to do so; and
that consent, to be voluntary, requires an effective choice
152 The Legitimation of Power

between alternatives. Earlier in the modern era, consent to

government was given, not through election of the executive
itself, but through the election of a representative assembly which
had a limited power of veto over an executive appointed by the
monarch. Consent to government, in other words, was transmit-
ted indirectly through representatives authorised to act on their
constituents' behalf. In the typical liberal-democratic system of
the present day, consent is also given directly through elections
for the head of the executive. Effective choice is guaranteed by
the freedoms of expression and association that allow parties and
policies to be formed and modified in a manner that reflects a
range of public opinion. Exercising a choice between them
ensures express consent to the government that is so chosen: on
the part of the majority, because they have voted for it; on the
part of the minority, because by participating in the election they
are assumed to have demonstrated their acceptance of the rules
by which the government was chosen.
In the previous section I noted an important advantage that
traditional systems enjoyed in preserving the supreme office from
the impact of public competition. In the context of consent they
also derived a significant source of stability from the fact that
the political community - those whose cooperation was needed
to realise the goals of the ruler, and whose consent was necessary
to confirm his or her legitimacy - was largely coterminous with
the socially and economically privileged and powerful. The distri-
bution of political power reflected the distribution of power in
the economic and social spheres. In the contemporary world,
by contrast, this coincidence is no longer guaranteed, since the
requirements of popular consent have extended the political com-
munity to the economically and socially subordinate, and their
right to electoral choice has given them the potential capacity to
alter the terms of their subordination through a revision of the
rules on which it rests. In terms of the dimensions analysed in
the first half of this book, the formal distribution of political
power no longer reflects the power relations of class or gender,
as it did in the traditional order. This disjunction constitutes a
potential source of tension and conflict that is ever-present within
the electoral mode of consent.
The classical bourgeois era succeeded in protecting itself from
this disjunction and its consequences in a period of intensified
Dimensions of State Legitimacy 153

class and gender inequality by limiting the suffrage to male prop-

erty-owners. The limitation was justified by arguments that dem-
onstrated more ingenuity than consistency with the universalist
premises of individualism itself. Thus Locke developed the idea
of 'tacit consent', arguing that, whereas consent to taxation
required explicit authorisation by the propertied through rep-
resentatives, consent to government as such could be inferred
from the mere fact of living under it and enjoying its benefits
(Locke, 1967, pp.365-81). In like manner the constitution makers
of the first National Assembly in France drew a distinction
between 'passive' and 'active' citizenship: between the enjoyment
of legal rights, which was universal, and the right to influence
public affairs through representatives, which was limited to tax-
payers, to those who, in Abbe Sieyes' words, were 'the true
shareholders in the great public enterprise' (Stewart, 1951,
pp .129ff.). While the exclusion of the working classes was thus
justified by appeal to a thoroughly bourgeois image of political
society as a shareholders' enterprise, the exclusion of women
was confirmed by the persistence of a thoroughly traditionalist
conception of consent. Since women had no distinct interests
separate from their menfolk, James Mill argued, they could safely
be represented by either their fathers or their husbands. Consent,
in other words, could continue to be given on someone else's
behalf without their explicit authorisation (Mill, 1937, p.45).
Nowadays such arguments seem only too evidently the special
pleading of the male and the propertied, and eventually they
succumbed to the universalist assumptions inherent in classical
liberalism itself, though only when harnessed by mass popular
struggles on the part of the working class and women in turn.
Today there is no longer any defensible point at which a line
can be drawn to limit those eligible to vote, except on grounds
of age or criminal conviction. In the Republic of South Africa,
for instance, the racial principle of eligibility has provoked a
deep-seated crisis of government. The exclusion of non-whites
from the political community has led to the active withdrawal of
consent, and the delegitimation of the state in respect of its
authority over the black and coloured populations. The RSA
proves the truth of the proposition that, if any are to have the
vote, it has to be given to all; if any are to be excluded, then
everyone has to be.
154 The Legitimation of Power

The powerful and privileged can thus no longer protect them-

selves against the threat from below by means of a discriminatory
electoral system. Social and economic inequality has to coexist
with a formal political equality in the status of citizenship, and
with the potential ability of the subordinate to use their electoral
power to alter the terms of their subordination. However, just
as there are informal social processes at work, explored in the
first part of the book, which serve to reproduce the legitimacy
of class and gender power, so also can these power relations
reassert themselves within the formally equal sphere of citizen-
ship. This is most apparent in relation to gender. The gendered
division of labour reproduces itself in the political domain, as
in other areas of life, through the cumulative pressure of role
expectations, of time available from domestic responsibilities,
and the differential desire or opportunity to develop the qualities
deemed appropriate to the exercise of political power (Randall,
1982, ch.3). As a consequence, parties and legislatures - the
main instruments of political change - remain dominated by men,
who are thus able, if not to block all reform, at least to control
its pace and content in a manner that does not prove threatening
to themselves. This control, combined with the divisions among
women themselves, has ensured that, since the era of the suffrage
struggles, the political activity of women as women (rather than
as members of other social groups) has not threatened the legit-
imacy of the political order.
There are some respects in which the history of class reflects
that of gender. Most obviously, wealth provides a powerful politi-
cal resource that can be deployed to influence the outcomes
of electoral choice. Moreover, as with gender, the hierarchical
relations of production can be readily reproduced within political
parties, to consolidate vertical social bonds at the expense of
lateral, class-based modes of organisation. This happens particu-
larly in rural areas, where electoral politics can serve to perpetu-
ate traditional clientelist relations. Voting for a social superior
is rewarded by particularistic benefits for the locality, which dis-
courage more programmatic demands that might challenge the
terms of subordination themselves. However, it is when class-
based organisations begin to be formed and win a following that
a crisis of electoral politics typically develops. Whether it can be
resolved within the framework of the electoral system depends
Dimensions of State Legitimacy 155

crucially upon the ability and willingness of the dominant classes

to effect a compromise that gives enough to the subordinate to
sustain their support for the political order (see below,
pp.177-8).
If they lack that ability or willingness, the dominant classes
will seek instead to secure some restriction on electoral choice.
This may be done by outlawing particular movements or parties;
or by employing state forces or private armies to repress them;
or by general restrictions on the freedoms of expression and
association. Any of these may enable the electoral process to
continue in a manner that does not threaten the system of class
power or the privileges of the powerful. It follows, however, that
such elections cannot have the same legitimating force as before,
since they infringe the conditions necessary to consent in the
liberal mode. Once popular choice is artificially restricted from
above, through the operation of the law, or of force, to limit
the freedom of association and organisation, then the outcome
cannot be said to have the voluntary consent of the governed.
At best we could speak of a semi-legitimate, or incompletely
legitimated, government, and of a degeneration in the electoral
mode of consent, whose logical conclusion is the single-party
system or the no-party regime, offering at most a choice of 'yes'
or 'no' to the individual leader.
A lack of electoral choice does not of itself, however, indicate
the absence of all popular legitimation. Here we need to consider
the alternative, mobilisation mode of consent. In the mobilisation
mode, consent is expressed through continuous mass partici-
pation in political activity supportive of the regime and contribu-
tory to the realisation of its political goals (cf. Denitch, 1976,
ch.S). Regimes legitimated in this way typically take their origin
from a revolution, and the continuing popular mobilisation can
be seen as a perpetuation of the revolutionary process into the
post-revolutionary era. Although the majority of the population
may in fact be politically inactive, the commitment of the sub-
stantial minority makes up in degree for what it lacks in univer-
sality. Moreover, because the 'vanguard' movement or party is
in principle open to all, and those participating in it are dis-
tinguished simply by their greater commitment rather than by
the privilege of birth or property, they can be seen as representa-
tive of the people as a whole, and their activity in the regime's
156 The Legitimation of Power

cause as demonstrative of the continuing support of society at

large.
In contrast to the electoral mode, here popular participation
is divorced from the process whereby office-holders are
appointed, and is directed more towards the execution of policy
at the grassroots than to its formation at the centre. This divorce
explains the frequently cited feature of Soviet-type systems where
a much greater proportion of the population than in liberal
democracies is politically active, its activity, however, being con-
cerned with 'low' rather than 'high' politics (e.g. Bialer, 1980,
pp.185ff.). And the ruling party has a correspondingly different
popular function from parties in electoral systems: not to prepare
leaders and policies for presentation to the electorate, but to
stimulate the mass involvement on which the system depends
both for the administration of policy and for its own legitimation.
Although in encouraging this involvement the party has material
inducements and political privileges to offer, the effectiveness of
participation as a legitimating process depends on the commit-
ment of those involved to a cause over and above that of their
own personal advancement.
Because of the need for continuous mobilisation, leaders who
possess the capacity for arousing mass enthusiasm have a particu-
lar place in this mode of legitimation, as exemplified by such
figures as Lenin, Hitler, Mao, Khomeini. However, the Weber-
ian concept of 'charismatic authority' is more of a hindrance than
a help in analysing this phenomenon, since its assumption that
the source of legitimacy resides in the special qualities of the
leader assigns far too exclusive an importance to the individual,
and leads to fruitless, because unresolvable, disputes about
whether particular leaders possess the indefinable quality of
'charisma' or not. Most of the figures named above achieved
their authority as the focus of popular resistance to a discredited
political order, and as the exponents of a belief system that came
to underpin the rules of its successor. They might embody the
faith in a particularly exemplary fashion, but the ability of the
political order and its legitimacy to survive them indicates that
it is the belief system itself, with its distinctive institutions and
mobilising power, that is crucial, rather than the particular per-
sonality of the individual leader. If Hitler and his Ftihrerstaat
constitutes perhaps an exception here, this only demonstrates
Dimensions of State Legitimacy 157

that leaders play different roles in different circumstances; and

that these need careful differentiation rather than inclusion in a
general category such as charismatic authority, which has sown
more confusion than almost any other in the history of twentieth-
century social science.
Since what is central to popular legitimation here is the
mobilising power of a belief system or cause, rather than the
exercise of choice between different leaders and policies, this
mode of legitimation is associated with monopolistic claims on
the part of the ruling party: the claim to a monopoly of truth in
the realm of doctrine, and a monopoly of organisation in the
sphere of political activity. Whereas the electoral mode can toler-
ate, indeed requires, a plurality of ideas and groupings, in the
mobilisation mode the public expression of alternative ideas,
or even of opposition to official policy, articulated outside the
framework of the ruling party, constitutes a threat to its legit-
imacy, since it challenges the truth claims on which its authority
is based. Such alternatives, while they may be entertained in
private, cannot be allowed to emerge into the public arena, and
have to be blocked in advance of their public expression. The
counterpart to legitimation through mobilisation is thus a system
of surveillance and repression, not as an alternative to legitimacy,
but as its necessary complement.
The mistake frequently made by liberal commentators is to
assume that, because there is no effective electoral choice, and
little freedom of expression and association in mobilisation
regimes, they can therefore have no legitimacy. This would cer-
tainly be true if they depended upon elections for their legitim-
ation; as we have already seen, the restriction of electoral choice
constitutes a characteristic form in which popular legitimation is
weakened in an electoral regime. Such a restriction does not
have the same implications for a mobilisation regime, however.
What is critical here, on the other hand, is any general erosion
of belief in the doctrine or cause that provides the regime with
its justification and its mobilising capacity; without such belief it
becomes a system of privilege and repression devoid of any social
or moral purpose. Although it is more difficult to chart this
process of erosion than that where restrictions are imposed on
electoral choice, nevertheless the cumulative indices of degener-
ation over time are palpable.
158 The Legitimation of Power

What alternatives are possible for a regime whose legitimacy

has been eroded in this way? As will be discussed more fully in
the next chapter, the recent history of Soviet-type systems shows
the clear possibility of a popular relegitimation through a shift
from the mobilisation to the electoral mode, with elections offer-
ing a real choice between candidates and policies rather than
being a mere formality. Such a shift represents nothing short of
a fundamental change in the political system, its rules of office
and the character of its political parties. On the other hand the
reverse process - of the leaders of an electoral system in decline
seeking to relegitimate themselves through a shift to a mobilis-
ation regime- is highly unlikely to prove effective, for the simple
reason that the continuous mobilisation necessary to legitimation
depends upon the vitality of a belief system or cause, and these
are typically the product of popular movements of opposition,
which cannot simply be called up to order from above.

Conclusion

It is worth at this point summarising what has been a long

chapter, so that the wood does not get lost for the trees. With
regard to forms of power, I have argued, legitimacy is at once
most urgent and most problematic in respect of the contemporary
state, in view of its enormous concentration of power on the one
side, and the vulnerability of its rules of power on the other,
lacking as they do any superior legal authority to validate and
enforce them. In considering the different elements necessary to
that legitimacy - legality, the justifiability of rules, expressed
consent - we need to be able to demonstrate the relationship
between underlying principles and the institutions that embody
or give effect to them. It is the ability to give a convincing
account of this relationship that is lacking in most discussions of
political legitimacy, and that this chapter has sought to provide.
At the level of legality it is a relatively commonplace obser-
vation that respect for the rule of law on the part of government
requires an institutional separation of powers and the effective
subordination of the military to civilian control. However, legal-
ity itself is ultimately dependent on the acceptability of the consti-
tutional order, and the justifiability of the rules governing
Dimensions of State Legitimacy 159

appointment to office and the spatial distribution of power. Rules

of appointment (access to office, degree of competition, mode
of removal, etc.) must conform to a source of authority recog-
nised within the society on the one hand, and facilitate rather
than hinder the realisation of a general interest and the acknowl-
edged ends of government, on the other. In the attainment of
both, the spatial distribution of power must correspond to peo-
ple's self-definition of themselves as a distinctive people. Finally,
the realisation of popular consent requires a particular kind of
party and form of popular participation, according to the system's
particular mode of legitimation.
At each point, in other words, it is possible to trace a relation-
ship between a different aspect or principle of legitimacy, and
the manner in which it is realised, or fails to be realised, within
specific institutions. For legality we look in the first instance (but
only in the first instance) to the relationship between the different
branches of the state: legislative, executive and judicial. For rule-
justifiability we need to examine the extent to which the basic
rules of appointment to high office embody an accepted source
of authority, and facilitate the attainment of the acknowledged
ends of government (the 'right' and the 'good'). For legitimation,
or expressed consent, we must consider the form and extent of
political participation, and the way in which the party system is
arranged to give effect to it. Finally, and most crucially, we need
to assess how far these different institutional arrangements are
mutually compatible, in the sense that, in realising one aspect
or dimension of legitimacy, they are at least consistent with, and
do not undermine, another.
Constitutional rules, we could conclude, comprise institutional
arrangements designed not only to ensure a sufficient concen-
tration and organisation of power to make the government of a
given territory possible, but also to make it legitimate. The differ-
ent political institutions, from parties through electoral systems
to arrangements for the division and distribution of power, can
each be evaluated according to the manner and extent to which
they embody or give effect to a particular requirement of legit-
imacy. In doing so in this chapter I have considered them primar-
ily according to the three different dimensions of legitimacy
established in the first half of the book, rather than as political
systems in the round. The next chapter will consider different
160 The Legitimation of Power

types of political system - liberal-democratic, communist, tra-

ditional, theocratic - and seek to identify not only how they
meet the different requirements of legitimacy already outlined,
but what are their distinctive points of vulnerability in turn.
6 Crisis Tendencies of
Political Systems
In the light of the previous chapter's discussion, a political system
can be defined as a set of constitutional rules whose purpose is
not only to effect a particular arrangement of state power, but
also to secure legitimacy for that arrangement, and hence for
those who exercise power under it, in respect of the different
criteria outlined: the maintenance of legality; the derivation from
a recognised source of authority; the satisfaction of a general
interest in regard to the acknowledged ends of government; the
demonstration of consent. Where the previous chapter con-
sidered each of these criteria separately, the present chapter will
explore their mutual relationship within different types of politi-
cal system.
The two main types of political system to be discussed - the
liberal-democratic and the communist - have manifested con-
siderable variation from one country to the next, and it will be
necessary to represent them in their most characteristic form for
purposes of analysis: liberal-democracy as typified in the
advanced capitalist countries since 1945; the communist system
as it evolved in what might be called its classical period from
the death of Stalin to the accession of Gorbachev. Each will be
presented as an abstract type or model, so as to identify its most
essential features, and the internal relationship between them. It
will be assumed that each type possesses a certain internal coher-
ence or consistency, which derives from the requirements implicit
in its legitimating source of authority.
Besides being definable as a 'pure type' for analytical purposes,
the liberal-democratic and communist systems also constitute a
'model' in a different, normative sense: as an ideal for others to
copy or emulate. Indeed, it would seem to be a characteristic

161
162 The Legitimation of Power

feature of legitimate political systems in the modern world, that

they should claim an exemplary quality that demands imitation
by others. Both liberal-democratic and communist systems have
sought imitators or followers outside their respective 'heartlands',
and the competition thus generated between them has deter-
mined much of world politics since 1945. If on the one side the
terminal crisis which the communist systems had entered by 1989
demonstrates only too clearly the deficiencies of the 'classical'
communist type as a desirable or even a viable model, on the
other side the inability of the liberal-democratic type to survive
outside the capitalist heartlands, without deformation or a recur-
rent slide into military dictatorship, suggests that it represents
no unproblematic model either. The aim of the following analysis
will be to identify the chief source of legitimacy weakness or
deficit in each type: to show its crisis tendencies, as they might
be called, not as contingent or unconnected, but as systematically
generated.
The final section of the chapter will consider political systems
in the Islamic world. One reason for including them is that the
attempt to derive political legitimacy from a religious source of
authority not only constitutes a conscious rejection of the exemp-
lary claims of both liberal-democratic and communist models; it
also challenges a basic premise of Western social science which
links 'modernisation' firmly with the process of secularisation.
As I shall argue, the idea of the 'Islamic state' is both inherently
ambiguous and politically contested, and has been used to give
a fa~ade of legitimacy to many regimes that inherently lack it.
However, there are two political systems in the Islamic world
that are of particular interest for any theory of legitimacy, by
virtue of their distinctiveness. One, Saudi Arabia, represents
an almost pure form of traditional legitimacy. The other, Iran,
constitutes a novel version of the theocratic state, combining a
religious principle of authority with consent through mass mobil-
isation. Both claim an exemplary significance, if only within the
region or within Islam, rather than universally; and the compe-
tition between them has constituted a potent additional source
of tension within the Middle East, at least until it became sub-
merged by Iraq's invasion of Kuwait. Both systems require a
special effort of the imagination, particularly on the part of
Westerners, if their legitimating principles are to be made intelli-
Crisis Tendencies of Political Systems 163

gible, and their inherent points of weakness to be accurately

identified (cf. Said, 1981).
In considering the distinctive character of the different political
systems mentioned, and the points at which their legitimacy is
typically most vulnerable to erosion, it will be necessary to take
into account the form of economy with which they are associated,
since economy and polity are so closely intertwined and mutually
determining in the contemporary world. It is not by chance,
for example, that liberal-democratic systems have been typically
associated with a capitalist economy, or communist ones with a
system of extensive public ownership and state planning. Further-
more, when it comes to the actual erosion of a political order's
legitimacy, it is usually changes or problems in the economy that
provide the catalyst, since these often lie beyond the control of
particular governments. On a historical plane, it was the twin
processes of urbanisation and industrialisation that gave the sub-
ordinate classes the mobilising potential which made the restric-
tion of the political community to the ranks of the propertied
ultimately untenable. In the contemporary world, it is the pres-
sures generated by economic change, or the failure to change,
that expose the weak points in a political order, whether it be
in its source of authority or its ability to resolve distributional
conflicts. If we are to talk of the systemic character of legitimacy
crises, therefore, then it is the politico-economic system, rather
than the political system on its own, that should properly be the
subject of our analysis. This will be the procedure adopted in
what follows.

Liberal democracy

In its evolved contemporary form, as exemplified in the advanced

capitalist countries of Western Europe, North America, Aus-
tralia, New Zealand and Japan, this type of politico-economic
system is characterised by the following legitimating features:

1. The principle of popular sovereignty is carried through con-

sistently in electoral terms, in that not only the legislature
but the head of the executive also is chosen directly or
indirectly by ballot, on the basis of universal suffrage. The
164 The Legitimation of Power

formal guarantee of the freedoms of belief, expression and

association, exercised within an autonomous sphere of civil
society, ensures that, in principle, electoral outcomes reflect
the 'popular will', and that public opinion can be brought to
bear on government in a continuous manner. Since all beliefs
are possible that do not challenge the idea of popular sover-
eignty itself, the ultimate test of correct policy can only
be its popular acceptability, electorally determined, not its
conformity to some established body of doctrine. Elections
provide not only the method of appointment to political
office, but also the avenue through which consent to govern-
ment is expressed, though the operation of the majority
principle means that the minority only consents to govern-
ment indirectly, through its participation in the electoral pro-
cess, not by virtue of having chosen it in preference to the
alternatives. The role of political parties in such a system is
to prepare candidates and policies for approval, to focus
electoral choice, and to provide the discipline needed to
secure effective government in the event of being elected.
2. Economic activity is based upon the private ownership of
property and upon private decisions about investment,
employment and consumption within a formally free market,
on the principle that the market constitutes both an arena
for individual choice and an effective mechanism for coordin-
ating the manifold decisions of individuals and firms in the
general interest. However, since the state is itself the ulti-
mate guardian of the public interest, it is necessarily involved
in guaranteeing the conditions for profitable investment and
the public acceptability of economic processes, where the
market on its own cannot. This requirement involves the
state in wide-ranging market interventions and supplement-
ations: in providing the material and immaterial infrastruc-
ture for production; in ensuring a sufficient workforce with
the requisite skills; in remedying the damaging consequences
of production for individuals and the environment; in tem-
pering the distributional inequalities of the market and mod-
erating its cyclical disturbances. In effect, the contemporary
state underwrites the legitimacy of capitalism by ensuring its
social acceptability as well as its legal validity. Moreover,
since any radical challenge to the rights of property, although
Crisis Tendencies of Political Systems 165

in theory compatible with democratic principle, would pro-

voke a crisis in the political system itself, in underwriting
the legitimacy of capitalism the state is indirectly securing a
condition of its own legitimacy also.

This politico-economic system of capitalist democracy is one

that has evolved over a long period of time. In particular, the
relationship between state and economy has developed from the
nineteenth to the late twentieth century under the impact of two
separate processes. One, economically determined, under the
pressure of market competition, has been the shift from what
might be called liberal to oligopolistic capitalism: from small-
scale, low-technology, labour-intensive production to large-scale,
high-technology, capital-intensive industries. Although a new
shift may be occurring with so-called 'post-Fordism', its general
direction has required increasing state intervention through
market-regulating and market-supplementing activities, to ensure
the overall conditions for capitalist profitability. The other, more
politically determined, process has been the development from
limited suffrage to welfare democracy under the pressure of
popular struggles, which has involved the state in modifying the
unequal distributional outcomes of the market, both indirectly,
through providing the legal rights necessary to trade-union
activity, and directly, through taxation policy, transfer payments
and welfare provision. While both the above types of state inter-
vention are necessary to secure the legitimacy of capitalism, the
second, distributional kind, requires a capitalist class able and
willing to concede a sufficient share of the surplus to the working
class to reconcile it to the inequalities of property ownership and
to its own subordinate position at the workplace. Provided this
condition is satisfied, which it has been in the situation of class
compromise and economic growth that has prevailed in the
advanced capitalist countries since 1945, the legitimacy of capital-
ist democracy seems secure.
Not everyone, however, would subscribe to this conclusion.
Most persuasive among the sceptics have been the theorists of
so-called 'legitimation crisis', Jiirgen Habermas and Claus Offe,
who have argued that the contemporary state is unable to fulfil
its required tasks in relation to the capitalist economy, without
encountering severe legitimation problems of its own, problems
166 The Legitimation of Power

which became particularly evident during the course of the 1970s.

An examination of their arguments will enable us to explore
more fully the basis of legitimacy of capitalist democracy
(Habermas, 1973, 1976b; Offe, 1984, 1985).
The argument takes its starting point from the fundamental
'contradiction' of the capitalist system, between the social process
of production and the private appropriation and use of the prod-
uct - simply put, the essentially public activity of production is
dependent upon private ownership and private initiative, so that,
for example, decisions about the balance between investment
and consumption take the form of a conflict between classes,
and so on. This contradiction is not resolved by state intervention
to deal with its consequences, but is simply displaced onto the
state itself, where it manifests itself as incompatible steering
objectives, fiscal crisis and legitimacy deficit.
The idea of 'displacement' here can be elucidated by contrast-
ing the political economy of advanced capitalism with liberal
capitalism on one side, and a planned economy on the other. In
the period of liberal capitalism, when governments undertook
the minimal functions of guaranteeing the security of property
and contract, state legitimacy was protected from crises in the
economy, since these took the form of natural disturbances of
the market, beyond human control. Under a centrally planned
economy, on the other hand, the state cannot avoid responsibility
for any inadequacies or interruptions in production, and eco-
nomic crises are experienced directly and transparently as a
problem of government, with unavoidable consequences for the
legitimacy of the state. The system of advanced capitalism con-
forms to neither of these relatively simple, if also oversimplified,
models. Here the state's intervention to complement the market,
so that its privately determined activities produce socially accept-
able outcomes, creates problems for the state which appear as
autonomous problems of the political domain, quite separated
from the underlying contradiction which is their cause. It is this
complex process, whereby solving one set of problems systemati-
cally produces other problems of a quite different kind at a
different level, that the idea of 'displacement' suggests.
What are these problems for the state? One generic problem,
so it is argued, is a problem of performance: the range and
complexity of tasks the state is required to undertake far outrun
Crisis Tendencies of Political Systems 167

its capacity to fulfil them. Symptoms of this are the ability of

powerful economic groups to veto or nullify interventions made
in the public interest, or the fiscal crisis that regularly afflicts
state finances, as its labour-intensive services prove unable to
match the productivity levels of manufacturing industry. Persist-
ent increases in taxation erode the economic profitability on
which the state itself depends, while the scope for cutting services
is limited by electoral considerations. Underlying these conflict-
ing requirements is a more fundamental contradiction, according
to Offe. The state has to secure and protect the integrity of the
market, while simultaneously intervening in it in the social
interest. It must uphold the logic of market valuation and the
production of commodities for profit, while its own activity fol-
lows the antithetical and subversive principle of provision Clccord-
ing to social need, so, for example, undermining the labour
discipline and achievement orientation on which capitalist pro-
duction depends. The problem of the state's performance, in
other words, is not simply a lack of capacity, but a systematically
generated incoherence in its goals. The basic contradiction of
the capitalist system is reproduced as incompatible steering objec-
tives at the level of the state itself (see Keane, 1984).
Performance problems affect the legitimacy of the state by
undermining its authority, and hence the support on which it can
rely. However, its legitimacy is more directly impaired, according
to Habermas, by the absence of any normative agreement which
might serve to validate the expanded intervention of the state,
and make it acceptable to those it adversely affects. A purely
technocratic definition of the public sphere, Habermas argues, is
impossible when its necessary counter-part of civic privatism -
the definition of citizenship as a purely private status - is itself
disrupted by the intervention of the state in the family, in edu-
cation, and in the value autonomy of the private sphere. On the
other hand, the evolution of capitalism has destroyed the tra-
ditional moral legacy on which it depended to moderate the
unrestrained pursuit of self-interest, and which might validate
state involvement in the economy (e.g. 1976b, pp.75-9). One
area where this erosion is widely acknowledged to be significant
is the sphere of distribution. Now that distribution of the social
product is to such an extent politically determined, any perceived
unfairness cannot be attributed to the impersonal chances of the
168 The Legitimation of Power

market, but demands correction according to a publicly defens-

ible standard of justice; yet there is simply no agreement on
what such a standard might be. The lack of agreement results in
a potentially endless succession of demands on the state from
one sectional group after another. At this point the insufficient
legitimacy exacerbates the steering problems in a mutually rein-
forcing cycle (see Plant, 1982).
The above is an all too brief summary of the main claims of
the 'legitimation crisis' literature. How valid are they? In my
view the idea that the basic contradiction of capitalism is dis-
placed onto the political plane, and constitutes the source of the
latter's recurrent problems, is a highly plausible one. But do
these problems amount to a legitimation crisis, or even to crisis
tendencies (as Habermas is careful to call them), rather than a
set of tensions to be managed, of competing demands between
which a compromise has repeatedly to be struck? Before we can
decide whether recurrent problems have crisis potential, we need
a precise account of what a crisis of legitimacy would actually
amount to. In particular, since it is claimed that the contradic-
tions of capitalism have been displaced onto the state, we need an
account of what would constitute a crisis of the liberal-democratic
political system. It is here that the literature on legitimation crisis
seems to me to be least adequate, because insufficiently specific
about the political domain.
According to the concept of legitimacy I have been arguing
for throughout this book, a legitimacy crisis could be said to
occur when there is a serious threat or challenge to the rules of
power, or a substantial erosion in the beliefs which provide their
justification. In the case of a liberal-democratic order, that means
a threat to the electoral rules and their associated freedoms, or
an erosion of commitment to the idea of popular sovereignty
underpinning them. Agreement about these rules is, in principle,
the only normative consensus that a liberal democracy requires
for its legitimacy. The problems of state performance and incom-
patible steering objectives, on the other hand, which the litera-
ture of legitimation crisis is preoccupied with, may serve to
weaken or undermine the authority of particular governments,
but do not thereby erode support for the rules themselves.
Indeed, as I have argued in the previous chapter, it is one of
the strengths of the electoral process that it enables ineffective
Crisis Tendencies of Political Systems 169

administrations in due course to be dismissed, alternative sol-

utions to be canvassed and tried out, and the authority of govern-
ment to be regularly renewed. Theorems of legitimation crisis,
in other words, do not adequately distinguish between the
authority of governments and the legitimacy of the system;
between the normative agreement necessary to sustain particular
policies and that needed to support the rules of the political
order; between agreement on a substantive criterion of distribu-
tive justice, and agreement on procedures whereby competing
definitions of justice might be resolved through compromise.
The ability of the electoral process to produce changes in state
policy and personnel, and to renew the authority of governments,
is demonstrated by the history of liberal democracies in the
1980s. The decade saw the election of right-wing governments
in many countries, committed to a neo-liberal programme and
ideology that directly addressed the problems of state capacity
identified in the 'legitimation crisis' literature, but who treated
the source of these problems as primarily political rather than
economic or structural, and thus as amenable to political sol-
ution. According to their analysis, the problem of 'government
overload' was not attributable to capitalism per se; it was the
result of a combination of misguided Keynesian theory and irres-
ponsible electioneering drawing governments into areas of eco-
nomic intervention which they were incapable of handling, that
had aroused exaggerated expectations and eroded monetary disci-
pline. The solution was to reestablish the autonomy of the
market and the authority of the state together by means of
a thoroughgoing withdrawal of government from the economy,
through cuts in taxation, the privatisation of public services, and
the reduction in state support to the losers in market competition.
This strategy would leave governments strengthened to concen-
trate on those limited market-supporting activities that only they
could perform: providing security for property and contract,
guaranteeing sound money, setting and policing the rules of com-
petition itself (Hayek, 1976; see King, 1987).
What was surprising about such a programme, involving
reductions in welfare provision and increased levels of unemploy-
ment, was that it proved at all electorally viable. However, its
emphasis on the virtues of individual consumer choice in contrast
to the least attractive aspects of collective action and collective
170 The Legitimation of Power

provision found considerable popular resonance. And the chief

losers from the neo-liberal agenda of rolling back the state were
unorganised groups of the population who were least capable of
mounting any political challenge to it: part-time workers, the
unemployed, the homeless, women forced to care for dependent
relatives at home. Here, right-wing governments took advantage
of a characteristic feature of liberal democracy, whereby interests
that are not organised and articulated in the political domain can
be ignored with impunity. Although injured minorities can prove
the Achilles' heel of democracy, where they are powerful or well
organised, minorities that lack collective voice or presence can
be the object of considerable injustice, without any consequence
for the legitimacy of the state.
Now it cannot be claimed that the neo-liberal strategy, where
it was put into effect, has conclusively solved the underlying
contradictions of advanced capitalism, rather than reducing their
political salience through a reassertion of market forces at the
expense of society's weakest groups. Yet there is nothing to
suggest that liberal democracies cannot zig-zag indefinitely
between a market-oriented solution to the problems of the state,
and a governmental solution to the problems of the market, with
a shifting point of balance struck between the two. Indeed, the
electoral process facilitates such a fluctuating course in the alter-
nation of pfirties between left and right, with in-built self-correc-
tion when the movement in one direction tends to get out of
hand. The rules of the democratic system, in other words, serve
to blunt the contradictory tendencies of the capitalist political
economy by extending their effects over time.
The evidence of the whole period since 1945, in fact, suggests
that the rules of the democratic proces~ provide considerable
resilience in problem-solving in the advanced capitalist states,
once the basic conditions of class compromise have been secured;
and that claims about legitimation crisis constitute an exaggerated
extrapolation from particular difficulties experienced during the
1970s. It does not follow, however, that the possibility can be
ruled out of crisis tendencies developing, or crises occurring,
that are specific to particular states within the liberal-democratic
universe. In Western Europe these have typically occurred when
the management of post-imperial political decline and resurgent
sub-state nationalism has exposed particular weaknesses in the
Crisis Tendencies of Political Systems 171

constitutional arrangements. In the USA, by contrast, it has been

the tensions generated by its neo-imperial role that have most
frequently subjected the political system to stress. In almost every
case, however, the liberal democratic order has shown itself
capable of an internal resolution of the crisis, without any break
threatened in the legality of the constitutional arrangements. It
is the sharp contrast between this experience of liberal democracy
in the advanced capitalist countries, and that of the less
developed world, where legitimation crises can more properly be
described as systemic, that I turn to next.

Liberal democracy in developing countries

The success of the liberal-democratic system in the capitali~t

heartlands, at least since 1945, and its active self-promotion of
its own virtues have made it a prestigious model for others to
copy. Yet its record of achieving or sustaining political legitimacy
outside these heartlands is derisory. There are very few countries
in the less developed world which have been able to sustain a
system of open party competition allied to freedoms of
expression and association for any length of time. Such a system
has usually been brought to an end by military coup, or by
transformation into a one-party state that breaks with the prin-
ciple of free association central to liberal democracy.
Purely cultural explanations for this sorry record, to the effect
that the peoples of these societies lack a commitment to demo-
cratic procedures and the ability to work them, apart from being
patronising, tend to circularity, as I have already argued, since
one of the conditions for a settled commitment to the rules of
party competition is a past history of successful alternation of
parties in power; and it is precisely this that remains elusive. A
more plausible explanation is a situational or structural one:
there are certain politico-economic conditions, typical of most
developing countries, in which the different requirements of legit-
imacy come into conflict with one another, because effective
electoral choice can only be realised at the cost of an intensifi-
cation of social and political divisiQn. In other words, the ability
of liberal democracies to satisfy the legitimating criteria of an
accepted source of authority and expressed popular consent, is
172 The Legitimation of Power

incompatible with the attainment of a recognisable general

interest in the circumstances of most developing countries.
What are these circumstances? At the cost of grossly over-
simplifying the enormous diversity within the so-called Third
World, I would single out two characteristics of underdevelop-
ment which make the attainment of legitimacy particularly diffi-
cult within a liberal-democratic system. The first is an overdevel-
oped state in relation to society, which is in part a consequence
of the leading role the state has come to play in the process of
economic development. The most obvious indication of this is
the size of the state apparatus, and the fact that the public sector
constitutes a, if not the, major site of privileged employment
opportunities and means of social advancement. This, in turn,
has a number of consequences. One is that the competition to
win control of government assumes a special intensity, in view
of the command over appointments and the means of patronage
that it provides, including the denial of such resources to political
opponents. The electoral system has to carry the strain of compe-
tition for scarce economic resources as well as the contest for
political power (Clapham, 1985, ch.3).
A further consequence is that the state's monopolisation of
talent and social initiative hinders the development of an inde-
pendent 'civil society', with autonomous associations that have
an accepted place in the political process. The relation between
state and society assumes a correspondingly paternalist form,
which has great difficulty in accommodating political demands
and initiatives that are not prompted or controlled from above
(Thomas, 1984; Mouzelis, 1986). It is not so much the character
of the state as such, it should be emphasised, as its relation to
society that is important here. The relative overdevelopment of
the state renders exceedingly difficult the autonomous organis-
ation of civil society and its acceptance as a normal part of the
political process, which is a necessary concomitant of consent
expressed through electoral choice.
The second relevant feature of underdevelopment is to be
found in the sharp economic inequalities and dislocations that
characterise the process of capitalist penetration of traditional
economic relationships, and the early stages of industrialisation.
Although these are conditions that the older capitalist economies
themselves underwent, they are intensified and prolonged by the
Crisis Tendencies of Political Systems 173

subordinate position in which the late-developers stand to the

capitalist 'centre'. The form of their industrialisation is one in
which the most advanced sectors of production, typically foreign-
owned, are detached from the other sectors, and make no posi-
tive contribution to their development. In short, developing
economies tend to be ones which experience the sharp inequalit-
ies and dislocations of the industrialisation process on a capitalist
basis, without generating the self-sustaining economic expansion
that might enable its benefits in due course to be spread more
widely (Cardoso and Faletto, 1978; Szentes, 1983; Szeftel, 1987).
From the standpoint of political legitimacy, the most significant
contrast with the early industrialisers is that the latter experi-
enced the process of industrialisation, when capitalist legitimacy
is at its weakest, under a political system based upon a restricted
suffrage. In other words, the legitimacy of the state did not
depend upon securing the consent of those most adversely
affected by the economic transformations taking place. The suf-
frage was only extended to the propertyless at the point when
they were no longer considered a threat to the rights of property,
and when the strains of the early period of industrialisation had
been overcome. The late industrialisers, on the other hand, have
no such restriction available to them, if they are to meet the
requirements of consent through electoral choice; and they are
thus fully exposed to the potential contradiction between the
costs that economic transformation inflicts upon the subordinate
classes, and the electoral power that the political system formally
assigns to them.
Whether a competitive electoral system can survive at all in
such circumstances depends very largely on the character of pol-
itical parties, and on the extent to which they cut across rather
than coincide with potential divisions within society. A common
form of party is the hierarchically structured patronage party, in
which subordinate classes are organised by their social superiors,
and which is directed more towards particularistic benefits than
to programmatic demands that might pose a challenge to the
interests of property or the powers of the state. In rural areas
such parties may be organised as the political extension of tra-
ditional 'clientelist' networks, whereby casting one's vote
becomes another form of service to the local patron, landowner
or tribal chief (Weingrod, 1968; Schmidt, 1977, pt.2). Although
174 The Legitimation of Power

such networks can be recreated in urban contexts, patronage

parties there will be organised around more modern institutions
such as trade unions, which, however, tend to enjoy much less
autonomy than their counterparts in the West, being initiated and
controlled from above along paternalist lines (Mouzelis, 1986).
Wherever such parties exist, we could speak of a system of
managed consent, or guided democracy, in which the organis-
ation of patronage within established vertical relationships
ensures that the exercise of electoral choice by subordinate
classes poses no threat to the dominant powers within society or
state.
However, a system of competition between hierarchically
structured patronage parties is vulnerable to two different forms
of evolution, according to the context. Where class formation is
well developed, as in much of Asia and Latin America, the
formal recognition of freedoms of expression and association
periodically allows new types of party organisation or electoral
programme to develop, which escape beyond the limits tolerable
to dominant classes or political elites. It is often said by commen-
tators on clientelism, for example, that clientelist relations are
not incompatible with the existence of a potentially conflictual
class structure. What they hinder is the emergence of forms of
consciousness and organisation on the part of subordinate classes
which would bring that conflict into the open (Flynn, 1974; Scott
and Kerkvliet, 1977). (In Marxist terms, they hinder a class 'in
itself' from becoming a class 'for itself'.) However, when eco-
nomic changes weaken the hold of traditional relationships, or
severely limit the benefits available for local distribution by ruling
parties, then more class-based organisations can emerge with
programmes that directly challenge the powers or prerogatives
of property. At this point the latter can only be preserved intact
by the use of legal or extra-legal repression to limit the freedom
of association and electoral choice, or else to abolish it alto-
gether. The better organised the movement or the demands of
the subordinate classes, the more extensive the repression will
have to be to achieve its purpose (O'Donnell, 1973; J.C. Scott,
1976, ch.7).
In many African countries, on the other hand, it is not so
much the erosion of a hierarchically organised party system under
the pressure of class conflict that has undermined electoral
Crisis Tendencies of Political Systems 175

democracy, as the intensification of regional or ethnic divisions

through the competition for control over state patronage. While
vertically integrated parties are able to moderate the political
salience of vertical or class-based divisions, they only serve to
intensify those divisions that are laterally or geographically
based. The process of competition for access to economic oppor-
tunities, through control of the state, itself heightens the signifi-
cance of the ethnic or regional connections through which oppor-
tunities for business or upward mobility can be organised
(Szeftel, 1983; Joseph, 1987). As I argued in the previous chap-
ter, where the cost to those excluded from the benefits of state
patronage is high, and the excluded comprise identifiable sections
of the population, corruption ceases to be something reluctantly
tolerated, and threatens to discredit the political proce~s alto-
gether. In such circumstances, military intervention to abolish
the rule of corrupt politicians can claim the merit of 'clean hands'
as well as salvation of the general interest; and it is typically
accompanied by much less repression than where a mass move-
ment of the subordinate classes has to be suppressed.
In view of the social divisiveness of party competition that
forms the typical counterpart to freedoms of association and
popular consent in liberal democracies in the Third World, it is
not surprising that those systems are most durable in which there
exists a dominant party enjoying a marked electoral ascendancy
and long continuity in office, such as those of India or Mexico
(Randall, 1988). Such parties have successfully incorporated the
major social forces within them, i.e. they are truly national
parties, while their continuity in office ensures the effective sub-
ordination of state officials, and in particular the military, to
political control. Such features are of course more securely
guaranteed in a formal one-party state, such as those in East
Africa, which may also provide the opportunity for limited elec-
toral choice between candidates, if not between parties or poli-
cies. In the absence of any competition from other parties, how-
ever, these ruling parties tend to become simply an extension of
the state administration, and lose their mobilising capacity (Tor-
doff, 1984, ch.5). Although the stability of these systems may
look impressive in comparison witb their competitive counter-
parts, the absence of legally recognised freedoms of association
and political organisation constitutes a substantial departure from
176 The Legitimation of Power

the liberal-democratic model, and from the requirements of

popular consent through electoral choice.
It must be concluded, then, that the evidence from the Third
World casts doubt on the claim of the liberal-democratic model
to provide an effective solution to the problems of political legit-
imacy in the circumstances of underdevelopment. In particular,
its inherent tension between the requirements of popular author-
ity and popular consent on the one hand, and the attainment of
a general interest on the other, proves unresolvable in these
circumstances. The evidence also confirms doubts about the exist-
ence of any generally positive connection between a capitalist
economy and liberal democracy, such as has most recently been
reasserted in the context of the changes in Eastern Europe. I
have argued that we should consider the political and economic
systems together, since the legitimacy of each is bound up with
the conditions of the other. By way of conclusion, therefore, it
will be useful to consider more systematically the interrelation-
ship between the two elements in the portmanteau concept of
'capitalist democracy'.
Proponents of a positive connection between capitalism and
liberal democracy have usually argued that there are principles
common to both, such as the freedoms of association, movement,
ideas and so on, which have a mutually reinforcing effect on the
legitimacy of each domain (e.g. Friedman, 1962). Thus the free
market in commodities is paralleled by the electoral market in
programmes and policies, and success in each is measured by
the number of people who can be attracted to support a given
product or programme. There is a democratic test of validity, in
other words, which applies equally to goods as to policies, to
firms as to parties, and which in turn requires open competition
in the development and promotion of new products or policies.
At its strongest it is argued that a free market in commodities
is a necessary condition for the political freedoms intrinsic to a
democratic order. A weaker version points to a congruence
between the two that is conducive to the maintenance of both.
Now although there is undoubted validity in the idea of mutu-
ally supporting principles in different domains, the above argu-
ment only tells part of the story. In particular, it treats economy
and state simply as parallel and coexisting spheres, rather than
as systematically interacting on the basis of different logics. As
Crisis Tendencies of Political Systems 177

a consequence, it overlooks the problems that a capitalist econ-

omy poses for liberal democracy. Among these problems two
stand out from the experience of developing countries.
First, the driving force of a capitalist economy is the pursuit
and maximisation of economic self-interest, from which a general
interest emerges only as unintended consequence, via the 'hidden
hand' of the market. The state, in contrast, must act as the
conscious agent of a public interest, and is required to intervene
both to provide the basic regulatory framework for the market,
and to apply corrective action where the market fails to satisfy
the public interest on its own. In the context of a contemporary
developing economy, in particular, these interventions have to
be substantial, to ensure that the market power of international
capital does not undermine national priorities and interests, and
that limited indigenous resources are utilised in the most effective
way. Yet the greater the intervention, the greater the potential
scope for the state to be colonised by the self-interest of the
market place, to be penetrated by the forces it seeks to direct
and regulate, and to become simply an extension of the particular
interests and dominant motivations of the economic sphere. Party
competition, as I suggested earlier, is more likely to intensify
than moderate this process; indeed, parties can become one of
the main instruments whereby the competition of the market
place is systematically transposed into the political domain. In
such conditions, therefore, far from the common principles of
economy and polity mutually supporting one another, the logic
of the market can undermine the distinctiveness and integrity of
the public sphere, which are essential to its legitimacy as guardian
of the public interest.
Secondly, the market, far from being a uniformly democratic
device, tends if left on its own to intensify economic inequalities
in a manner that frustrates the principle of political equality
central to a democratic order. This inegalitarian process can only
be moderated by political intervention, which may take broadly
one of two forms. Redistribution can be effected from above,
through a paternalist or authoritarian state dispensing benefits in
a way designed to pre-empt or frustrate the autonomous action
of the disadvantaged themselves. However, where the political
system acknowledges at least the formal freedom of association,
such action cannot be prevented indefinitely. The other mode of
178 The Legitimation of Power

economic redistribution is through the operation of autonomous

institutions of a subordinate class (trade union, political party),
and on the basis of a conscious class compromise with the econ-
omically privileged. This mode, however, requires a dominant
class with sufficient economic room to make such a compromise,
and also willing to coexist with autonomous institutions that
involve some diminution of its own power. Neither condition is
easy to achieve in the circumstances of developing countries. As
a result, attempts to make the difficult transition from one mode
to the other, which is essential to the long term viability of a
liberal-democratic system, become blocked or frustrated.
What we have here, in fact, is precisely those features which
theorists of 'legitimation crisis' have pointed to - the capitalist
erosion of traditional values that might inhibit or restrain the
pursuit of self-interest; the absence of any agreed conception of
distributive justice, or of compromise between competing concep-
tions, that might validate the state's intervention in the market
- only proving much more critical to political legitimacy in
developing countries than in the advanced capitalist states. The
reasons for this, as I have suggested, are more situational or
structural than cultural, and lie in the contradictoriness of the
tasks the development state is required to perform: to satisfy the
requirements of a developing indigenous entrepreneurial class
and the expectations of subordinate classes simultaneously, in
the context of a world market that is systematically tilted against
both. Such contradictions are much more acute than anything
confronting the state in the advanced capitalist economies. For
the principles of the market and liberal democracy to be mutually
supportive, in other words, rather than mutually destructive,
requires specially favourable politico-economic conditions to
secure. And if it could be shown that the economic development
of the advanced capitalist societies was systematically connected
to the underdevelopment of the colonial and post-colonial societ-
ies of the Third World, then we should have to conclude, further,
that the security of liberal democracy in the former was depen-
dent upon the conditions making for its insecurity in the latter.
In other words, it would be doubtful whether the successful
legitimacy of capitalist democracy could ever be other than a
localised, not a universal, one.
Crisis Tendencies of Political Systems 179

The communist model

Anyone who writes about the communist or Soviet system today

has to begin from the obvious fact that it entered a phase of
deep crisis during the 1980s; that the programme of radical
change initiated in the Soviet Union by Gorbachev to resolve
the crisis prepared the ground for the spectacular collapse of the
communist regimes across Eastern Europe in 1989; and that the
brutal suppression of the Tiananmen Square demonstrations in
China radically altered the basis of communist rule in that
country even as it sought to immunise the political system from
the pressures of change. In other words, the communist system
in the form we have known it since the end of the second world
war must now be regarded as a historical phenomenon, in the
sense that, although it has not been entirely superseded, it no
longer provides a viable alternative model to liberal democracy,
owing to its loss of legitimacy in its own heartlands.
How are we to explain the process of change that reached
such a dramatic denoument in 1989? Was it the product of a
classic legitimation crisis, in which the limits of system develop-
ment had been reached, and no further evolution was possible
without a transformation in its rules of organisation and under-
lying principles of legitimacy? Or did it reveal, as some would
argue, that the communist regimes had never enjoyed any legit-
imacy at all, and had simply run out of stratagems for managing
recalcitrant populations by a mixture of inducement and
coercion? Was it a crisis of legitimacy, or one borne of chronic
illegitimacy? How we answer this question depends upon our
analysis of the communist system in what could be called its
'classical' form prior to Gorbachev. This, far from uncontentious,
task must form the starting point of any discussion.
To begin with, it will be important to draw a distinction
between those communist systems that were established as the
result of an indigenous revolution, such as the USSR, Yugosla-
via, China, Vietnam or Cuba, and those that were imposed by
conquest or kept in place by the threat of military intervention,
as in most of Eastern Europe (Denitch, 1976; Rigby and Feher,
1982). The former, as the product of popular struggles against a
discredited regime, which were also national struggles against an
external enemy, have always derived part of their legitimacy
180 The Legitimation of Power

from the sufferings and exploits of the revolutionary period.

The communist systems of Eastern Europe, in contrast, were
associated with the infringement of national autonomy, and the
externally dictated suppression of any spontaneous evolution in
their political process - most dramatically in Hungary in 1956,
Czechoslovakia in 1968, and Poland in 1981. A system associated
with national subordination is prima facie an unlikely candidate
to enjoy widespread moral authority (Lewis, 1982). Even with
the Eastern European regimes, however, the form that their
recurrent crises took is only intelligible in terms of the generic
features common to communist systems, and the principles from
which they have sought to derive their legitimacy. The commu-
nist model as it developed in its classical form in the Soviet
Union must therefore provide the reference point for understand-
ing these regimes as well.
At this point a further source of difficulty lies in the persistent
disagreement that exists among those who have written on the
legitimacy of the Soviet model, as to what its basis might be.
Those who have started from a Weberian typology have been
unable to agree whether its source of legitimacy was primarily
charismatic, focused upon an exceptional leader such as Stalin,
Mao, Tito or Castro (Gill, 1982); or traditional, deriving from the
revolutionary inheritance or the persistence of longer-standing
national traditions (Lane, 1984); or goal-rational, based upon
progress towards an ideal communist society (Rigby, 1980, 1982);
or some combination of all three (Heller, 1982). Even those who
have not been constrained by the Weberian straight-jacket have
disagreed as to whether the legitimating basis was doctrinal and
scientific, located in the official doctrine of Marxism-Leninism
(Brunner, 1982); or unofficial and paternalistic, conforming to
the authoritarian family and personality type characteristic of an
enduring social culture (Feher, 1982); or again whether legit-
imacy was something that concerned elites alone (Bialer, 1980;
Rigby, 1982), or had a mass dimension (Lane, 1984; Lomax,
1984). Such confusion is less an index of fundamental disagree-
ment about the nature of the Soviet system in its classical period
than a demonstration of the disarray attending the concept of
legitimacy in political science, and the absence of any agreed
theoretical foundation that might enable us to distinguish legitim-
Crisis Tendencies of Political Systems 181

ating elements from other more derivative features of a political

system (Ludz, 1979; Pakulski, 1986).
My own analysis will start from the premise that the legitimacy
of any political system is intimately linked with its basic organis-
ing principles and rules of power, which it serves to validate in
the different ways already distinguished; and that, as with the
system of capitalist democracy, we should treat the polity and
economy as two parts of an interrelated whole. Starting from
this premise I shall identify the legitimating elements of the
Soviet model as it existed prior to Gorbachev, and consider
possible explanations for the legitimation crisis of the 1980s. The
fact that at the present time of writing the precise shape of the
future evolution in the USSR remains unclear is immaterial,
since we can at least be certain that a return to the old system
as a legitimate politico-economic order is no longer possible.
Although we can therefore no longer talk of it in the present
tense as a contemporary system, the advantage of hindsight
which the close of an epoch affords should help us to identify
its legitimating features with greater clarity. I take these to be
as follows:

1. The communist system of rule in its classical form derived

its authority from a dual source. First was from the people;
here it stood in the mainstream of legitimacy from the French
revolution onwards. However, it was the people only as
defined in a certain way - as the working class or classes -
and as represented uniquely by the Communist party in its
role as 'vanguard'. This definition of the people and limi-
tation on their representation was justified in turn by refer-
ence to a second source of authority, which qualified if not
actually displaced the first: the doctrine of Marxism-Lenin-
ism, and its claims to exclusive knowledge of the conditions
of evolution towards a communist future. It was this doctrine
that privileged the working class as the most progressive
social class, with its interests identical to those of society as
a whole, and the Communist party as the exclusive
interpreter and representative of those interests. Popular
consent in this system was not expressed primarily through
elections, which lacked the alternatives necessary for effec-
tive choice, and were largely propagandistic in purpose. It
182 The Legitimation of Power

was expressed rather through the extent of mass activism at

the grass roots, which provided a continuous demonstration
of popular commitment to the party's cause, and served to
validate its claim to the leading role in society. As explained
in the previous chapter, in this 'mobilisation' model the
expression of popular consent is divorced from the process
of appointment to office and choice of policy alternatives,
which remain an elite prerogative in accordance with the
superior knowledge or capacities to which they lay claim.

From these legitimating elements of the polity a number of

features can be seen to follow, which are relevant to the disputes
in the literature mentioned above. First, whereas under liberal
democracy the source of authority is identified by reference to
a simple principle - that of popular sovereignty - and a straight-
forward electoral device for ascertaining the people's will; in
the communist system, authority was primarily located in an
elaborated body of doctrine, Marxism-Leninism. Hence the
reproduction of that doctrine, the protection of its orthodoxy,
the identification of authentic interpreters, etc. assumed a central
place in the organisation of power. Moreover, since the test of
'correct' policies lay ultimately not in what the people, freely
organised, would accept, but in what the party hierarchy,
informed by scientific doctrine, would determine, the relation
between rulers and ruled necessarily assumed a paternalist
character, as it does wherever it rests on the claim to superior
knowledge. The doctrinal source of authority determined the
paternalist character of rule, rather than being in conflict with
it.
Secondly, the process of legitimation operated at both elite
and mass levels, albeit in different ways; and the party was the
instrument of both. At the elite level, a life-time's socialisation
into the party's norms, and the careful filtering out of non-
conformists, guaranteed loyalty to the system and ensured the
necessary elite cohesion in face of struggles for succession and
position, as well as guaranteeing the subordination of the state
and military apparatus to the party. At the mass level, the party
mobilised the grass-roots activity which provided the continuous
evidence of popular commitment to its rule, and which served
as the surrogate consent of the population as a whole.
Crisis Tendencies of Political Systems 183

Thirdly, while the expression of consent which confirmed the

party's claim to rule depended, on the one side, upon continuous
popular activity in its cause, it also required, on the other, that
any divergent opinions which might challenge the party's mon-
opoly of representation should not reach the public domain.
Unlike in liberal democracies, where public opposition to govern-
ment policy is validated by the basic rules of free association, in
a communist system any public dissent expressed outside the
party, especially by the working class, had a delegitimating conse-
quence for the system as a whole, since it challenged the principle
of the party's leading role and its monopolistic claim on represen-
tation. The systematic obstruction of an alternative or indepen-
dent public opinion and the means to its organisation, therefore,
through the state's monopoly of communication and its apparatus
of surveillance, constituted a necessary feature of the system's
legitimacy, rather than an alternative to it. The state's demobilis-
ation of dissent, if need be by illegal or extra-legal means, formed
the necessary counter-part to the party's mobilisation of consent.
If, then, legitimacy and suppression went hand-in-hand in this
manner, it helps explain why commentators have had such diffi-
culty in deciding whether communist systems had any legitimacy
at all. In determining this question, we need to draw a distinction
between the everyday processes geared to preventing the emerg-
ence of dissent, which themselves operated in secret, and the
highly visible and dramatic repression of a mass protest that had
somehow managed to escape the constraining effects of demobil-
isation, and had emerged into the public domain. The former
could be regarded as systemically normal, and was a necessary
counterpart to its legitimation. The latter was abnormal - the
response to a process of public delegitimation, which only further
compounded it. It was one of the system's weaknesses, however,
that the first, everyday, mode of repression constituted a recur-
rent source of grievance; and that, in stifling any expression of
dissent outside accepted party channels, it ensured its emergence
with all the more force when it did eventually break out.

2. The economy in the classical form of communist system was

organised on the basis of public ownership and centralised
planning, on the grounds that only in this way could the
systematic exploitation and unemployment that the working
184 The Legitimation of Power

class suffers under capitalism be overcome; and that, since

determining investment priorities and the distribution of the
social product are matters of essential public interest, they
should be consciously decided by political authority rather
than left to the sway of private interests and the uncertain
outcomes of the market. These organising principles, which
derived their credibility from capitalism's most vulnerable
features, especially as exemplified during the 1930s, com-
prised a part of what was understood by 'socialism'. The
other part was conceived of dynamically: as a process of
economic development towards a future society in which
material needs would be met for all, and the 'all-round devel-
opment of human powers' would become possible. Progress
towards this goal provided a necessary confirmation of the
validity of socialism's organising principles, and of the histori-
cal schema that lay at the core of Marxist-Leninist doctrine.

The system of central planning through administrative control

proved itself capable of developing an industrial base and the
public services necessary to support it, such as transport, edu-
cation, health, scientific research, and so on. It laid the foun-
dation for impressive industrial and economic development. But
it also showed itself chronically incapable of producing food and
consumer goods in sufficient amounts, variety or quality to satisfy
the basic needs of the population. The reasons for this are intrin-
sic to a system of centralised planning itself, and are located in
its inability to ensure effective lateral coordination between firms
and economic sectors on the basis of vertical command. A 'com-
mand' economy stifles initiative and self-determination at the
level of the enterprise, without, however, being able to secure
the information or control the responses necessary to make the
plan work, much less provide corrections when it fails to do so.
This basic defect proves increasingly evident as the command
economy moves from an 'extensive' phase of development,
characterised by the production of heavy goods and expanded
utilisation of labour, to an 'intensive' phase, where the provision
of more sophisticated consumer goods and improvements in pro-
ductivity are the key (Brus, 1972; Nove, 1983).
This chronic problem of the command economy in the commu-
nist systems gave rise, on the one hand, to a parallel unofficial
Crisis Tendencies of Political Systems 185

economy, and, on the other, to an expansion of the apparatus

of administrative supervision, as it sought in vain to control the
uncontrollable. The bureaucratic apparatus itself constituted the
main obstacle to reform in a more market-oriented or decent-
ralised direction, since such a reform would deprive it of its chief
raison d'etre. Here lay a systemic contradiction more acute than
anything under capitalism - between the goals of the system and
its basic principles of organisation - which was experienced as
an increasingly acute conflict between the economic needs of
society and the interests of the administrative apparatus
employed to realise them (Bahro, 1978; Feher et al., 1983).
The problem at issue here was not just a matter of economic
performance, but of its consequences for the belief system of
socialism, which played such a central part in both economy and
polity. The communist model was publicly dependent upon a
uniformity of belief system to a degree that capitalist democracies
are not. In the latter, a belief in individual liberty, popular
sovereignty, etc., is needed to underpin the rules of political and
economic organisation. But these beliefs constitute a background
condition, as it were, against which a pluralism of particular and
competing beliefs occupy the foreground. Nor are they
threatened by a retreat of the population into a privatised exist-
ence of purely personal motivations, which indeed they serve to
justify.
In the communist model, in contrast, beliefs played a continu-
ously public role. This was not just a question of the status of
Marxism-Leninism as the official doctrine, with its exclusive
claim to truth. It was also that society was organised for the
pursuit of a collective purpose - the realisation of socialism -
and belief in the validity of that purpose was necessary not only
to the legitimacy of its rules of power, but to key motivations
also, at a number of levels. In the absence of procedures of
public accountability, this belief was necessary to ensure a mini-
mum of integrity in the holders of power. In the absence of
electoral consent, it was necessary to the mobilisation of a mass
base for the party, with any degree of commitment. Because of
the deficiency of material incentives for economic production (or
the disincentive of bankruptcy or unemployment) it was neces-
sary to the work motivations of the population at large.
Without the belief in the validity of a collective purpose, in
186 The Legitimation of Power

short, the communist model lost its vital stimulus; it became

reduced to a pursuit of private interests destined to frustration,
and a structure of elite prerogatives and civil restrictions and
harassments devoid of any public justification. Yet it was pre-
cisely the system's own evolution that undermined this belief, as
the contradiction became more apparent between the societal
goal, of expanded human powers and fulfilment of needs, and
the framework of bureaucratic planning fashioned to realise it;
between the socialist end and the socialist means, as these had
come to be defined within Marxist-Leninist doctrine. What we
have here is the classic pattern of a 'legitimation crisis', in which
the evolution of the system brings it up against the limits of its
own organising principles, and in so doing erodes the beliefs that
are necessary to sustain them.
At the same time the course of economic development also
produced the agents of a new politico-economic order in increas-
ing strata of educated personnel, for whom acting upon com-
mands from above proved less and less tolerable. Most visible
of these were the dissident intelligentsia, to whom the decay of
communism was not a failure of economic system merely, but a
moral bankruptcy exposed in the party's distortion of truth, and
in the gulf between official rhetoric and the reality of everyday
experience (e.g. Havel, 1987). Yet it was only when the
expression of dissent extended to the working classes, and to a
public delegitimation by those whose interests the regime claimed
to represent, that the underlying contradictions and legitimacy
weakness of the system could develop into a full scale crisis of
political authority.
The process of disenchantment described above ran its course
much more rapidly in the Eastern European countries, where
the communist system was associated with national subordination
rather than with the prestige of leadership of a world movement,
and the achievement of great power status, as in the Soviet
Union. That much is evident from the history of repeated upris-
ings against communist rule, which could only be maintained in
being by military intervention, from Hungary in 1956 through to
Poland in 1981. Although the Czech situation in 1968 was differ-
ent, in that the reforms initiated by the party enabled it to stem
its own crisis of legitimacy, the invasion which cut them short
demonstrated the gulf that existed between the archetypal Soviet
Crisis Tendencies of Political Systems 187

model that was reimposed, and the direction that the system
would have taken under the impetus of an autonomous, nation-
ally determined, evolution. The fact that a general recognition
of the limits of Soviet tolerance became thereafter a self-imposed
limitation on demands for political change in Eastern Europe
cannot be taken as evidence of legitimacy, i.e. of an acceptance
of the rules of power as justifiable; only of a prudent recognition
of the conditions for survival in the face of superior force. Any
doubt on this score was resolved by the dramatic collapse of the
communist regimes during 1989, once Gorbachev's renunciation
of the Brezhnev doctrine of intervention was proved to be seri-
ous. The rapidity of the collapse provided a copybook demon-
stration of the fate of political systems lacking moral authority,
once they are unable or unwilling to suppress their populations
by force.
The policy of renunciation which set this process in train was
prompted by the internal impasse that the communist model had
reached in the Soviet Union itself, compounded by the costs of
sustaining a great power role in competition with the USA. At
the heart of the impasse was the impossibility of improving eco-
nomic performance within a centralised planning regime, and the
necessity of a radical shift to a system of market coordination
and a more consumer-oriented economy. However such a shift
in turn could not be carried through, as Gorbachev clearly recog-
nised, without a fundamental change in the rules of the political
system and its underlying principles of legitimacy.
The reasons for the necessary linkage between the economic
and the political dimensions of reform in the Soviet Union can
be judged from the past history of economic reform programmes
in Eastern Europe (McFarlane, 1984). Without a clear mandate
from society, such as competitive elections would provide, the
economic reform process of the communist system has proved
liable to obstruction at the hands of the party and state appar-
atus, whose rationale as well as privileges depend upon its own
administrative control over the economy. And without a more
secure legitimacy, the government has been unable to sustain the
short-run costs involved in a market-oriented reform, since, to
paraphrase Jaruzelski, only a government that enjoys public con-
fidence can demand sacrifices. The result has been that, as soon
as reforms have run into difficulties, the conservatives within the
188 The Legitimation of Power

party have been able to play on popular discontent to secure

their reversal. On this evidence, economic reform could only
become irreversible if the reformers could secure a basis of legit-
imacy that rendered them immune from manoeuvrings within the
party hierarchy, and a popular mandate that would enable the
reform process to withstand the short-term costs and dislocations
it would impose upon the working class. And that necessitated
a shift from a mobilisation to an electoral mode of legitimation.
At the institutional level such a shift in the basis of political
legitimacy has been relatively simple to effect, since the commu-
nist states already had an apparatus of elections, parliamentary
assemblies and so on, ready to take on a quite new significance
once they were released from the processes of pre-selection and
coordination by the party hierarchy. But the significance of the
shift has been profound, since the guarantee of effective choice
necessary to electoral legitimation has opened up the sphere of
high politics to popular influence, and signalled the abandonment
of doctrinal orthodoxy as the basic principle of political cohesion.
The very rapidity with which the old political order has unravel-
led as a consequence has brought serious problems in its train.
Of these problems, the most obvious is that the process of
democratisation initiated by Gorbachev from above has called
forth responses from below that reach far beyond his own pro-
grammatic aims. The national conflicts in the republics, and the
demands for secession, threaten the spatial integrity of the Soviet
state. And the dramatic eclipse of the Communist Party, not
merely as the ruling party but as a serious electoral force, in the
popular revolutions of Eastern Europe poses a serious question
mark over its future in the Soviet Union as well. However much
the party may be needed there to provide a basis of cohesion
and continuity in the difficult process of 'perestroika', the logic
of electoral legitimacy, with freedom of association and a choice
between policies as well as candidates, leads powerfully in the
direction of a multi-party system. Whether the evolving political
order can combine the open articulation of social conflicts at the
highest level of the state, with the unity of direction needed to
implement the tough decisions involved in the marketisation of
the economy, must remain an open question.
A second major problem concerns the very different time
scales of political and economic reform; of 'glasnost' and 'peres-
Crisis Tendencies of Political Systems 189

troika' respectively. Removing controls on the flow of infor-

mation, on the expression of ideas and on political organisation
produces immediate and dramatic effects, whereas economic
reorganisation is a slow and painful process that takes time to
produce the results that provide its justification. The danger is
that the strains and dislocations of the economic process will
undermine the support necessary to sustain the new political
order, and give its opponents powerful ammunition to discredit
it. Although the cost of reversing the process of democratisation
in the Soviet Union would be enormous, yet the possibility that
social and economic dislocation may become so deep as to pro-
vide the occasion and justification for a military or 'Bonapartist'
solution to the political situation cannot be discounted. No
society in history has been called on to manage the end of
empire, the marketisation of its economy and the democratis-
ation of its political system simultaneously.
Whatever the precise outcome of the process of change
initiated in the Soviet Union by Gorbachev, the disintegration
of the communist model as a legitimate politico-economic order
marks the close of an epoch in contemporary state systems.
Does it follow, therefore, as Francis Fukuyama has argued, that
capitalist democracy is left holding the field, not merely as the
only legitimate form of contemporary system, but as the ultimate
embodiment of human aspirations, as the 'end of history'? (Fuku-
yama, 1989).
Let me try to answer this question by recalling two potent
images from the revolutions of 1989, both involving mass action
in the streets. The first image is that of the popular demon-
strations in Leipzig, Prague and elsewhere: of people collectively
calling their rulers to account, and demanding a form of govern-
ment in which they had a say in the choice of leaders and the
determination of policy. These actions demonstrate the power of
the principle of popular sovereignty to disrupt those institutional
arrangements that limit or frustrate its realisation, once the
beliefs that support them have become eroded in the population
at large. The communist model in its indigenous revolutionary
form may be seen initially to have been politically progressive
in comparison with the dictatorships and traditional systems it
replaced, in that it established the principle of equal opportunity
and extended the political community to include the whole popu-
190 The Legitimation of Power

lation. Yet its doctrinal base in Marxism-Leninism perpetuated a

paternalist relationship between rulers and society, which became
vulnerable once the credibility of its belief system was eroded,
and the people claimed the right to decide their interests for
themselves. Here my own analysis of the irreversibility of the
principle of popular sovereignty in its unqualified form gives
some support to Fukuyama's claim.
The other image that stays in the memory is of a different form
of mass action: the East German population pouring through the
Berlin wall with their shopping bags to share in the consumer
abundance of the West. This reminds us that the communist
model was not undermined by internal processes alone, but by
the competition with Western capitalism that it had itself
endorsed, and which it had been able to withstand in the
industrialising phase of its development, but was hopelessly
mismatched for in the mass consumption phase of advanced
capitalism (Bauman, 1990). At first sight the image of the East
German consumer invasion endorses the idea of a positive link
between the democracy of the market and the democratisation
of the political order. However, we have also subsequently wit-
nessed the effects of exposure to the international market on the
productive base of the East European economies: the other face
of capitalism as a disruptive force, as harbinger of unemployment
and intensifier of inequality, both within and between countries.
This other side of capitalism, which I discussed earlier, exposes
the limitations of Fukuyama's perspective. Liberal democracy
could only achieve a universal significance, not to mention a final
one, if the conditions for its realisation could be universalised;
and that would require a different kind of international economic
order from any that capitalism has so far produced.
The dual character of capitalism as an attractive and disruptive
force could stand as a paradigm for the impact of external factors
in general upon the internal legitimacy of states: for the attraction
(and repulsion) of normative models on the one side, and the
intrusiveness of empirical powers on the other. Both are evident
in the examples from the Islamic world to be considered in the
final section.
Crisis Tendencies of Political Systems 191

Types of Islamic state

One reason for including a discussion of Islam at this point,

however cursory it must be, is that the revival in the political
significance of Islam over the past two decades has provided the
most direct challenge to a longstanding presupposition of West-
ern social science, to the effect that political legitimacy in the
contemporary world can only be securely grounded on a secular
basis. Central to this presupposition is the belief that secularis-
ation is an essential feature of modernisation, and that modernis-
ation itself is a multi-faceted process whose elements each
provide a necessary complement or support to one another.
These assumptions are directly challenged by the recent Islamic
revival, which emerged as the unexpected product of modernis-
ation, and has proved strongest precisely in those countries
whose economic development, fuelled by oil revenues, has been
most rapid (Keddie, 1988).
Now it is of course possible to preserve the original assump-
tions by arguing that the Islamic revival is not only the product
of modernisation, but also a reaction to it, and one that finds
strongest support among those social groups which have ben-
efited least from the process of economic expansion; in other
words, that it constitutes an essentially reactionary phenomenon,
which can only hinder further social and economic development
because it stands in opposition to some of its essential require-
ments. Such a judgement, however, only achieves plausibility by
ignoring the point on which the Islamic revival challenges West-
ern assumptions most strongly: its rejection of any universal
validity inhering in Western development models, whether capi-
talist or communist, and its search for a development path that
is more in tune with its own history and traditions. The Islamic
revival as a political phenomenon, in other words, has attained its
momentum from the discrediting of the capitalist and communist
models, both for their intrinsic defects - pursuit of profit without
social responsibility on one side, hostility to private property on
the other - and above all because they have involved subordi-
nation to foreign influences and interests, whether those of the
USA or the USSR.
Like nationalism, then, to which it is closely related, the Isla-
mic revival draws its appeal from an explicit rejection of external
192 The Legitimation of Power

models and external interference in an anti-imperialist age. It

also has more specific political determinants, deriving from the
Arab - Israeli conflict, and in particular from the Egyptian defeat
in 1967, which discredited the Nasserite socialist version of devel-
opment, and from which many concluded that a secular ideology
was no match for an enemy inspired by religious conviction, and
whose state had a primarily religious purpose. The strength of
the more fundamentalist movements, such as the Muslim
brothers, can be systematically related to the different phases of
the Palestinian crisis, and the reactions of the Western powers
to it (Enayat, 1982, ch.3).
Despite these common features making for a revival of Islam
as a politically significant force, the form of that revival has
varied considerably according to differences of local situations
and tradition. Islam is no more monolithic an entity than Christi-
anity. The political ideal of an 'Islamic state', in particular, has
been subject to an enormous range of interpretations. At a mini-
mum it can mean a state which gives protection to the institutions
and practice of the Islamic faith. More substantially, it can mean
a state whose policies are themselves conducive to Muslims living
a life in accordance with the requirements of their religion. A
more stringent demand is that the state should embody the
shari' a, or Islamic law, in its own legislation and judicial proce-
dures; though, here again, what this amounts to in practice will
depend on how narrowly the shari' a itself is to be interpreted.
Finally the idea of an 'Islamic state' can mean the subordination
of the state to the religious project of creating a spiritually trans-
formed community, inspired by 'the true Islamic mentality and
moral attitudes'. Each of these conceptions has very different
implications for the relationship between religion and politics,
and for the kind of critique that is made of the established order
(Enayat, 1982; Ahmed, 1987).
From the standpoint of political legitimacy, however, it is
important to note that, apart from the idea of a full-blown theoc-
racy, subject to the rule of the ulama, or jurists, none of the
conceptions mentioned above carries any implications for the
particular form of political system. The ideal of an Islamic state
or Islamic order is compatible with, and can be realised within,
any constitutional form: monarchical, representative, or what-
ever. This is because its primary concern is to specify the pur-
Crisis Tendencies of Political Systems 193

poses or goals which the state should serve, rather than the
source from which its political authority is to be derived. By the
same token, however, the project of establishing an Islamic order
has proved insufficient to secure the legitimacy of regimes whose
source of authority is itself in doubt. Many rulers have sought
to bolster a failing legitimacy by appeal to their Islamic creden-
tials, and by instituting an Islamic revival from above; but with-
out notable success. President Sadat's grandiose mosque-building
programme in Egypt revealed a religion based more on outward
show than authentic conviction (Gilsenan, 1988). President Zia's
introduction of an Islamic code in Pakistan aroused widespread
opposition, particularly among women, and demonstrated the
gulf that existed between the idea of a state for Muslims and his
more robust conception of an Islamic state (Alavri, 1988). Presi-
dent Numeiri's introduction of a similar code in the Sudan antag-
onised the Christian south, and intensified the divisions in that
country (Niblock, 1987, epilogue).
These attempts suggest that the manipulation of religion from
above to secure political legitimacy is ineffectual if not counter-
productive. Islam began as a revivalist movement from below
against a corrupt society, and its most dynamic political manifes-
tations in the modern period have all contained elements of
such revivalism (Mortimer, 1982, ch.2). In particular, it has only
provided a secure basis for political legitimacy in the twentieth
century when a movement from below has been successfully
harnessed to a political project at the level of the state. The two
distinctive types of political system in the contemporary Islamic
world - the conservative monarchy of Saudi Arabia and the
revolutionary theocracy of Iran - both had their origins in such
a conjunction. In the former, the revivalist movement was har-
nessed to the expansionist ambitions of a hereditary dynasty, and
served to reinforce a purely traditional type of legitimacy. In the
latter, the revivalist current took a politically revolutionary form,
and created a unique mass-mobilisation regime organised on
theocratic lines. Both reveal the limitations as well as the streng-
ths of Islam as a support for state legitimacy. As I shall also
explore, both demonstrate the complex interrelation between
politics, religion and an economic development based on oil.
The distinctive feature of the Saudi Arabian political system
is that it is an absolute monarchy whose legitimacy is based
194 The Legitimation of Power

entirely on the hereditary principle, and yields no place to ideas

of popular sovereignty. Power resides with the Al Saud family,
who approve the succession, advise t]le monarch and staff a
number of the leading ministries of state. Proposals to create a
constitution to limit the king's powers, and provide a means of
popular representation have never been accepted. Consent to
rule is expressed in traditional manner by meetings of tribal
leaders and other figures of influence, acting on behalf of society
as a whole (Niblock, 1982, ch.S).
The obvious question to ask is how such a purely traditionalist
system has been able to survive, not only in Saudi Arabia, but
in most of the other Gulf states as well. One reason has to do
with their small size. In other parts of the world the process
of colonisation created states whose territory either cut across
traditional political units, or incorporated a number of them.
The alteration of the spatial range of government necessarily
prevented the traditional structures from forming the basis for
the post-colonial state, and the hereditary principle from provid-
ing its source of legitimacy, even had it survived erosion from
other forces. In the area of the Gulf, in contrast, British influence
served to consolidate rather than disrupt the traditional political
units and system of rule, through the principle of 'non-inter-
ference' in the internal affairs of the sheikdoms.
Such an explanation, however, will not suffice for Saudi Arabia
itself. Although the house of Al Saud traces its ancestry back
for 500 years or more, the kingdom of Saudi Arabia is the
product of relatively recent conquest, dating from the 1920s and
1930s. What requires explanation, therefore, is not only the sur-
vival of a traditionalist system of rule, but its ability to maintain
its hold over a substantial territorial state. It is here that both
religion and oil have had a crucial part to play.
The significance of religion in the Saudi state derives from the
fact that the process of conquest of the Arabian peninsular was
simultaneously a religious crusade to spread Wahabism- a move-
ment of Islamic revival associated with the Al Saud family since
its origins in the eighteenth century - among the Bedouin tribes
of the interior (Hopwood, 1982). The converts provided a mili-
tant force, both religious and military, for use in the further
conquest of the region, and in driving the unpopular Hashemite
monarchy out of the area bordering the Red Sea (the Hijaz)
Crisis Tendencies of Political Systems 195

which contained the holy cities of Mecca and Medina. Although

the crusading force thus created, the Ikhwan, had in due course
to be suppressed by King Abdul-Aziz when it threatened British
control of Iraq, the prestige that the royal house has enjoyed in
the peninsular has been intimately bound up with its religious
credentials and mission: as principled upholder of Islamic law
and as protector of Islam's holy places (Habib, 1978). This mis-
sion has necessarily given the ulama a substantial influence over
legislation in the kingdom, especially in areas of public morality;
in return they have underwritten the legitimacy of the monarchi-
cal order (Mortimer, 1982, ch.6). Indeed, the traditionalisms
of religion and of dynastic succession have proved as mutually
supportive in Saudi Arabia as in Christian Europe of the past.
Moreover, it is a connection which the recent revivalist tendenc-
ies in the Islamic world have served to strengthen, even when
Saudi rule has been criticised for its conservatism and support
for US interests.
However, if the historic family link with a revivalist strain of
Islam has given the Al Saud their claim to moral influence,
their hold over the peninsula has also depended on the material
resources provided by the thoroughly modern phenomenon of
oil extraction (Niblock, 1982). Holding together a predominantly
tribal system requires substantial economic resources, since the
allegiance of local leaders has to be continually confirmed by
grants, personal gifts, hospitality and other favours. During the
early years of the state the revenue derived from customs and
pilgrimage dues was quite i.nsufficient to meet these require-
ments, and it was only the exploitation of oil in the 1940s that
enabled the state finances to be put on a sound footing. From
that time on, the central control exercised from Riyadh over
ever-increasing oil revenues has provided an effective deterrent
against secessionist tendencies, and a potent means for consoli-
dating the traditionalist system of rule. Indeed, an oil-based
'rentier state' can be particularly well-suited to the preservation
of a traditional legitimacy in the modern world, since it maxi-
mises the largesse available for distribution by rulers, while
allowing change in established patterns of economic production
to be kept to a minimum. It is only since the 1970s, in fact, that
substantial industrialisation has taken place in the country
through downstream development from petrochemicals.
196 The Legitimation of Power

In conclusion, it is only a unique combination of religious

authority and enormous oil revenues that has enabled the tra-
ditional system of rule to survive in Saudi Arabia, and to resist
the threat from the secular movements of Arab nationalism that
brought down the monarchies in Egypt, Syria and Iraq in the
1950s, as well as to withstand the repeated shocks of the Arab
- Israeli conflict, and most recently of the Iranian revolution.
However, it must be doubtful whether a purely traditional
legitimacy can survive indefinitely without some concession to
popular representation, now that the state has embarked on a
substantial programme of industrialisation. All the evidence of
industrialisation elsewhere indicates that its characteristic by-
products are an intensification of social divisions, and pressures
for working-class self-organisation that translate in time into a
demand for political rights. Thus, although the AI Saud have so
far successfully resisted demands for a constitution, it is unlikely
that they can control the consequences of economic development
sufficiently to prevent their recurrence in a more insistent form.
In other words, while the survival of a purely traditional system
is impressive, it has been dependent not only upon a unique
conjunction of factors, but also on a particular phase in the
country's economic development which is unlikely to last. The
shock waves created by Iraq's invasion of Kuwait have only
emphasised the vulnerability of traditionalist rule to future
expressions of mass discontent in a world where the principle of
popular sovereignty is a legitimating norm, and have underlined
the dilemmas of a regime so manifestly dependent upon US
protection.
When compared with Saudi Arabia, Iran presents a marked
contrast with respect to both the political systems it has had
over the past decades. The Pahlavi monarchy demonstrated the
weakness of a system of supposedly traditional rule which actively
sought to divorce itself from all other elements of traditionalism
within the society, Islam included. And the Islamic republic
which replaced it by revolutionary overthrow is as different as
anything could be from the traditional monarchy of Saudi
Arabia, not least in the way religion has been used for the
purpose of radical mass-mobilisation. The change from one
regime type to the other is instructive for any analysis of political
Crisis Tendencies of Political Systems 197

legitimacy in the contemporary Islamic world (see Keddie, 1981;

Abrahamian, 1982).
Although Iran had had a monarchical system of rule for cen-
turies, the Pahlavi dynasty enjoyed only limited legitimacy, for
a number of reasons. First, the dynasty was only created in the
present century, and rested on nothing more solid than the mili-
tary exploits of Reza Khan, not on the prestige of a royal lineage
stretching unbroken into the past, or on the support of an
extended family network, as in Saudi Arabia. Secondly, both
Reza Khan and his son Muhammed resisted all attempts to limit
their powers by making them accountable to a representative
assembly, as required by the 1905 constitution. Their autocratic
form of rule was of doubtful legality, since the 1905 constitution
was never formally replaced. Muhammed Reza Shah's powers in
particular were consolidated by a CIA-financed coup which
brought down the ~overnment of premier Mossadeq in 1953, in
order to prevent his proposed nationalisation of Iran's oil assets.
The Shah's later attempt to demonstrate the continuity of his
rule with Iran's pre-Islamic past in the grandiose celebrations for
2500 years of the monarchy in 1971 served more to expose the
weakness of his authority than to consolidate it. Lacking any
credible claim to traditional right, his regime was also hostile to
any genuinely popular source of legitimacy that cooperation with
a freely elected assembly might have provided.
However, the absence of any legitimating basis in a recognised
source of authority is not itself sufficient to explain the collapse
of a regime, especially one sustained by such a powerful army
and extensive means of repression. It was the actively delegi-
timating force of the huge mass demonstrations during 1978 that
undermined the will and the capacity of the Shah to maintain
his repressive system in being. These demonstrations marked the
culmination of two interrelated processes that had been at work
during the 1970s: on the one side the intensification of popular
grievances brought about by the Shah's programme of economic
development; on the other, the increasingly dominant role played
by religious leaders in the opposition to the Shah, and the
increasingly Islamic terms in which that opposition came to be
expressed. Both processes merit examination.
That a state-sponsored development programme based on
enormous oil revenues should come to undermine the authority
198 The Legitimation of Power

of the state, rather than reinforce it as it did in Saudi Arabia,

is to be explained by the particular interrelationship between
the pattern of economic development and the character of state
institutions, rather than by any general category such as that of
the 'rentier state' (Mahdavy, 1970). In Saudi Arabia the oil
revenues were deployed in a way which strengthened the tra-
ditional order based upon patronage, and confirmed the loyalty
of local chieftains and political leaders to the ruling family, which
controlled resources at the centre. In Iran, by contrast, the Shah's
more ambitious programme of 'modernisation' simply intensified
the gulf between a privileged Westernised elite and the vast
majority of society. This gulf was most evident in the economic
disparities of the cities, between those who could afford to sustain
a Western consumer lifestyle, based largely on imported goods,
and the impoverished masses who had been driven off the land
by agrarian reforms into the insecure employment, slum housing,
and inadequate infrastructure of urban life. The austerity measu-
res introduced by a new premier in 1977 to deal with the overhea-
ted economy intensified discontent, and underlined the contrast
between the resources lavished on the military and bureaucratic
apparatus of the state and the impoverished condition of the
majority (Halliday, 1979; Keddie, 1981, ch.7).
The gulf between regime and society was further widened by
the emphasis given to Western, and particularly US, interests in
the Shah's policy. Ever since the coup that brought down premier
Mossadeq, the regime had been associated with subordination to
the USA. The diplomatic immunity granted to US military per-
sonnel and 'advisers' in 1964 had caused particular offence. The
pace of economic development, which accelerated with the
increase in oil prices in the early 1970s, brought a huge influx
of foreign staff to the multinational companies established around
Tehran, and a sharp rise in rents in the capital. And with the
import of Western consumer goods for the Iranian elite came
also Western lifestyles and culture, films and TV programmes
and the pervasiveness of consumer values. The Shah's policies,
besides intensifying economic inequalities, thus came also to be
associated with the undermining of any distinctive Iranian ident-
ity and its submergence in an alien culture (so called 'Westoxic-
ation').
In such a context, although there was a strong tradition of
Crisis Tendencies of Political Systems 199

secular opposition in twentieth-century Iran, the link between

royal dictatorship, economic inequality and cultural alienation
proved to be much more effectively exposed in Islamic terms
(Keddie, 1982). The categories of the secular opposition, whether
those of a democratic republic or a socialised economy, had the
disadvantage of themselves being of Western origin, and were
incapable of addressing the cultural dimensions of the society's
malaise. An Islamic analysis, in contrast, which linked the Shah's
lack of legitimacy with his policies of secularisation and his disre-
gard for the country's religious heritage, offered a much more
persuasive socio-political critique, and one that spoke much more
powerfully to the sense of national humiliation. A return to Islam
offered a solution to moral decline, an assertion of national
independence and a restoration of political legitimacy simul-
taneously. In delivering this message, the Shah's religious critics
had the advantage over the secular opposition of much closer
links with the deprived masses, both urban and rural, and an
institutional autonomy that enabled them to survive the extremes
of repression, for all that the Pahlavis had systematically sought
to limit the role and influence of the mullahs over many decades.
They were thus able to link their own particular and longstanding
grievances with the general discontent towards the regime, and
so give their critique of the Shah's policies a universal significance
for the society.
Once a religious definition of the country's discontents had
become prevalent, it was not surprising that the symbolic leader-
ship of the opposition should focus around the person of the
Shah's most consistent and unyielding religious critic, Ayatollah
Khomeini, who had been exiled as far back as 1964, and whom
the Shah singled out for special condemnation in January 1978.
What took everyone by surprise, however, was the mobilising
power of Khomeini's message, evidenced in the enormous mass
demonstrations of 1978 and the exemplary readiness of demon-
strators to sacrifice their lives in public opposition to the regime.
Although the role of armed mujehaddin units in the final stages
of the revolution in February 1979 should not be overlooked,
Iran was distinctive among Third World revolutions in that the
regime did not succumb to guerrilla warfare in the countryside,
but to the urban masses demonstrating peacefully on the streets;
it was not defeated primarily through armed struggle, but through
200 The Legitimation of Power

a process of popular delegitimation which undermined its author-

ity from within. Here Iran set an example of 'people power' that
was to be repeated throughout the 1980s (Ahmad, 1982; Halli-
day, 1988).
However, the difference between Iran and the later examples
of 'people power' is that none of the others resulted in a mobilis-
ation regime, i.e. one legitimated primarily through mass-mobilis-
ation rather than electoral competition. For that to take place,
other preconditions were required besides the revolutionary pro-
cess itself, in particular the establishment of a monopolistic doc-
trine laying exclusive claim to the truth, and an organisation
capable of mobilising the masses in its service on a continuing
basis. Of crucial importance in this context was Khomeini's dis-
tinctive innovation in Shi'ite political thought. Traditionally,
Shi'ism has always been pluralistic, acknowledging no single
authority or source of religious interpretation; and it has defined
only a limited political role for the ulama, as critics of rulers
who threatened the practice of the faith or affronted the beliefs
of the faithful, as the Shah had done. All shades of religious
opinion had been agreed on a negative critique of the Shah.
What Khomeini proposed, however, went much further: it was
that the ulama should themselves wield political power, with the
aim of transforming society according to Islamic law, of creating
'conditions conducive to the production of morally upright and
virtuous human beings'. If the only legitimate government, as he
argued, was government which actively sought to implement the
shari' a in all spheres of life, then only those trained to interpret
Islamic law were competent to oversee legislation, and any oppo-
sition to them must by definition be illegitimate (Rose, 1983; cf.
Zubaida, 1982).
The symbolic leadership that Khomeini came to exercise over
the revolution enabled him to appropriate its energies for his
own political project; to treat it, as he subsequently claimed, as
a referendum for an Islamic constitution. Yet the political system
he advocated was not established all at once, or without a strug-
gle (Bakhash, 1986; Hiro, 1987, pt.2). The early years after the
revolution were marked by a condition of dual power, between
the President and Khomeini as Vice-Regent, between the formal
institutions of the state and the revolutionary institutions of law
enforcement, and between two incompatible principles of legit-
Crisis Tendencies of Political Systems 201

imacy: appointment to high office through electoral choice in

open competition; and appointment by virtue of qualifications in
religious law, confirmed by mass acclaim and the power of mass-
mobilisation. That the second came to prevail over the first was
due not only to the prestige of Khomeini himself, but to the
forces of a militant revivalism pushing him from below, and to
the impact on society of external events such as the Iraqi invasion
and the running sore of the Arab - Israeli conflict. To be sure,
elements of electoral legitimation remained, but they were sig-
nificantly weakened by the requirement that candidates for the
presidency be approved by the Council of Guardians, and by the
effective suppression of alternatives to the Islamic Republic
Party. And that party, in turn, with its links to the mosques and
local prayer leaders, operated more as an organisation for mass-
mobilisation than electoral competition.
Expectations that the regime would quickly collapse under-
estimated the potency of the Islamic revival when harnessed to
an anti-imperialist crusade. They also overlooked the distinctive
strength of a mobilisation regime, which Iran shared with both
the fascist and the communist exemplars. This is that the mass
mobilisation of the ideologically committed serves at the same
time to provide a public legitimation for the leadership, to over-
awe the UI).committed and to generate support for the repression
of opponents. From the standpoint of its victims Khomeini's
repression was no different from that of the Shah. From the
standpoint of the character of the regime, however, it was less
a sign of weakness or lack of support than a necessity for preserv-
ing the doctrinal exclusivity around which its support was mobil-
ised. In a mobilisation regime the legitimation of the leadership
and its policies and the radical delegitimation of opponents are
systematically connected; and repression serves as part of a politi-
cal crusade as much as the interests of 'state security'. Although
a more pragmatic temper may have come to inform state policy
since the end of the Iran- Iraq war and the death of Khomeini,
the potential for the reawakening of revolutionary fervour
remains present within Iranian society. Nor is there to date any
evidence of the political system itself evolving in a more pluralist
direction, as the shift to a more genuinely electoral mode of
legitimation would require.
Anyone comparing Iran and Saudi Arabia, in conclusion, must
202 The Legitimation of Power

be struck by the difference in the forms of Islamic revival with

which their regimes have been associated, and the difference in
the types of political system that Islam has served to validate.
In Saudi Arabia, Islam has underwritten a conservative mon-
archy, whose family and tribal networks, nourished by oil rev-
enues, have given it sufficient authority to keep the masses at
arm's length from the political process, and to sustain collabor-
ation with US interests in the region. In Iran, Islam has provided
the ideological focus for a revolutionary movement, radically
anti-monarchical and anti-imperialist, which has brought the
deprived masses onto the centre of the political stage, first in
undermining the Shah's regime, then in legitimating the rule of
the ulama.
What is common to both states, however - and this is the key
to their political legitimacy - is the organic connection forged
between the definition of a religious purpose and the acknowl-
edged source of authority from which the rules of the political
system are derived. In the one, a traditionalist religious purpose,
which seeks to maintain the precepts of the shari' a intact in a
changing world, has made common cause with a hereditary dyn-
asty that also derives its authority from the distant past, and
seeks to preserve its essential principles of rule unchanged under
modern conditions. In the other a radical, transformative pur-
pose, whose goal is to transform individual character and with
it social relations also, has given a leading role in the state to
the ulama, and has required a politically mobilised society rather
than one that is merely quietist or apolitical. In each case there
is a congruence between the religious purpose and the basic
principles of political authority.
As with all legitimate political systems, both Saudi Arabia and
Iran have seen themselves as models for others, if not uni-
versally, then at least within the Islamic world. Saudi Arabia is
the leading representative of conservative forces within the
Middle East. And Iran has actively sought to export its revo-
lution to neighbouring states. Yet there are considerable
obstacles to either providing a replicable model for others to
copy, not least because of the particular historical trajectory that
produced each type of system. It has required specially favour-
able circumstances for monarchies to survive in the Middle East;
once a ruling family has been overthrown, it is difficult to see
Crisis Tendencies of Political Systems 203

how it could be restored to anything other than a decorative

function, since the whole rationale for traditional legitimacy rests
on the proof of superiority that comes from unbroken continuity
with the past. Iran for its part may seek, or at least may have
sought, the exemplary role played by previous revolutions,
whether the US, French or Bolshevik; but it is handicapped
by its minority position, as Shi'ite and non-Arab, among the
populations it seeks to influence. At most it has the disruptive
power to cause problems for other states, rather than to provide
a compelling model for them to emulate.
If neither political system offers a generalisable model to
others, what alternative legitimate order is available to set against
the general run of personal dictatorships in the Islamic world?
The communist model has lost its attraction, and in any case its
militant secularism was always an affront to religious believers.
Liberal democracy, on the other hand, besides facing the diffi-
culties that are general in Third World countries, confronts a
further handicap where fundamentalist movements have estab-
lished themselves in any strength. If by 'fundamentalism' we
mean the demand for strict adherence to a literalist interpretation
of sacred texts, then this typically threatens the political equality
between the sexes that is entailed in universal suffrage, and the
equal right to hold public office. If we include further the claim
to possess a monopoly of truth, then this entails opposition to
the freedoms of expression and association necessary to electoral
legitimacy.
Islam is no more inherently fundamentalist than any other
belief system. Its history embraces pluralism of interpretation
and coexistence with other religions. The modern period has
witnessed many attempts to develop a creative synthesis from
Western and Islamic elements in the social and political spheres.
For these to be effective in helping establish a more generalisable
form of legitimate political order depends upon a decline in
fundamentalism; and among the necessary conditions for that
must be a resolution of the Palestinian issue, and a reduction in
neo-imperialist tendencies on the part of Western powers, neither
of which looks particularly likely at present.
Throughout this chapter, in fact, what has become apparent
is the importance of external influences for the legitimacy of the
supposedly sovereign state. Sovereignty, as I have argued, is a
204 The Legitimation of Power

condition of vulnerability as well as of legal supremacy and self-

sufficiency, in that there is no superior authority to guarantee
the rules of political power, which have to be legitimised intern-
ally. Yet that legitimacy is itself conditioned in form and degree
by the impact of external forces, both ideological and material,
and will increasingly be so conditioned in future, with the grow-
ing interdependence of states. At the ideological level, the suc-
cess of capitalist democracy in its heartlands continues to prove
a potent source of attraction to other societies, as the communist
model does no longer. At the same time, the power associated
with that success has worked to undermine alternative bases of
political legitimacy, as well as to provoke significant movements
of reaction appealing to indigenous sources of authority. If we
cannot always predict the consequences that the impact of such
external forces will have, their continued significance for political
legitimacy is beyond doubt.
7 Modes of Non-
legitimate Power
The previous two chapters have discussed the legitimation prob-
lems of the contemporary state in general, and of particular types
of political system. The final topic to consider is the actual pro-
cess of breakdown of political order, and the way in which
legitimacy deficits develop into delegitimation and then illegality.
In the first section of this chapter I shall distinguish the different
modes of non-legitimate power at a conceptual level, and show
how they are systematically related to one another as a dynamic
or sequential process. The second and third sections will then
analyse two different types of such a process: those leading to
revolution and coup d'etat respectively; and will consider the
prospects for the successful relegitimation of revolutionary and
military regimes in turn. In a concluding section I shall draw
together the different elements in an explanation for the repeated
breakdown of legitimacy experienced by states in the contempor-
ary world.
A recurrent theme throughout the book has been that the
legitimacy of power must be understood as multi-dimensional,
comprising legality or rule-conformity, the justifiability of rules,
and confirmation through expressed consent. One consequence
of clarifying this threefold structure is that it becomes possible
to distinguish different ways in which power may fail to achieve
legitimacy. It may be acquired in contravention of the rules, in
which case it is illegitimate. The rules may find only weak justifi-
cation in terms of accepted beliefs about the proper source of
authority or ends of government; here we should speak more
correctly of a legitimacy deficit, or weakness. Finally, those whose
consent is necessary to the confirmation of legitimacy may act in
ways which demonstrate the withdrawal of their consent; this is

205
206 The Legitimation of Power

called delegitimation. I shall consider each of these in turn, before

showing how they are related to one another as a sequential
process.
Illegitimate power is, simply, power that is acquired through
a breach of the constitutional rules. The typical forms of breach
are coup d'etat, revolution, conquest, secession. The first two
involve an illegal assumption of power - in the one case by the
military, in the other as the culmination of mass struggle - which
leaves state sovereignty over an existing territory in principle
intact. The latter two involve an infringement of territorial sover-
eignty, in the one through subjection to an external power, in
the other through its rejection in favour of a new self-governing
state. Most of these breaches of legality will involve the use of
force; where not, they usually require the use of force to defend
and consolidate the power that has been illegally acquired.
The point of using the strong negative term 'illegitimate· for
these acquisitions of power is that they constitute a clear and
indisputable negation of the first condition of legitimacy, which
is legality. As such they confront particular problems of legitim-
ation which many of them are unable to overcome. First, as with
any infringement of the law, they require special justification by
reference to extra-legal norms that are widely acknowledged,
and an exigency sufficiently compelling to warrant such action. In
particular, since those who have breached the law will themselves
require that legality be subsequently observed, they have to pro-
vide convincing reasons why others should not copy their
behaviour. Then secondly, in creating a definitive break with an
established constitutional order, they have to find a new basis of
legitimacy for the system of rule they inaugurate in its place.
The term 'illegitimacy' thus conveys at one~ the idea of manifest
illegality, the definitiveness of the break with an established con-
stitutional order, and a task or programme of relegitimation that
may or may not be successfully carried out.
All the above examples of the illegal acquisition of power are
relatively clear-cut. However, there are others that are much less
so. As with any system of law, there are areas of uncertainty
that give rise to constitutional dispute. Did the Head of State
have the power to invite this party leader to form a government,
or to dismiss that one, in the given circumstances? Was the
province legally entitled to secede from the federation by a
Modes of Non-legitimate Power 207

decision of its own parliament? The constitutional position may

be unclear, or the validity of the constitutional system as a whole
may be disputed, e.g. on the grounds that its own establishment
was of doubtful legality. Or again, as with any other system of
law, the factual basis of any infringement may be uncertain. The
most common example is electoral fraud. It is not always easy
to detect, to assign precise responsibility for, or to assess what
effect it has upon the outcome, especially where it extends back
to the process of electoral registration, or even to the compilation
of the census returns. Is the victor legitimate or illegitimate? In
all these examples it is best to talk of 'disputed legality' or
'contested legitimacy'. As with the must clear-cut cases, any of
them may provide both occasion and justification for further
violations of legality.
The term 'illegitimacy', then, denotes a breach of the consti-
tutional order; and it is with the clear cut instances of revolution
and coup d'etat that the present chapter will primarily be con-
cerned. Illegitimacy is to be distinguished from a condition of
'legitimacy deficit' or 'weakness', where the constitutional order
diverges from accepted beliefs about the proper source of politi-
cal authority; or where it either magnifies, or is incapable of
resolving, chronic problems of government performance; or
where judgement about these questions is a matter of fundamen-
tal disagreement within the society.
As the previous chapters will have made clear, there are many
forms of legitimacy deficit or weakness, with widely differing
significance. However, the basic distinction drawn earlier
between the source of political authority and the ends or pur-
poses of government, between the 'right' and the 'good', is of
equal importance here. A legitimacy deficit involving the former
can best be described as a divergence or discrepancy between
the constitutional rules and the beliefs that should provide their
justification, whether the divergence exists because the rules have
been established or altered in a manner that is incompatible
with established beliefs about the rightful source of authority, or
because the beliefs of a society have themselves evolved over
time in a way that weakens support for the constitutional order.
A legitimacy deficit involving the ends or purposes of govern-
ment, on the other hand, can be described as an inadequacy
or incapacity of the constitutional rules to facilitate successful
208 The Legitimation of Power

government performance, or to provide resolution in the event

of failure, whether the failure is one of ineffectiveness or of
partiality in respect of the ends that government exists to attain.
This distinction can be further clarified by relating it to two
very different metaphors or analogies that have historically been
used by political theorists to describe constitutional systems: the
organic and the mechanical respectively (e.g. J.S. Mill, 1964,
pp.175-7). The organic analogy sees political systems as a kind
of natural growth, evolving over time, and dependent upon a
particular soil or environment for their flourishing. This seems
an appropriate analogy to use to describe the conformity or lack
of conformity of a constitutional order to the 'culture' of a given
society, and in particular to its beliefs about the valid source of
political authority. The mechanical analogy, on the other hand,
sees constitutional systems as deliberate constructions, designed
more or less effectively to fulfil the ends or purposes of govern-
ment. This seems a more appropriate analogy to use to describe
the ability of a political system to facilitate government perform-
ance or resolve government failure. My argument, of course, is
that legitimacy embraces both; and that a legitimacy deficit can
occur either because of a divergence between constitutional rules
and given beliefs about the source of political authority, or
because of their inadequacy to resolve problems of government
performance; or indeed both simultaneously.
If we are to look for an analogy that combines both elements,
then perhaps it is to be found in civil engineering, where a given
structure has to be both appropriate to its particular site and
environment, and effective for the function it has to serve. The
advantage of this analogy is that it enables us to make a further
distinction, between the inherent strength vr weakness of a struc-
ture, in its context and for its purpose, and the stresses to which
it may be subject. If we translate this into political terms, then
the strength or weakness of the structure represents the degree
of legitimacy of a constitutional system, and the stress to which
it is subject represents the impact of government failure upon it.
By 'failure' here is meant some evident inability of government
to fulfil the purposes expected of it, or to do so without intensify-
ing social divisions. It is government failures, in this sense, that
constitute the pressure or load with which a political system has
to cope. The more secure its legitimacy, the greater the stress it
Modes of Non-legitimate Power 209

will be able to withstand. On the other hand, where legitimacy

is weak, government failure will deliver the decisive pressure
that will set in train a crisis of political authority.
Analogies should not be pushed too hard. Those drawn from
the natural sciences convey an illusion of exactness that is out
of place in the analysis of societies. Those drawn from construc-
tion tend to be static, and cannot adequately capture the dynamic
process of interaction and 'feedback' between different elements
of a social or political system. Even 'systems' analogies run the
danger of treating societies as if they were composed of imper-
sonal elements rather than living people. Thus the 'level of toler-
ance' of a political system is actually the tolerance level of the
human agents involved in it; and legitimacy as a system character-
istic is actually a relationship between rules and the beliefs that
underpin them. However, provided due caution is observed, the
ideas of 'stress' and 'critical load' can help illuminate the relation
between the pressures imposed by government failure, and the
inherent strength or weakness of the political order on which
they bear.
A legitimacy deficit or weakness, then, is a condition of inap-
propriateness or inadequacy in the constitutional rules, which
limits the degree of support that they, and those deriving power
from them, can command when subjected to the stress of govern-
ment failure. 'Delegitimation', finally, denotes a process whereby
those whose consent is necessary to the legitimation of govern-
ment act in a manner that indicates their withdrawal of consent.
Mass demonstrations, strikes, acts of civil disobedience: such
actions can have damaging consequences for the moral standing
of a government and also its capacity to rule; the more so, the
larger the numbers involved or the more crucial their cooperation
is to the attainment of the government's purposes.
However, the precise significance of such actions must depend
upon the context and the nature of the political system. In a
liberal democracy, the expression and organisation of opposition
to particular policies is an accepted feature of politics, and has
no delegitimating significance for government. It is only when it
is designed to make the policies of government unworkable, or
actively to bring it down, or to demonstrate allegiance to a
different political order, that it crosses the threshold from normal
opposition to delegitimation, or attempted delegitimation. The
210 The Legitimation of Power

difference here is in the declared intentions of the actors

involved. On the other hand, in a political system where there
is little freedom of expression and association allowed, the organ-
isation of opposition to particular policies itself constitutes a
challenge to government authority, whatever the intentions of
the actors. In a communist system for example, the expression
of opposition outside the ranks of the party, especially by the
working class, challenges the party's claim to a monopoly of
representation, and therefore the terms on which its own legit-
imacy is defined. And under authoritarian regimes of whatever
hue, the banning of a demonstration or protest against a particu-
lar policy turns it into an action of very different significance,
and one which, if it takes place, thereby becomes a challenge to
the authority of government as a whole.
What makes these actions all examples of delegitimation,
whether through the declared intentions of the actors, the nature
of the political system, or the response of government itself, is
that they challenge the claims of government to govern with the
consent of the population. As such they serve to weaken or
undermine whatever moral authority a government possesses.
The more so, where the action involves an explicit refusal of
people to 'do as they are told', since the humiliation of impotence
intensifies the loss of moral authority. It is of the essence of such
actions, however, if they are to have any delegitimating effect,
that they should take place in public, and be publicly known. It
is for this reason that governments will do all they can to prevent
such manifestations from taking place at all; to suppress them if
they do take place, so as to discourage repetition; and to stop
news about what is happening from spreading to the rest of
society. Just as consent has no legitima£ing force unless it is
expressed, so the withdrawal of consent has no delegitimating
impact if it is not generally known. Recent history shows many
examples of the crucial role that journalists play in amplifying the
delegitimating effects of civil disobedience, from the townships of
South Africa through the 'intafada' in occupied Palestine, to
Tiananmen Square and the mass demonstrations throughout
Eastern Europe in the autumn of 1989. Governments have good
reason to try and curtail or suppress news coverage, since they
understand only too well that the dissemination of information
in such a context can have profound political consequences.
Modes of Non-legitimate Power 211

The distinctions drawn above between the different ways in

which political power may lose or lack legitimacy - through a
breach of legality in the acquisition of power, through normative
inadequacy of the constitutional order, through the active delegi-
timation of government - follow logically from the three-dimen-
sional conception of legitimacy developed earlier. To set them
out in this analytical way may appear somewhat schematic or
formalistic. In practice, however, there is an integral connection
between them. Thus a major breach of legality typically follows
the loss of authority produced by acts of delegitimation, and
these in turn typically take place, with most damaging conse-
quences, where the legitimacy of the wnstitutional order is itself
weak. In other words, we can posit a developmental sequence
whereby some failure of government performance, compounding
a normative inadequacy of the constitutional rules, sets in train
a process of delegitimation and loss of authority, which culmi-
nates in a seizure of power in contravention of the legal order.
Such a sequence accords with the reasoning of the political agents
involved, since a normative inadequacy of the constitutional
order, when accentuated by government failure, provides a suf-
ficient ground for disobedience or withdrawal of consent, and
the consequent loss of authority provides both reason and
occasion for a breach of constitutional legality.
The main part of this chapter will focus on two such sequential
processes that are typical of the contemporary world: one leading
through revolution to a democratic or mobilisation regime; the
other through coup d'etat to a military dictatorship. The first,
revolutionary, process typically takes place under an authori-
tarian system, whose source of authority is problematic because
it is insufficiently anchored in the principle of popular sover-
eignty, and which lacks popular legitimation because the masses
are partially or completely excluded from the political process.
Although the regime is further weakened by its inability to over-
come problems of government performance, the key element in
its legitimacy deficit lies in an inadequate source of authority. In
this context some damaging measure of government policy
provokes widespread protest, which breaches the authorised
limits of political expression, and involves a direct challenge
to the government's authority. Responding with characteristic
repression, the government only succeeds in widening the scale
212 The Legitimation of Power

of the protest beyond its particular policies to include the rules

of power themselves, and their exclusion of the people from the
political process. The terminal point for the regime is only
reached when its loss of authority over society penetrates the
armed forces themselves, and substantial sections refuse to obey
orders, or defect to the popular movement. At this point power
can be seized in the name of the people, and a new order be
constituted, deriving from a valid source of authority and legit-
imated through elections or continued popular mobilisation.
The second sequential process typically takes place under a
democratic regime, whose authority is grounded in the acknowl-
edged principle of popular sovereignty, but which is beset by
problems of performance, and where the rules of the political
system come to be perceived as themselves part of the problem,
rather than a means to its solution. The electoral process offers
only an intensification of competition, and the continued vulner-
ability of the losers. At the same time the freedoms of speech
and association, which are necessary to electoral legitimacy and
to the continuous expression of public opinion, are pushed to
the limit by sectional interests determined to frustrate govern-
ment policy and so to undermine its authority. Here it is not so
much the exclusion of society from the political process, as the
consequences of its inclusion, that is the problem. Here delegi-
timation - acts of disobedience, withdrawal of consent - takes
the form, less of an assertion by the populace of its right to
intervene, than of the use of that right by particular interests to
undermine government policy. And the force of delegitimation
is to expose, not so much the government's lack of a valid source
of authority, as its inability to secure the general interest. It is
in such a context that a section of the military defects and seizes
power in the name of a general societal interest that has been
compromised, and the reassertion of order and authority that
have been undermined.
The two processes of revolution and coup d'etat sketched out
above should be seen as ideal-typical ones only, which do not
exhaust the full range of historic possibilities. There is, for exam-
ple, a third such typical process, which begins with a basic
disagreement within society, particularly over the spatial organis-
ation of the state, and which moves through various forms of
delegitimation towards secession and/or civil war, depending on
Modes of Non-legitimate Power 213

how the respective populations are distributed over a given terri-

tory. There is not the space to deal with this process in any detail
here. Even with revolution and coup d'etat, not all instances fit
neatly into the typical aetiology outlined above. Not all coups
have their origin in democratic systems: military regimes them-
selves, lacking legitimacy, are particularly vulnerable to repeated
coup attempts; and the examples of junior officer revolts in the
Middle East in the 1950s show that they can also take the form
of substitute revolutions where a traditional legitimacy has been
exhausted, and the source of political authority is itself in ques-
tion. Equally, as the example of Nazism shows, it is possible for
a revolutionary or quasi-revolutionary mobilisation to take place
in a democratic system, and to lead to a mobilisation regime
based upon an exclusive claim to truth. However, the two pro-
cesses as I have outlined them are sufficiently typical in the
contemporary world to provide at least a framework for the
more detailed analysis that follows.

Revolutions and legitimacy

Revolutions are forced changes of political system which take

place under the pressure of extra-legal mass action, and lead to
an extension of popular involvement in a new political order.
Revolutions share with rebellions or revolts the feature of mass
insubordination to government; they differ from them in that
they result in a change of political system rather than merely of
policy or personnel. They share with coups d'etat the character-
istic of system change through extra-legal action or overthrow;
they differ, as I have already argued, in that the prime agency
of change is the subordinate masses rather than the military
apparatus, and that they result in an extension of popular partici-
pation in the political process, rather than its curtailment.
Against those who hold that the defining characteristic of revo-
lution is violence, I would urge that it is rather that of extra-
legal mass action, which may or may not be violent. Against
those who maintain that the only 'true' revolutions are social
revolutions, I would insist that revolutions are by definition politi-
cal events, though they may or may not be accompanied by
profound social transformation. Against those who argue that
214 The Legitimation of Power

the only 'true' revolutions are those which seek to establish a

quite new order, or realise a completely new ideal, I would
contend that revolutions may be restorative as well as innovative,
as when they reestablish a democratic constitution that had pre-
viously been corrupted or violated. In other words, there can be
different types of revolution. But what distinguishes them all as
revolutions is the combination of a particular form of agency -
extra-legal mass action - with a particular type of outcome: a
forced change of political system involving expanded popular
participation.
The literature on revolution is enormous, and the range of
explanatory theorising about it is very wide, not to say contradic-
tory, reflecting the sheer complexity and variability of the
phenomenon. No attempt will be made here to survey these
theories, or to adjudicate between them. (For comparative stud-
ies see Dunn, 1972, Skocpol, 1979, Krejci, 1983; for surveys of
theory see Stone, 1965, Freeman, 1972, Kramnick, 1972, Zago-
rin, 1973, Cohan, 1975, Kimmel, 1990). My purpose is more
modest: it is to identify the contribution that an understanding
of legitimacy can make to the analysis of revolution, as one
element in a complex whole. This contribution is twofold: first,
in explicating the loss of moral authority by the old regime, its
character and different phases. The second part lies in identifying
what is involved in the relegitimation of a new political order.
In discussing each of these in turn, I shall make reference to the
most recent decade of revolutions, from Iran and Nicaragua in
1979 to the revolutionary year of 1989, as well as the classic
revolutions of the modern period that have been much analysed
(France, Russia, China).
At the heart of every process which leads to the overthrow of
a political system lies a legitimacy deficit: a conviction that the
existing regime cannot satisfy the general interests of society,
whether through persistent incapacity or chronic divisiveness and
partiality. What makes the process a revolutionary one is not
only that this conviction becomes widespread, but that the
regime's inadequacy comes to be linked to the absence of any
valid source of its authority. In particular, those who wield power
are seen to lack effective authorisation 'from the people'. It is
the failure to acknowledge the people as the ultimate source of
authority in the system's rules of power that is not only seen as
Modes of Non-legitimate Power 215

unjustifiable in itself, but comes to be defined as the root cause

of the regime's failings. In other words, the process of revolution
is inextricably linked with the emergence of the principle of
popular sovereignty in the modern age. Revolutions result not
only from the exclusion of the masses from political influence,
but from their assertion of the right to be included. Before the
emergence of popular sovereignty as a principle, there could only
be rebellions or civil wars, not revolutions.
What form the deman<;l for a political order deriving its author-
ity 'from the people' comes to take, itself depends upon the
nature of the old order, and the point where the latter's source
of authority has become eroded. Where it is the principle of
traditional or dynastic legitimacy that has suffered erosion, at
the forefront will be the assertion of popular sovereignty as
such, in any form. Where it is foreign domination that is under
challenge, it will be the principle of national autonomy and the
authority of indigenous political forces that will be asserted.
Where it is the limitation of popular sovereignty by a more
hierarchical principle of appointment that is at issue, the demand
will be for free elections for the chief offices of state. Where it
is the restriction of the suffrage to a minority of the population,
it will be for a fully representative test of the people's will. At
the heart of revolution lies the erosion of some principle of
power which has served to exclude the masses, partially or com-
pletely, from the political process; to the particular form of
exclusion or limitation there stands a corresponding demand for
their inclusion. Revolutions are the eruption of this demand with
the force of a quasi-natural event.
To trace an integral link between the erosion of a hierarchical
principle of authority, the mobilisation of the masses and the
form of the demand for their political inclusion, is not to commit
the error of seeing revolutions as consciously intended, or their
outcomes as predictable. Revolutions are paradoxical phenom-
ena: intelligible but also unintended, explicable at the same time
as they are unpredictable. Thus the erosion of a regime's source
of authority can take place over a long time before there is any
widespread recognition of it, much less any readiness to act on
that recognition. Dissident intellectuals may labour for years at
considerable risk to expose the moral bankruptcy of a regime,
without any apparent effect. It usually requires a serious assault
216 The Legitimation of Power

on people's interests by a government to make them aware of

the lack of justification for its rules of power. And the link
between problems of government performance and the inad-
equate basis of its authority is often only consciously made under
the impact of political struggle itself.
Thus a popular mobilisation to protest at some damaging
policy, which is answered by repression, will turn people's atten-
tion from the specific grievance to the rules of power themselves,
and lead them to connect the damage to their interests with their
systematic exclusion from the political process. It is this that
explains the typical 'agenda shift' that takes place in a revolution:
what starts out as a protest against a particular policy escalates
rapidly into a challenge to the political order as a whole, without
this consciously being intended. Yet it is only possible where the
potential for such an outcome is already present in an exclusivist
principle of authority whose intellectual basis has been eroded
over time.
One essential element, then, in a regime's loss of moral author-
ity is the process whereby people come to make the link between
the policy failures from which they suffer, and the invalidity of
a principle of authority which excludes them from political influ-
ence: to make the link, so to say, between the 'bad' and the
'wrong'. Such a connection occurs over different timescales for
different people; what is characteristic of revolution is that it is
sparked simultaneously for large numbers by some notable event.
This is the 'shift of consciousness' that contributes to a regime's
loss of moral authority. However, there is another essential con-
tributory factor, and that is provided by the acts of mass dis-
obedience and insubordination - the demonstrations, strikes,
etc., undertaken in explicit defiance of authority - which very
publicly delegitimate the regime. Although the shift of conscious-
ness and the acts of delegitimation may in practice occur simul-
taneously, it is helpful to distinguish them analytically.
Delegitimation is the dramatic loss of prestige incurred by a
regime when large numbers of its subjects refuse to do as they
are told in a public and collective way. Those who have always
claimed to 'represent' the nation are now exposed as unrepresent-
ative; their pretensions to serve the public interest are declared
fraudulent by the public in person. Above all, those who wield
enormous power are suddenly glimpsed as impotent. It is not so
Modes of Non-legitimate Power 217

much that the emperor is revealed to have no clothes; the act

itself divests him. And the memory of this public stripping will
persist even where the 'emperor' reasserts his authority through
force. Moreover, the suffering endured by an actively mobilised
population under fire only increases its moral authority in face
of the regime. Where the 'martyrs' in the consciousness-raising,
pre-revolutionary stage are typically drawn from the intelligentsia
and established leaders of moral opinion, here in the phase of
public delegitimation it is typically the young who are the most
determined, most self-sacrificing agents of defiance and popular
mobilisation.
Acts of public delegitimation, bringing a massive loss of
authority, constitute the crisis point for a regime. Facing a revol-
utionary situation, it is confronted with the unpalatable choice
between concession or repression. Any concession short of a
radical and negotiated change of political system - de facto sur-
render of power- will prove insufficient. So often in a revolution-
ary situation governments are caught out by the rapidity of the
'agenda shift', and find themselves lagging behind the point of
popular demand; they offer concessions on policy or personnel,
when it is the system itself that has become the target. Insufficient
concessions both increase resentment and signal the regime's
weakness and vulnerability. At the same time the strategy of
making far-reaching concessions to demobilise people and pre-
pare the ground for later repression is a highly risky one, since
the reform process may become irreversible under its own
momentum.
Repression, on the other hand, has to be massive and ruthless
to succeed. The outcome of the repressive choice depends cru-
cially on the unity of the armed forces, and on their willingness
to carry out orders to fire on a defenceless population, the
defence of whom is their chief rationale. The regime's loss of
moral authority through its public delegitimation produces an
acute conflict of loyalty among the military command. Are they
with the people or against them? The outcome will always be a
matter of local circumstance, as the revolutions of 1989 demon-
strated. In Bulgaria, Czechoslovakia and East Germany the high
command signalled its unwillingness or inability to undertake
wholesale repression, and the regimes simply crumbled without
a fight. In Romania the security forces initiated the massacre for
218 The Legitimation of Power

which they had been recruited and trained, but the regular army,
with its own grievances against Ceausescu, sided with the people
in a bitterly fought struggle. In China, the loyalty of the army,
or at least strategically deployed units of it, held sufficiently to
carry through the massacre of Tiananmen Square - an act from
which its prestige, and the legitimacy of the party that ordered
it, can hardly recover.
Where the repressive strategy succeeds, the population is in
for a long haul; the remnants of opposition are forced under-
ground, or take to the countryside to prosecute the armed strug-
gle. This is the prolonged rural route to revolution via guerrilla
warfare and liberated zones. Where the regime is overthrown,
however, the process of its delegitimation has one further stage
to run. This is the retrospective exposure of its past crimes, as
final justification for the breach of legality. The contrast between
the private opulence of the regime and the extent of popular
extortion and repression is now made fully available to public
view. There is the simultaneous opening up of the contents of
torture chambers and luxury villas; of security files and secret
bank accounts; of empty public coffers and burgeoning private
wardrobes. The ability to cover up acts of illegality, which served
to prolong the regime, now adds retrospectively to its discredit.
The importance of the loss of moral authority to regime col-
lapse, such as I have outlined above, may seem self-evident. Yet
it has to be asserted against those who would claim that legit-
imacy has no part to play, or at most a trivial or subsidiary one,
in any account of revolution. Among the most distinguished of
these sceptics is Theda Skocpol, who in her book States and
Social Revolutions advances what she calls an 'organisational' or
'realist' perspective on revolutions. According to her account it
is the loss of organisational capacity on the part of the state,
particularly under the pressure of external competition, working
to diminish the self-confidence of key elites, and with it the
repressive power of the state apparatus, that is the central
explanatory factor in revolution. This account she seeks to sus-
tain, on the one side, against traditional Marxist explanations of
revolution in terms of class conflict; and, on the other, against
those who 'treat the legitimacy of political authorities as an
important explanatory concept.' The relevant passage is worth
quoting at length:
Modes of Non-legitimate Power 219

'If state organisations cope with whatever tasks they already

claim smoothly and efficiently, legitimacy - either in the sense
of moral approval or in the probably much more usual sense
of sheer acceptance of the status quo - will probably be
accorded to the state's form and rulers by most groups in
society. In any event, what matters most is always the support
or acquiescence not of the popular majority of society but
of the politically powerful and mobilized groups, invariably
including the regime's own cadres. Loss of legitimacy,
especially among these crucial groups, tends to ensue with a
vengeance if and when ... the state fails consistently to cope
with existing tasks, or proves unable to cope with new tasks
suddenly thrust upon it by crisis circumstances. Even after
great loss of legitimacy has occurred, a state can remain quite
stable - and certainly invulnerable to internal mass-based
revolts- especially if its coercive organisations remain coherent
and effective.' (1979, pp.31-2).

The above passage illustrates the erroneous conclusions that

follow from an inadequate conception of legitimacy. Two points
can be made in response to Skocpol's basic line of argument.
Firstly, insufficient organisational capacity on the part of the state
is certainly one possible cause of a loss of political legitimacy. Yet
it only leads to a revolutionary situation, as opposed to some
other outcome, if it is associated with an erosion of belief in the
source of authority that underpins the system's rules of appoint-
ment. It is the link between the two, as I have argued - between
the failure to realise some basic purpose of government, and the
erosion in its source of authority, between the 'bad' and the
'wrong' - that makes a regime crisis a potentially revolutionary
one. Secondly, division within the state apparatus and its leading
cadres is certainly critical to a revolutionary outcome, and may
be engendered by a recognition of the regime's own failures. But
nothing brings such divisions to a head, or makes them so fateful,
as the delegitimating force of mass-mobilisation in defiance of
authority. In short, loss of moral authority induces organisational
incapacity every bit as much as incapacity contributes to a decline
in legitimacy.
Both these points are clearly illustrated from the example of
the Iranian revolution of 1979. It was manifestly no lack of
220 The Legitimation of Power

organisational or extractive capacity on the part of the Iranian

state, or of repressive capacity on the part of its military appar-
atus, that led to the Shah's downfall. The latter was the product
of socially damaging policies prosecuted by a regime whose
source of authority was insecure, and highly vulnerable to a
radical critique by intellectual leaders. And the division within
the military apparatus that precipitated the Shah's downfall was
manifestly not self-induced, but resulted from the delegitimating
effects of mass demonstrations. No doubt a certain relaxation of
repression may have given confidence to the popular forces to
seize the initiative, but the relaxation did not have its origin in
any organisational imperative. Indeed, it would be difficult to
find a clearer example of a loss of moral authority inducing
organisational incapacity, even paralysis, rather than vice-versa.
To be more pointed, the chasm between the enormous power of
the Shah and the justifiability of that power was what ultimately
rendered that power itself impotent. Skocpol's 'realism' is simply
not realistic enough at this point, since it does not take people
seriously as moral agents.
Revolutions are complex events, subject to multiple causation,
both long and short term, systemic and contingent, predisposing
and precipitating, etc. I have no wish to underestimate the role
that governmental failure, whether it be the product of internal
and external pressures, organisational incapacity, personal
incompetence, or whatever combination, plays in the creation of
a revolutionary situation; nor to discount the importance of intra-
elite division, especially among the military, to the final outcome.
Yet none of these are sufficient to make a governmental crisis a
revolutionary one, without the loss of moral authority that has
its origin in some weakness in the regime's source of authority,
and that is intensified by the impact of mass delegitimation. In
other words, Skocpol is correct in what she asserts, mistaken in
what she denies. And although her later analysis of the Iranian
revolution goes some way towards correcting her original dispar-
agement of the role of intellectual and cultural forces in the
revolutionary process, these are not integrated into a more gen-
eral account of legitimacy (Skocpol, 1982). The reason for this
is not far to seek. Skocpol associates the whole 'problematic' of
legitimacy with the Parsonian theory of the value-integrated
society, with its emphasis on social uniformity rather than diver-
Modes of Non-legitimate Power 221

sity, consensus over conflict, and its assumption that revolution

results from a self-conscious attempt to remove a 'dis-synchronis-
ation' between values and their environment (Skocpol, 1979,
pp.11-12). None of this follows if the starting point of legitimacy-
theory is located, not in some exaggerated conception of social
harmony, but in a theory of power, where it properly belongs; or
if we recognise that revolutions can be intellectually and morally
intelligible, without their outcomes being necessarily intended.
Both these points will become clearer through an analysis of the
process of relegitimation.

Revolutionary regimes and relegitimation

One aspect of relegitimation has already been discussed, and

that is the retrospective justification of the breach of legality
involved in the overthrow of the old regime. This primarily
negative task is relatively straightforward in comparison with the
positive task of establishing the legitimacy of the new order that
replaces it. Here the process of overthrow carries an ambivalent
legacy for the successor regime. On the one hand it derives
enormous legitimating force from the mass-mobilisations and sac-
rifices of the revolutionary period. As beneficiary of the popular
struggles it inherits all the moral prestige associated with their
achievements. On the other hand, the 'failures' of the old regime,
compounded by the dislocations of the revolutionary process,
make it exceedingly difficult for the new government to meet
the expectations of the population; the economy may be dis-
rupted, and the institutions of the state themselves may be in
disarray. There is a potentially serious discrepancy, in other
words, between the moral prestige of the new regime and its
capacity to accomplish the tasks with which it is confronted. The
prospects for its relegitimation will depend in part upon the
relationship between these divergent aspects of its inheritance.
The central issue in relegitimation, however, is the establish-
ment of a new principle of legitimacy, or source of authority,
that has sufficient popular basis to sustain the new rules of power.
Some writers argue that a new principle of legitimacy has to be
widely acknowledged in the pre-revolutionary society as a con-
dition for the delegitimation of the old regime; and that the
222 The Legitimation of Power

process of delegitimation itself therefore presupposes the alterna-

tive principle that will come to replace it. However, this is not
always so, or at least not in any straightforward way. People
may have a much clearer sense of the injustice of their exclusion
from the political process than of the precise form which their
inclusion ought to take. This will depend upon the context. In
many revolutions the collapse of the old regime only signals the
beginning of a fierce struggle between competing groups, each
striving to appropriate the revolution for their own alternative
project. This struggle may involve a conflict between different
principles of authority; or over how far a particular principle
should be extended; or between political and social definitions
of the revolution. It is precisely such struggles that make the
outcome of many revolutions unpredictable, since no single alter-
native is inscribed in the loss of legitimacy of the old regime.
Particular groups or individual leaders may have a clear idea of
the new order they envision; but the outcome may well be the
unintended result of a struggle between them under the unpre-
dictable conditions of spontaneous mass action on one side and
counter-revolution on the other.
The complexity of such processes defies any easy generalis-
ation. Even developing a typology of revolutions is problematic,
since each one will produce something different from its prede-
cessors. From the standpoint of relegitimation, however, a dis-
tinction can usefully be drawn between revolutions that are politi-
cally restorative, and those that are socially transformative. The
former, as the term implies, involve the return to a previously
established constitutional order, after a period of authoritarian
government or 'corrupted' democratic politics. Since the ideas
underpinning a representative system are already present within
the society's tradition, and what is involved in the demand for
a 'return to democracy' is widely understood, the need for intel-
lectual innovation and constitutional experimentation after the
overthrow of the old regime will be limited. Moreover, insofar
as the revolution is confined to a primarily political agenda, party
competition will be possible without threatening to tear society
apart, and free elections without the danger of handing power
back to the forces of counter-revolution. In restorative revo-
lutions, in other words, the task of relegitimation is a limited
one: the principle of authority underpinning the new rules of
Modes of Non-legitimate Power 223

power is already widely acknowledged, and the evolution of

'people power' from the mass-mobilisations of the revolution to
the electoral mode of legitimation is relatively straightforward.
Such an evolution also conforms to a desire to return to a less
politicised way of life after the excitement and intensity of the
revolutionary period.
Now there is a view that such revolutions do not count as
'real' revolutions at all (e.g. Kramnick, 1972). The point is often
made that, whereas until the middle of the eighteenth century,
the term revolution meant a turning full circle, a return to consti-
tutional order, the restoration of a relationship of trust between
rulers and ruled, etc., the distinctive feature of modern revo-
lutions, in contrast, has been precisely their radically innovative
character; and that this conception of revolution developed
alongside the Enlightenment view of history as unilinear, open-
ended, and above all progressive, rather than cyclical. There is
no doubting that many revolutions since 1789 have indeed confor-
med to such a conception. At the same time there is no reason
to abandon the older idea as characterising one type of revolution
that continues to the present day. The overthrow of President
Marcos in the Philippines exemplified this type, as did the popu-
lar mobilisations in Burma in 1988, had they proved successful.
And the Eastern European revolutions of 1989, although not
properly restorative, in that only Czechoslovakia had a convinc-
ingly democratic past, nevertheless shared some of the character-
istics of such a type, in that they represented the adoption of an
established norm, even if the norm was a Western European one
rather than deriving from each country's own past. Indeed, the
revolutions of 1989 signalled precisely the desire to end inno-
vation, and the readiness to adopt a common European politico-
economic model 'off the shelf' as it were. To say this is not to
underestimate the strains that the marketisation of the economy
will impose on the new democratic polities, nor the difficulty of
purging the apparatus of the state of the personnel and habits
of the past. It is simply to make a judgement about the character
of the revolutions themselves.
The point about politically restorative revolutions, then, is that
with an established principle of authority ready to hand, and
a straightforward transition possible to an electoral mode of
legitimation, the task of relegitimating the political order is a
224 The Legitimation of Power

relatively limited one. The matter is very different with those

that I have termed socially transformative, not least because
the trajectory of the revolutions themselves is so different. To
understand this trajectory we need to make a distinction between
two quite different ways in which revolutions can be socially
transformative. They can be so, first, because with the collapse
of authority of the old regime, and with the rights of exclusion
necessary to private property no longer enforceable, the people
may simply take possession for themselves of the estates and
factories in which they had previously worked. This could be
called the spontaneous social transformation 'from below'; and
it will happen where the principles underpinning the rules of
property as well as political power have become widely dis-
credited under the old regime. It should be distinguished from
the form of social transformation that takes place when a revol-
utionary party or leadership equipped with a socially transforma-
tive project wins state power, and succeeds in mobilising society
towards its own consciously chosen goals. This could be called
the organised social transformation 'from above'. Of these two
forms of revolutionary social transformation, the first can take
place without the establishment of the second (France) and the
second without the first (Iran); or they can take place successively
(Russia), or simultaneously, as in the piecemeal extension of
liberated zones through guerrilla warfare (China).
If we keep this distinction in view, two points can then be
made about the process of relegitimation of state power. The
first is a negative one. Wherever a socially transformative revo-
lution takes place it is very difficult if not impossible to establish
a settled parliamentary regime with electoral legitimation as its
outcome, even supposing the main political actors initially favour
it, and the popular basis of the revolution would seem to point
towards it. This is because of the degree of political resistance
that the social transformation provokes, and the consequences
that follow from its suppression. This is most evident where a
transfer of property rights has taken place. Any revolution
involving the dispossession of the propertied will provoke an
organised counter-revolution by the dispossessed, which will be
the more intense the larger their numbers. And counter-revo-
lution brings civil war, suspensions of legality, and terror, which
are hardly compatible with a settled parliamentary regime. Its
Modes of Non-legitimate Power 225

conclusion in turn produces widespread exhaustion, economic

hardship and disillusion that the fruits of revolution do not match
either its promise or the sacrifices that have been made. These
are hardly the circumstances in which an established revolution-
ary leadership will submit itself to electoral competition, with
the possibility, however remote, of losing power or seeing a
reversal of the changes brought by the revolution. Rather than
commit electoral suicide, it is more likely to consolidate the
authoritarian rule of the civil-war period.
The authoritarian tendencies provoked by counter-revolution
will be intensified where the successful party or leadership has
its own project of social transformation to effect. At the same
time, however - and this is the second point to be made -
such a project provides the necessary basis for the new regime's
legitimation. On the one hand its particular ideology of social
transformation will embody a principle of legitimacy to justify
its authoritarian structure of power. As we have already seen,
Marxism-Leninism assigned the leading role to the party hier-
archy in identifying the historically determined route to the
socialist or communist goal. In similar vein, the forms of Islamic
fundamentalism associated with Khomeini in Iran gave pride of
place to the ulama in overseeing the moral transformation of
society. What was required in addition was the dissemination
and consolidation of this belief system through the institutions
of formal and informal education, so that it achieved popular
implantation among at least a new revolutionary generation.
If one way in which the project of social transformation from
above serves also to legitimate the revolutionary regime is
through the source of authority it provides for a hierarchical
political order, a second is through the continued mass-mobilis-
ation that it requires. The transition from revolutionary mobilis-
ation to mobilisation for social transformation or reconstruction,
whether or not via military mobilisation for revolutionary
defence, forms an almost unbroken continuum. The agencies and
modes of mass activism that helped to bring down the old regime
- street committees, communes, factory councils, soviets,
religious assemblies, or whatever - can be readily adapted to the
work of civil defence and social transformation, albeit in a more
top-down and 'coordinated' manner. And the habit of political
activism will have been developed widely enough to survive the
226 The Legitimation of Power

inevitable exhaustion and desire for a return to normalcy. From

the standpoint of legitimation, continued mobilisation confirms
the popular and 'democratic' character of the new regime. The
electoral mode, by comparison, can be criticised for demobilising
the people, for stifling the process of social change, and 'freezing'
public opinion at a particular and arbitrary moment of electoral
choice.
If we compare a number of different socially transformative
revolutions in the light of the above discussion, then France
after 1789 can be seen to have experienced the vulnerability of
parliamentary rule in the face of counter-revolution and civil
war; but no group with an ongoing project of social transform-
ation was subsequently able to establish itself in power. It was
thus left to military dictatorship - to authoritarianism without
any legitimating principle of authority - to consolidate state
power, but leave the problem of its legitimacy essentially unre-
solved, as the history of the French state in the nineteenth cen-
tury amply demonstrated (cf. Richter, 1982). In Russia the
second form of social revolution - the Bolsheviks' project of a
heavy-industry-led economic transformation under Marxist aus-
pices - provided the ideological thrust for legitimising the recon-
structed state order. But the project proved antithetical to the
first, spontaneous social revolution - the peasants' redistribution
of land - and narrowed the party's base of mobilisation to the
urban working class, with fateful consequences for the character
of Stalin's subsequent regime. In China, the revolution's form
and timescale themselves ensured a much greater consistency
between the party's project for economic development and the
pattern of property redistribution in the countryside, since both
evolved together. The result was a much wider social base for
continued mobilisation, and a much more coherent model for
other peasant revolutions in the Third World than that provided
by the Soviet Union, for all the latter's prestige as the first
communist regime. In the Iranian revolution there was no change
of property rights, since the social transformation was primarily
cultural and spiritual in purpose, though its internal and external
crusades had significant material consequences, as well as an
identifiable social base among those who had benefited least from
the Shah's modernisation programme. Here it was the leadership
position that Khomeini came to hold over the opposition in the
Modes of Non-legitimate Power 227

year preceding the revolution that turned the movement against

the Shah from one of constitutional restoration to one of social
transformation, thereby excluding a subsequent liberal-demo-
cratic development with free electoral competition. This was a
revolution whose outcome was determined very largely at the
ideological level, rather than by the exigencies of counter-revo-
lution and civil war, and whose legitimacy was more solidly
grounded in traditional beliefs and institutions than the others
mentioned.
The contrast I have drawn between politically restorative and
socially transformative revolutions, then, is a contrast between
typical political outcomes, and their respective agendas for the
relegitimation of state power. The former result in competitive
multi-party systems, with electoral legitimation and limited ideo-
logical innovation. The latter produce systems of authoritarian
rule that require for their legitimacy the commitment to an on-
going project of social transformation, carrying with it both
extensive ideological activism and continued mass-mobilisation,
typically under the hegemony of a single party.
Not every example of revolution fits neatly into one or other
of the above categories. The Nicaraguan revolution, for instance,
was a bold attempt to embrace the two. On the one hand it
embodied a determined project of social transformation, not so
much in that there was widespread redistribution of property,
except of Somoza's extensive estates, but in terms of continued
mass-mobilisation around grass-roots health and welfare pro-
grammes, popular defence and education for a new civic ideal
(Coraggio, 1986). On the other hand the ruling front, in explicit
repudiation of the orthodox communist model, proved scrupu-
lous in preserving parliamentary legality and electoral compe-
tition in face of a ruthless counter-revolution and economic
blockade, both orchestrated by the US government (Harris and
Vilas, 1985, chs.8-10). Indeed the outcome of the 1990 election
in Nicaragua seemed to provide retrospective justification for all
those earlier revolutionary leaderships who had refused to submit
themselves to electoral competition on the grounds that it would
jeopardise the social gains of the revolution as well as their
own position; and that parliamentary democracy and progressive
social transformation in a developing country are indeed mutually
exclusive.
228 The Legitimation of Power

What the Nicaraguan experience most clearly demonstrates, in

fact, is the role of external forces, of crude intervention as well
as ideological attraction, in the outcome of modern revolutions.
In particular, Nicaragua's transition in 1990 under US pressure
from a socially transformative revolution to a purely restorative
one ('making the country safe for democracy', as even Daniel
Ortega described the achievement) indicates the lengths to which
the USA will go, even in defiance of international law, to stifle
any revolutionary process that does not conform to the model
of its own Lockean revolution, or that attempts to alter the
balance of economic inequality in favour of the poor and the
dispossessed. It will do so even in favour of right-wing dictator-
ships if necessary, though to be sure it will also later prod such
dictatorships in the direction of democratisation, once any pros-
pect of a progressive social transformation is safely over (Boros-
age and Marks, 1976). Given the importance of the USA as the
world's sole remaining superpower, the consistency of its record
in this regard has obvious significance for the fate of any future
revolutions.

Coups d'etat and military regimes

In this section I shall examine the place of legitimacy in an

understanding of coups and military governments, and their dif-
ference from revolutions and revolutionary regimes. At the
beginning of Chapter 5 I pointed to one of the central dilemmas
inherent in state sovereignty: the absence of any superior or
external coercive power which might deter the armed forces of
the state from breaching legality and usurping power themselves.
The factors restraining them must therefore be internal ones.
They are found in the organisation of state institutions, in the
relations established between the civil and military powers, above
all in the degree of legitimacy of the political order. Of these
the last is the most crucial. As I noted in Chapter 5, the proce-
dures and traditions that encourage the subordination of the
military to the civilian power may delay, but cannot ultimately
prevent, a military intervention when the civilian power has lost
its authority, and the constitutional rules prove inadequate to
resolve the crisis. If the legitimacy of government is important
Modes of Non-legitimate Power 229

to securing the cooperation of society at large, therefore, tt IS

above all so in ensuring the subordination of the military to the
civilian authority. From this it follows that, where the military
do seize power, it does not happen 'out of the blue', but only
as the culmination of a process of erosion in the legitimacy
of the civilian order, whose warning signs are evident well in
advance.
Such an erosion provides both a prime reason for military
intervention, and the means of justifying the breach of legality
after it has taken place. Here it is not only the erosion of
legitimacy itself that is important, but the contrast that can be
drawn between the deficiencies of the civilian regime on the one
side and the distinctive self-image and operating style of the
armed forces on the other. As I have already suggested, the
context in which coups take place is primarily one in which
the civilian regime has lost public confidence in its ability to
sustain some basic purpose of government; the crisis, in other
words, is one of government performance, and usually govern-
ment performance alone. It is also one where the military believe
that their distinctive qualities and mode of organisation are such
as to provide a resolution of the crisis. The contribution that the
theory of legitimacy can make to the study of coups, therefore,
lies in the connection to be drawn between the form of legitimacy
deficit and the distinctive character of the military in its particular
context.
In his book Political Order in Changing Societies, Samuel
Huntington coined the term 'breakthrough coup', which many
others have adopted, to identify a particular type of coup occur-
ring in terminally weakened monarchical or aristocratic political
systems. Here the military, equipped with a secular education
and trained in the use of sophisticated technology, constituted
the chief bearers of the ideology of modernisation in opposition
to the traditionalist values of the regime, and the chief advocates
of the career open to talent against the privileges of birth and
heredity. When the civilian regime faltered, they put themselves
forward as the representatives of progressive social forces, and
as agents of an urgent programme of economic modernisation.
Such, Huntington argues, were many of the coups in Latin Amer-
ica in the 1920s and 1930s, and in the Middle East in the 1950s
(1968, ch.4).
230 The Legitimation of Power

However adequate Huntington's view of the military as a 'mod-

ernising' force may have been at the time, such coups must now
be regarded as a feature of the past, if only because there are
so few monarchical or aristocratic regimes left to overthrow.
Most coups nowadays take place in democratic political systems,
and need to be understood in the light of the problems typical
to those systems. Although the circumstances vary enormously,
what is primarily at issue here is not the source of authority of
the regime, or the belief in the principle of popular sovereignty
itself, which underpins the rules of the democratic order; it is
rather problems of government performance, of weakness, cor-
ruption or divisiveness, as these are exacerbated by typical demo-
cratic procedures, or appear unresolvable by them. In other
words, it is the very institutional arrangements that contribute
to the legitimation of government in a liberal democracy - the
freedoms of expression and association, electoral competition,
majoritarian procedure - which themselves become the obstacle
to resolving the crisis of government performance, and thereby
lose public confidence. (For general accounts of democratic
breakdown see Linz and Stepan, 1978, Nordlinger, 1977; for
Latin America, Collier, 1979, Philip, 1985; for Africa, First,
1970).
Of the different problems mentioned above, that of govern-
ment weakness, or inability to fulfil the basic economic or secur-
ity purposes of the state, is most damaging where a government
is confronted by powerful groups taking advantage of the free-
doms of expression and association to press their claims. Where,
for example, chronic inflation is exacerbated by a government's
inability to resist the demands and sanctions of organised eco-
nomic interests; or where social conflict is intensified by press
agitation and popular mobilisation; the origins of economic and
social disorder will be attributed, less to deep-seated socio-
economic causes, than to government weakness on one side and
the opportunities for disruption afforded by political freedoms
on the other. At such a juncture, the imposition of military
'discipline' and the suppression of civil liberties may come to
seem, if not a positive virtue, then at least a regrettable necessity.
The phenomenon of corruption, as I suggested earlier, can
occur under any type of regime. However, it is likely to be
much more visible in a democratic system, where publicity is not
Modes of Non-legitimate Power 231

controlled by government. And in a society where the state

provides a major means of access to economic opportunities
or resources, the process of electoral competition will generate
manifold obligations to be honoured from the spoils of victory,
as well as powerful incentives to fraud on the route to office.
Instead of elections providing the means of resolving government
failure and restoring the authority of government, therefore, they
become identified as themselves the prime cause of the problem.
At this point the military offers the image of 'clean hands', and
the military coup seems to provide what the electoral process
should but cannot - the removal of a discredited government.
Here is the coup, less as eliminator of civil liberties than as a
form of substitute election.
A different source of military intervention from that of either
government weakness or corruption is where the policies of a
democratically elected government come to threaten the interests
of powerful minorities, especially the propertied, who can enlist
the sympathy of the military in their cause. Here it is not initially
government weakness that is a problem, but rather its determi-
nation to promote the interests of disadvantaged and expectant
supporters. Seeking to rectify injustices is always politically more
divisive than perpetuating them. However, it is rare for the
military openly to intervene on behalf of a privileged minority
as such, however threatened, and so risk appearing divisive itself.
The propertied will have to prepare the ground first for the
destabilisation of the regime by means of the economic resources
and external contacts available to them. Runs on the currency,
investment strikes, the withdrawal of funds, disruptions of trade:
any of these may serve to undermine confidence and precipitate
an economic downturn, which will adversely affect the govern-
ment's own supporters and bring them into conflict with its poli-
cies. In an atmosphere of general crisis, where the regime is
assailed from many directions simultaneously, the military will
more readily be able to present themselves as an instrument of
national salvation, and their intervention as serving a general
interest that the elected government has jeopardised through its
divisive policies.
The above are only some of the typical scenarios for a coup
d'etat. At the root of their justification lies not only an inability
of democratic processes to resolve a crisis of government per-
232 The Legitimation of Power

formance, but a corresponding rationale for military rule. Two

elements in this rationale can be distinguished. One is the estab-
lishment of a mode of government which 'solves' the problematic
features of democratic politics at a stroke, by simply abolishing
them. Military rule reasserts the authority of government over
society by removing the freedom of organised groups to pursue
their interests independently of the state; it imposes unity on a
divided nation by removing the political avenues for competition
and conflict; it 'abolishes' fraud and corruption by eliminating
every occasion for electoral pressure and manipulation; it
restores confidence in the ability and integrity of government by
removing every independent means for discovering and publicis-
ing what it is up to. In short, the military resolve the problems
of democratic politics by abolishing politics altogether; they
immunise the state from the problems of society by elevating the
state above society.
Along with a depoliticised mode of government the military
offer, secondly, a public image of themselves that is the opposite
in every key particular to that of politicians. The armed forces
traditionally pride themselves on being a national institution,
drawn from all sectors of society, and embodying a national
interest above the sectionalism of political parties. They are
trained to action rather than debate, for getting things done
rather than making speeches, for taking decisions on technocratic
grounds rather than through the wheeling and dealing of poli-
ticians. Above all, they embody discipline and order, in contrast
to the untidiness and disorder of politics. Such virtues become
all the more attractive, the more incapable politicians appear of
resolving chronic problems of society through political means.
It is one thing, however, to justify a cuup d'etat retrospec-
tively; quite another to secure the legitimacy of military rule on
an ongoing basis. The dilemma of the military is that the ration-
ale they offer for breaching constitutional legality and dispensing
with the rule of politicians, not to mention the human rights
abuses they may commit in the process, make it impossible for
them simply to return power to a civilian government after a
brief interlude, even should they wish to do so. The task of
making society once more 'safe for politics' almost always turns
out to be a protracted one, which requires the establishment of
military rule as an ongoing system of government. Such a govern-
Modes of Non-legitimate Power 233

ment may come to enjoy a certain internal stability, in the sense

that, where there is sufficient cohesion among the armed forces,
it may be immune from further coups. It may over time achieve
a certain social basis of support, for example among the proper-
tied. It may acquire a civilian fa~ade, through the incorporation
of carefully vetted politicians. But what it can never acquire is
legitimacy.
The reasons for this are twofold, and can be deduced from
the analysis of legitimacy so far developed. First, the rationale
for military rule is to be found in a crisis of performance of the
democratic regime it displaced; and its own justification rests on
the promise of resolving that crisis. In contrast with revolutions,
the basic source or principle of authority - that of the people,
or popular sovereignty - has not itself been discredited. A loss
of public confidence in certain institutions and procedures of
liberal democracy there may have been, but not an erosion of
belief in the principle of popular sovereignty itself. Moreover,
the military regime has no alternative source of authority to offer
in its place, that might underpin its own rules of appointment.
Indeed, it acknowledges the principle of popular sovereignty by
claiming to 'represent the people', to 'embody the national
interest' more effectively than the politicians themselves have
done.
In assessing such claims we should recall the important distinc-
tion between the 'good' and the 'right': between promoting
national ends and deriving authority from the nation; between
serving the people and representing them. The military may do
the former, but cannot do the latter. To do the latter requires
just those institutions of representation and election that the
military have seized power to dismantle. As a result they are
compelled to equate the good with the right: to make serving
the ends of government equivalent to an entitlement to office,
and promoting the nation's interests the same as deriving author-
ity from the people. Justifications for military rule, in short, rest
upon a conceptual and normative elision, which may serve to
confuse people for a while, but cannot make good a basic
deficiency in a crucial dimension of legitimacy.
The first weakness of a military regime, then, is that its rules
of office, if it has any, cannot be justified in terms of a recognis-
ably valid source of authority. The second is that it is not subject
234 The Legitimation of Power

to any public legitimation in terms of expressed consent. Its

ongms are not to be found in mass-mobilisation, as with a
revolutionary regime. On the contrary, power has been seized
precisely to exclude the people from politics, and to stifle the
mass-participation that is necessary to political legitimation in
the era of popular sovereignty. The military regime, therefore,
in both origin and purpose, is a non-legitimated one.
Now it does not follow from this that the military can have
no social basis of support, especially among the privileged or the
propertied. Here again, however, we need to observe another
important distinction: between enjoying support, and possessing
the institutional means for the regular expression of consent. The
military may construct a form of oligarchical order, with privi-
leged access for its supporters to influencing policy. But there is
no publicly justifiable rule that can limit political participation to
one section of the population only, however it is defined. In
abolishing formal political rights for one, they have to be abol-
ished for all. In so doing, the military regime necessarily deprives
itself of the means of any public legitimation.
The distinctions drawn above are essential if we are to under-
stand the lack of legitimacy of military regimes. Nowhere in fact
is it more important to analyse the internal structure of legit-
imacy, if we are not to be taken in by the 'legitimations' of the
powerful. If we operate with a purely subjective conception of
legitimacy as whatever people believe to be legitimate under the
propaganda influence of government; or if we define legitimacy,
with Lipset, as the ability of a regime to convince people that
its institutions are the most appropriate ones for the society;
then there is no reason to suppose that military rule might not
become as legitimate as any democratic system. After all, society
needs strong government to maintain order, eradicate corruption,
preserve national unity, promote economic development, and so
forth, does it not? Strong government a military regime can
supply in good measure. If people acknowledge this need, does
this not by definition guarantee the military's legitimacy?
The simple answer is 'no'. The legitimacy of any political
system rests, like a tripod, on three legs. The first is the acknowl-
edged source of authority underpinning the rules of appointment
to the leading offices of state. The second is the capacity of its
institutions to facilitate achievement of the ends or purposes of
Modes of Non-legitimate Power 235

government. The third lies in the mechanisms for the regular

expression of consent. A military regime rests on the second of
these alone; it is a tripod supported on one leg. To achieve this
balancing feat at all, the leg has to become grossly enlarged.
Performance is everything, and the achievements and capacities
of the regime have to be disproportionately magnified. Lacking
any title to rule, the regime has to affirm its unique qualifications
to fulfil the country's historic destiny. Lacking any authorisation
from the nation, it has to stimulate every nationalist sentiment
in its favour. Lacking any manifestation of popular consent, the
leader's charismatic qualities have to be repeatedly proclaimed.
The basic problem of regimes whose legitimacy rests 'on one
leg', however, and depends on performance alone, is that they
are highly vulnerable in the event of policy failure. Military
regimes may be better able than civilian ones to suppress dissatis-
faction with their policies. Yet failure of performance removes
the only justification for their rule, since they lack any valid
source of authority. Moreover, since there is no way of distin-
guishing between the particular government and the system of
military rule as such, in the absence of any legalised mode of
succession, any failure will discredit the regime as a whole (see
Clapham and Philip, 1985; O'Brien and Cammack, 1985).
This fundamental weakness drives military regimes in one of
two directions. One is the search for ever new policy initiatives
where they can secure some easy and preferably dramatic suc-
cess, to cover up failure in other areas together with any internal
opposition it may have provoked. Since they are military
regimes, this typically involves military exploits and external
adventurism, such as recovering disputed territory on their bor-
ders. So the Greek Junta tried to bring down President Makarios
in 1974 to further the unification of Cyprus with Greece; Idi
Amin's troops made incursions into Tanzanian territory in 1978;
the Argentinian generals invaded the Falkland Islands in 1982;
Saddam Hussein attacked Iran in 1980, and annexed Kuwait in
1990; and so on. In the first three examples the adventurism
provoked military retaliation that brought the regimes tumbling
down. The outcome of Iraq's invasion of Kuwait is at this
moment in doubt (1990), but may well follow a similar pattern.
The other course for a military regime to take is to seek a
more substantial legitimacy through a return to electoral politics.
236 The Legitimation of Power

Pressures also work in this direction from external 'sponsors',

responding to adverse publicity about human rights abuses. In
many cases there is a period of transition first, involving an
attempt to secure legitimation without the conditions necessary
to it. This is the period of elections without free choice: of single
parties constructed entirely from above, of plebiscitary votes
('yes' or 'no' to the leader), of experiments in 'indigenous' forms
of local democracy which do not affect the centre, and so on
(Bienen and Morell, 1976). How long this period lasts depends
upon the extent of the pressures, both internal and external, on
the one side, and how far the military judge society to have been
made safe for electoral competition on the other. In any case
they will always be there behind the scenes when the freedoms
of speech and association have been restored, to limit the
'excesses' of politicians. Yet the logic of a formal surrender of
power is implicit in the character of military rule itself.
Here, in conclusion, lies the key point of difference between
revolutionary and military regimes. Revolutionary regimes derive
enormous moral authority from the mass mobilisations and popu-
lar sacrifices of the revolutionary period. Their institutions
embody an acknowledged principle of legitimacy. And they are
subject to continued legitimation whether in the electoral or the
mobilisation mode, or, more rarely, through a combination of
the two. Military regimes, in contrast, at least those that have
their origins in a breakdown of democratic politics, can at most
possess a temporary or transitional legitimacy. That is to say,
their only justification lies in the necessity of overcoming a per-
formance crisis of civilian government, of resolving an impasse
which the political system has been unable to resolve, so as to
make a return to civilian rule on a sounder basis possible. This is
the temporary or conditional legitimacy indicated by the Roman
concept of dictatorship, meaning the temporary suspension of
constitutional rule in an emergency so that the conditions could
be created for its more secure restoration. Of course the military
find it necessary to stay much longer than they first intended, or
at least they convince themselves that they must. But the longer
they stay, the more problematic becomes their lack of any source
of authority or any popular legitimation, despite all the 'legitim-
ations' they may themselves contrive.
Modes of Non-legitimate Power 237

Conclusion: the dilemmas of political legitimacy

The analysis of the processes of political breakdown, and of the

relegitimation, or attempted relegitimation, of revolutionary and
military regimes, completes the agenda I proposed at the begin-
ning of Chapter 5. It also reinforces the question I raised there,
namely, why is legitimacy so hard to achieve for the contempor-
ary state? By way of conclusion I shall draw together the different
elements necessary to an answer.
For all its enormous and evident power, the contemporary
state is a highly problematic structure in those respects necessary
to its legitimacy. First there is what I have termed the vulner-
ability of sovereignty: as the supreme law-making and law-enforc-
ing body, which legitimises all other powers within society, there
is no higher authority than the state to guarantee its own consti-
tutional rules in the face of intense struggle to exercise and
influence its sovereign power. Its rules therefore remain vulner-
able unless they are anchored in the most fundamental source
of authority acknowledged and agreed within the society. This
is the first condition of rule-justifiability to be met.
Secondly, as the power-structure expressly devoted to promot-
ing the public interest, the state has to satisfy the most general
societal needs for security and economic sufficiency, and in a
way that is not manifestly partial or biassed. It has to do this in
a context where, externally, it cannot control many of the forces
that affect its security and economic performance, and, intern-
ally, the society it governs is divided along class, ethnic, regional
and other lines. Whereas all states are equal in the vulnerability
of their sovereignty, they vary enormously in their capacity to
moderate the external forces to which they are subject, and in
the pattern and intensity of their internal societal divisions.
Thirdly, the state has to secure express consent to its power,
in a world where all adults are considered entitled to be members
of the political community, and therefore to have the right to
the minimum political participation necessary to express that
consent. This formal equality of political status has to coexist,
and be rendered compatible with, inequalities of power in all
the dimensions of social life, the political included.
Legitimate political systems comprise legally established and
validated rules for the organisation and distribution of power,
238 The Legitimation of Power

for appointment to office, and so on, which can meet these

three central requirements: which embody an ultimate source of
authority acknowledged within the society; which are capable of
resolving performance problems and dealing with government
failure; which provide for express consent on the part of the
population as a whole. These criteria are not easy to fulfil simul-
taneously, and it is not surprising that there is only a limited
number of types of political system capable of meeting them. By
way of conclusion I shall distinguish three broad types, each of
which has its distinctive strengths and weaknesses.
One is a traditional monarchy, with popular representation
confined to a legislative assembly. Jordan and Morocco provide
better examples here than Saudi Arabia, whose lack of all popu-
lar representation constitutes the most vulnerable point in its
political arrangements. This political type combines two key
advantages of the traditional order - that the supreme executive
office constitutes a focus of social unity beyond competition, and
that political power formally reflects the structure of social and
economic power - with the minimum requirement for popular
consent through elections to a legislative assembly. Such a system
cannot be generally replicated, however, for the simple reason
that traditional institutions cannot be created; they can only
survive. Mostly they have not done so, partly_ because of the
disruptions of Western colonialism, partly because the hereditary
principle cannot withstand erosion from the principles of equal
opportunity and popular sovereignty in the circumstances of the
modern world. The remaining monarchies are therefore vulner-
able to the assertion of these principles within their own societies,
and to agitation on their behalf from without.
Second is the mobilisation regime, whose opening up of careers
to those from below and continuous mass-politicisation have their
origins in a revolutionary process. A central place in such regimes
is occupied by the monopolistic belief system, which, through
the agency of a dominant party, defines the collective goal for
society and provides the stimulus to mass-mobilisation, as well
as the source of authority for the ruling group. The strengths
and weaknesses of such regimes cannot properly be identified
apart from the specific logic of their respective belief systems,
since it is these that provide their distinctive dynamic and traject-
ory; to treat communism, fascism and Islamic fundamentalism as
Modes of Non-legitimate Power 239

equivalents would be absurd. Yet we can note common features

of their respective regimes. Most important is the way in which
the belief system provides for all three criteria of legitimacy
simultaneously: as source of authority for the rules of power; as
authoritative definition of the common interest or societal goal;
as stimulus to consent through mobilisation in the regime's ser-
vice. Here lies the source of this type's strength and weakness
simultaneously: its strength, that the belief system can accomplish
all this; its weakness, that, in doing so, it is required to do too
much. The regime is thus vulnerable to forces making for the
erosion of belief; its repression of alternative beliefs and diver-
gent modes of expression, which is neccs'iary to its legitimacy,
proves corrosive over the longer term; and its claim to privileged
knowledge of society's goal cannot withstand the independent
assertion of their interests by the people themselves, once they
are sufficiently determined to do so.
A liberal-democratic system, the third type, avoids the weak-
nesses of both the traditional and mobilisation regimes in the
unqualified expression it gives to the principle of popular sover-
eignty; in the tolerance allowed to a variety of beliefs that is the
necessary counterpart to the electoral mode of consent; and in
its acknowledgement that the ultimate test of the public good is
what the people, freely organised, assert, not what the
interpreters of tradition or doctrinal orthodoxy may determine.
Its key points of vulnerability lie in the fact that electoral compe-
tition for the supreme executive office is liable to intensify
societal divisions, and that the tension between formal equality
of electoral power and socio-economic inequalities can become
unmanageable; how far either becomes critical depends upon the
particular institutional forms on the one side, and on the intensity
of social divisions on the other. The problems may be moderated,
as we have seen, by various devices, formal or informal, for
limiting the scope of electoral choice; the corresponding weak-
ness, however, is that this reduces the legitimating force of elec-
tions, and their value in removing politicians who have failed,
and in restoring governmental authority. These semi-legitimated
systems are vulnerable to popular mobilisation outside the frame-
work allowed, in a way that liberal-democracies proper are not.
In contrast to these legitimate types of political system, military
dictatorships constitute the limiting case of a non-legitimate pol-
240 The Legitimation of Power

itical order, in that, however justifiable military intervention may

be, they lack the key characteristics of an authoritative source
for their power, and popular consent in either the electoral or
the mobilisation mode. Their reliance on performance alone
explains their typical evolutionary pattern, either towards
destructive adventurism, or a gradual return to electoral legitim-
ation. They should thus be seen as an interim, transitional or
exceptional type of system, even though the 'transition' may in
practice prove long drawn-out.
Now if we compare the three types of legitimate political
system analysed above, it is evident that their respective weak-
nesses and points of vulnerability are not just different, but
qualitatively so. The traditional order is vulnerable to demands
for popular participation, and for equal opportunity to compete
for political office. Mobilisation regimes are vulnerable through
their exclusion of the population, not from access to office, but
from any influence on policy or the choice of leaders, and
through their suppression of independent means of popular-will
formation. Both traditional and mobilisation types, that is to say,
are vulnerable to the force of popular sovereignty, and to the
erosion of the belief system that validates a paternalist definition
of the public interest. Liberal democracy, in contrast, is not
vulnerable to the erosion of its basic belief system; its problem
lies in obtaining the social conditions and institutional forms in
which its principles can be realised in a non-conftictual manner.
The former types, in other words, are vulnerable to the collapse
of their underlying principle of authority; the problem for liberal
democracy is to achieve the conditions to enable its principle of
authority to be realised.
Such a contrast can only be made sense of by arranging the
three types in a developmental series. A mobilisation regime is
more advanced than a traditional one in its opening up of politi-
cal office and its expansion of popular participation. Liberal
democracies represent a higher stage than either (although not
necessarily a 'final' one) in the extent to which they realise the
principle of popular sovereignty, admit a non-paternalist defi-
nition of interests, and allow the possibility of consent to the
rules of power through their revisability by electoral process. To
posit such a developmental series is not to indulge in historical
determinism. Traditional systems may develop into liberal-demo-
Modes of Non-legitimate Power 241

cratic ones, with or without a mobilisation regime as intermedi-

ary; each historical transition may in turn take place with or
without the interposition of a military dictatorship. The signifi-
cant point is only that, once a given type of legitimate order is
attained, it is impossible to regress to an earlier one.
It follows that both traditional and mobilisation regimes await
to be superseded by a further legitimate political order. Liberal
democracies, on the other hand, though certainly open to
improvement in the extent to which they realise their underlying
principles, can only be overthrown by force. This explains why
the former are both vulnerable to revolutionary movements from
below; the latter only to coups d'etat from above. Among the
most significant features of the political evolution in the USSR,
and of the revolutions of 1989, is the confirmation they provide
for such a developmental analysis. Viewed from the historical
vantage point of the end of the twentieth century, the legitim-
ation problems of traditional and mobilisation regimes can be
seen as problems inherent in a lower stage of development,
awaiting to be surpassed; those of liberal-democratic regimes as
problems of how the conditions necessary to a higher stage can
be attained and consolidated.
One drawback to a developmental analysis such as I have
proposed is that it encourages those who live in a secure liberal-
democratic order - which is by no means in practice the final
realisation of democratic principles - to assume that the failure
of other societies to sustain such an order is because their popu-
lations are not ready for it, are not 'mature' or 'advanced'
enough; that they still require a paternalist definition of their
interests. Such an assumption has historically provided all too
convenient a legitimation for imperialist and nco-imperialist
interventions in other societies. Yet once a people has asserted
the principle of popular sovereignty for itself, it is primarily
structural or contextual conditions, so I have argued, not cultural
ones, that frustrate its realisation in a settled democratic order.
Culturalist explanations, besides being themselves paternalistic,
provide an alibi to those in the West from acknowledging that
they share some of the responsibility for the persistence of socio-
economic conditions in the developing world that are adverse to
democratisation. For such conditions to be overcome would
242 The Legitimation of Power

require the emergence of a more just and sustainable inter-

national economic order as its necessary condition.
8 Legitimacy in Political
Science and Political
Philosophy
It remains in a brief concluding chapter to draw out the impli-
cations of my account of legitimacy for the relationship between
political science and political philosophy. In the first chapter I
referred to the extraordinary disjunction between the two within
Anglo-American departments of politics in their respective treat-
ments of legitimacy, and in the different types of literature that
students are expected to read for each. Such a disjunction is
symptomatic of a more general separation between the two, that
has its intellectual rationale in the distinction between facts and
values, and in the different purposes of explanatory and norma-
tive theorising respectively. It is a separation that has been pro-
gressively reinforced through the twentieth century by the
demands of the academic division of labour.
There can be no denying that the activities of explanatory and
normative theorising are in principle distinct, since explaining
what is differs from exploring the grounds for what ought to be;
nor can it be denied that observing the fact-value distinction is
a necessary discipline if social scientists are not to intrude their
own values into the situation under investigation. Yet such con-
siderations do not entail the separation of the two activities,
when they are also closely interconnected. The costs of such
separation are evident in the difficulty many social and political
scientists experience when it comes to handling the normative
dimension of social relations. Either they discount it altogether,
as in 'realist' theories of power. Or they treat values in a
reductionist way, as deriving their force from psychological or
pre-rational motivations. In doing so, they forfeit any under-

243
244 The Legitimation of Power

standing of the logic of reasons at work when people follow

rules, keep obligations or seek to realise their ideals; and of that
logic of institutional arrangements which derives in part from
the distinctive character and requirements of their legitimating
principles and procedures. Such 'logic' can only be grasped by
an internal analysis of ideas and arguments, of the kind that is
central to the practice of normative philosophy.
On the other hand the work of normative philosophy cannot
proceed in isolation from empirical analysis, for a number of
reasons. We only have the moral and political principles we do
because the world is as it is, and not otherwise. Self-evidently,
if we were invulnerable to a stab in the back or a bullet in the
head, then the prohibition against killing would lose its point; if
we did not need food and shelter for survival, disputes about
the principles by which the social product should be distributed
would lose their force (Hart, 1961, ch.9). Less self-evidently, it
is because we share some capacities as well as needs in common
that democracy can be justified; and because we also have differ-
ent capacities, and to different degrees, that competitive recruit-
ment to those positions of power that are socially necessary is
justifiable. Moreover, what positions are socially necessary in the
first place is a question that social-scientific evidence is needed to
answer; just as it requires such evidence to identify the conditions
necessary to the realisation of the principles or ideals that we
espouse. In these and other ways empirical evidence is required
for the solution to normative problems (Runciman, 1965). So
the work of normative philosophy and that of social science can,
neither one, proceed in isolation from the other.
If there is anywhere that we might expect the interconnect-
edness of the two activities to be most clearly apparent, it is in
the analysis of legitimacy. Evidence of the connection is provided
by the fact that I set out at the beginning of this book to write
a social-scientific account, and ended up perforce in providing a
philosophical one, not as a deviation from the main enquiry, but
in order to fulfil its explanatory purpose. It will be worth recalling
the different points where this happened, since they suggest a
rather different account of the relationship at each stage.
First was the point in analysing the concept of legitimacy when
I identified a common normative structure underlying the variety
of historically specific legitimations, and explained this structure
Political Science and Political Philosophy 245

in terms of the different ways in which power offends our moral

sense, requiring in turn a principle of differentiation, a concep-
tion of common interest, and evidence of consent from those
qualified to give it. With regard to political power, I argued that
its distinctive legitimating requirement of being grounded in the
most fundamental source of authority acknowledged within the
society, derives from its vulnerable position as final adjudicator
and guarantor of positive law. Given this common normative
structure of legitimacy, however, social science and political
philosophy relate to it in different ways. For the former, I
argued, it provides a guide to identifying the specificity of particu-
lar historical forms of legitimation, and to making judgements
about legitimacy-in-context, as part of an explanatory enterprise.
For the latter it provides the framework for elucidating the prin-
ciples and conditions of an ideal legitimacy that transcends the
constraints of given power relations, as part of an essentially
normative purpose. For both, the recognition of legitimacy as
multi-dimensional should counteract any tendency to reduce it
to a single dimension, whether in social science to reduce it to
mere belief, or in political philosophy to posit a complete anti-
thesis between social-contract and utilitarian modes of analysis.
Now I must acknowledge at this point that the account of
political philosophy I have given above, and previously in the
book, is not the only one that can be given. There is, for exam-
ple, a more limited, conservative definition of political philo-
sophy which lies much closer to the activity of social science, in
that it regards the norms and traditions of particular historical
societies as constituting the boundary of meaningful theorising.
According to this approach, the task of political philosophy is
to explore the intimations or implications of a given tradition
(Oakeshott, 1962). However, there are substantial problems to
limiting political philosophy in this way, which parallel the prob-
lems of a social science that is limited to a Winchian programme,
and revolve around the same point: how to deal with changes
in, or challenges to, established norms of legitimacy, whether
from within or without the society. Just as social scientists cannot
explain these if they remain confined within the terms of a par-
ticular set of society-specific norms or beliefs, so neither can
philosophers evaluate them, except from a standpoint which is
outside the given tradition, and from which they might prioritise
246 The Legitimation of Power

particular elements within it. For this reason I do not see how
it is possible to avoid some external or transcendent position
altogether in political philosophy, even if it remains implicit and
unconscious.
One significant connection, then, between political science and
political philosophy with regard to legitimacy lies in their pos-
session of a common concept, with a common structure, which
they deploy for different purposes. There is a second level of
analysis, however, where they can be seen to share, not only a
common concept, but a common standpoint as well. As I argued
in Chapter 4, when it comes to explaining, not people's
behaviour, but legitimacy itself, and how the beliefs and the
consent that comprise it are maintained and reproduced within
established relations of power, the social scientist as well as the
political philosopher is compelled to adopt a standpoint that
transcends these self-confirming processes of legitimation; and to
confront the question as to whether there are some conditions
or principles in which this self-confirming cycle is less closed
than others, and what these might be. In other words, only a
comparative analysis from a position outside particular historical
power relations will suffice. And this position is identical to that
of the normative philosopher who has to stand outside all power
relations to establish philosophically valid principles of legit-
imacy, and to identify the conditions under which consent to
power rules is truly voluntary because it is unconstrained by their
effects.
This common standpoint shared by explanatory and normative
theorising about legitimacy is most evident in the work of Rous-
seau, who, in his Discourse on the Origin of Inequality showed
how legitimacy is historically established within pre-existing
relations of power and inequality; and in his Social Contract
demonstrated that the only way to escape this pre-structuring of
consent and of the common interest, to the advantage of the
powerful, is through a legislative assembly of all citizens, in
which power rules are agreed to and revised from a position of
equality. Despite the oversimplification in the historical part of
Rousseau's analysis, his work is particularly significant for my
argument that an explanatory account of how legitimacy is con-
structed, and a normative account of how it should ideally be
so, are logically connected by the common intellectual standpoint
Political Science and Political Philosophy 247

they share; and that it is reflection on the empirical processes

through which legitimacy is reproduced, that leads to the dis-
covery of the principles and procedures of an ideal legitimacy
which is not conditioned by the very power that it validates
(Rousseau, 1963).
Rousseau's work is significant in another way, and that is in
taking the assumptions of social contract theorising to their logi-
cal conclusion, in the institution of the democratic assembly. The
importance of his book The Social Contract lies in showing that
the social contract cannot be conceived of either as a once-for-
all agreement on rules of power from a position of equality prior
to the establishment of all power relations; nor as a hypothetical
device for imagining what rules or principles we would agree to
if we were in such a position. Both are ahistorical: the former,
because it binds all future generations whatever the change in
circumstances that may subsequently occur; the latter, because
in seeking to demonstrate what principles people would choose
if they were in such an 'original' position, it makes assumptions
about human nature that are contestable, and about the human
condition that are historically variable (the problem with all such
theorising from Hobbes through to Rawls). We cannot antici-
pate, say, what principles of distributional justice people would
agree to in an original position, outside all power relations and
in ignorance of their future position within them. What we must
do instead is to construct a set of institutions in which the idea
of the original or hypothetical contract between equals is as
nearly as possible approached in practice, by making all rules of
power open to discussion, criticism and revision by the decision
of equal citizens. This means that the democratic legislative
assembly must have preeminent place in any normative theory
of legitimacy; and that theorising about democracy must be logi-
cally prior to all other questions in political philosophy (see
Beetham, 1987, ch.3).
The problems that confront the ideal of the citizen assembly
when it is subjected to the practical requirements of represen-
tation and majority voting are not my subject here. Suffice it to
say that Rousseau's response to both requirements was inad-
equate: to the former, because he refused to entertain the idea
of representation at all; to the latter, because his concept of the
'general will' notoriously constitutes an evasion of the problem
248 The Legitimation of Power

of majority - minority relations, rather than a solution to them.

In practice the interests of minorities can only be protected by
making the equality of citizenship status that is central to democ-
racy subject to constitutional safeguard; and by devising pro-
cedural arrangements that encourage compromise between diver-
gent interests rather than the overriding of some by others. The
fuller discussion of such issues belongs elsewhere.
At a second level, then, the activities of explanatory and nor-
mative theorising share not only a common concept of legitimacy,
but also a common epistemological standpoint. There is a final
level, however, where they become effectively one and the same
enterprise. This is the point where, in order to explain the break-
down of non-democratic regimes, the social scientist has recourse
to an evolutionary account that depends for its explanatory force
upon the validity of the normative hierarchy established by the
philosopher; and, in explaining the circumstances in which liberal
democracies typically break down, the social scientist makes a
necessary contribution to the philosopher's exploration of the
conditions necessary to the realisation of this political formation.
Let me take each of these in turn.
In analysing the different way in which non-democratic regimes
break down from the way in which liberal democracies do, so I
have argued, the normative philosopher's reasons for the superi-
ority of democratic principles become part of the social scientist's
explanation for the vulnerability of non-democratic forms to com-
petition from, and supersession by, more democratic ones. The
link between the two is forged by social agents themselves, whose
struggle for a more emancipatory principle of legitimacy is one
of the key causes of social change. It is here that the analogy
Habermas draws with developmental psychology is important, in
that it shows how a normative hierarchy can be united with an
explanatory account of social change by postulating processes of
breakdown as part of an evolutionary process towards a more
advanced stage of development (Habermas, 1979, ch.3). It also
suggests that such a process corresponds to a developmental
potential inherent in the human species itself, which is realised
in the course of social and political struggle.
Such an account is similar to that offered by Hegel in his
Reason in History, which provides a classical example of the
integration of normative and explanatory theorising within an
Political Science and Political Philosophy 249

evolutionary perspective (Hegel, 1953). To carry conviction,

however, such a perspective needs to be freed from the idealist
and teleological tendencies to which it is prone. Overcoming the
first requires a post-Marxian understanding of the way in which
developments at the level of ideas are linked to developments
in productive forms, albeit in more complex ways than economic
determinism allows. Overcoming the second, teleological, tend-
ency requires a post-Darwinian recognition that the processes
whereby new social forms first develop are very different from
those that ensure their survival and expansion in competition
with others; and that the two require correspondingly different
types of explanation. In particular, there is room for all the
contingency of a historical account in explaining the original
emergence of higher social forms. Yet what validates them as
'higher' is that they prove in practice non-regressive, and able
to survive and replicate themselves in competition with other
social and political forms; not that they are foreordained accord-
ing to some Aristotelian teleology, whereby all the stages of
future development are prefigured in an original starting point,
as the oak is in the acorn from which it grows.
Many social and political scientists are sceptical of evolutionary
models, for a variety of reasons, some of which I would endorse
(see p.241). In the final analysis, however, I have found that the
breakdown of non-democratic regimes in comparison with liberal
democracies is only intelligible by assuming the validity of an
evolutionary perspective. Provided it does not constitute the star-
ting point of enquiry, where it will discourage us from analysing
different political systems in their own terms, but only as its
culmination, at this point such a perspective offers the means of
uniting not only explanatory and normative theorising, but theory
and practice also. That this is so will be evident from the final
point of connection between the two modes of theorising.
Whereas the breakdown of non-democratic regimes exposes
the limitations of a lower stage of development, the breakdown
of liberal democracies reveals the difficulties of realising a higher
one. (I refrain from saying 'highest' or 'final'.) Here the political
scientist's explanation for the breakdown of liberal democracy is
but the other side of the coin to the philosopher's account of
the conditions necessary to its realisation, since they are one and
the same. Among the key difficulties I have identified is the
250 The Legitimation of Power

paradox that the principle of popular sovereignty embraces both

a universalist criterion of citizenship, and a particularistic limi-
tation of 'the people' to members of a given nation, so that any
political conception of the general interest becomes confined to
the boundaries of the nation state. It is precisely the commitment
of liberal-democratic governments (and their electorates) to
realising the general interest within national boundaries that con-
stitutes one of the chief obstacles to generalising the conditions
for liberal democracy on a global scale. To be sure, the interde-
pendence of all our interests in securing the conditions for human
survival is becoming every day more evident. Yet it may take
the institutional embodiment of this interdependence to develop
at the global level before liberal democracy can be realised as a
universal, and not merely a localised, political form. If that is
the conclusion towards which theoretical analysis draws us, then
it cannot be a matter of theory alone, but also of practical
affirmation.
Bibliography
Abercrombie, N. and Turner, B.S. (1978) The dominant ideology
thesis', British Journal of Sociology, 29, pp.149-70.
Abrahamian, E. (1982) Iran Between Two Revolutions (Princeton Uni-
versity Press).
Ahmed, E. (1982) 'Comments on Skol:pol'. Theory and Society, 11,
pp.293-300.
Ahmed, I. (1987) The Concept of an Islamic State (London: Pinter).
Alavi, H. (1988) 'Pakistan and Islam: ethnicity and ideology', in Halli-
day, F. and Alavi, H. (1988), pp.64-lll.
Anderson, B. (1983) Imagined Communities (London: Verso).
Aristotle (1954) The Nichomachean Ethics (Oxford University Press).
Aristotle (1962) The Politics (Harmondsworth: Penguin).
Bahro, R. (1978) The Alternative in Eastern Europe (London: New Left
Books).
Bakhash, S. (1986) The Reign of the Ayatollahs (London: Allen and
Unwin).
Barker, R. (1990) Political Legitimacy and the State (Oxford University
Press).
Barrett, M. (1980) Women's Oppression Today (London: Verso).
Bauman, Z. (1990) 'Communism: a post-mortem', Praxis International,
10, pp.185-92.
Beetham, D. (1987) Bureaucracy (Open University Press).
Beetham, D. (1991) 'Weber and the legitimacy of the modern state',
Analyse & Kritik, 13, forthcoming.
Bensman, J. and Givant, M. (1975) 'Charisma and modernity: the use
and abuse of a concept', Social Research, 42, pp.560-614.
Bentham, J. (1960) A Fragment on Government and an Introduction to
the Principles of Morals and Legislation (Oxford: Blackwell).
Bialer, S. (1980) Stalin's Successors (Cambridge University Press).
Bienen, H. and Morell, D. (eds.) (1976) Political Participation under
Military Regimes (London: Sage).
Blau, P.M. (1963) 'Critical remarks on Weber's theory of authority',
American Political Science Review, 57, pp.305-16.
Bloch, M. (1961) Feudal Society (London: Routledge).
Borosage, R.L. and Marks, J. (eds.) (1976) The CIA File (New York:
Grossman).

251
252 Bibliography

Breuilly, J. (1982) Nationalism and the State (Manchester University

Press).
Brunner, G. (1982) 'Legitimacy doctrines and legitimation procedures
in East European systems', in Rigby and Feher (1982), pp.27-44.
Brus, W. (1972) The Market in a Socialist Economy (London: Rout-
ledge).
Bush, R. et al. (eds.) (1987) The World Order (Cambridge: Polity).
Cardoso, F.H. and Faletto, E. (1978) Dependency and Development in
Latin America (University of California Press).
Chodorow, N. (1978) The Reproduction of Mothering (University of
California Press).
Clapham, C. (1985) Third World Politics (London: Croom Helm).
Clapham, C. and Philip, G. (eds.) (1985) The Political Dilemmas of
Military Regimes (London: Croom Helm).
Cohan, A.S. (1975) Theories of Revolution: an Introduction (London:
Nelson).
Cohen, G. (1981) 'Freedom, justice and capitalism', New Left Review,
126, pp.3-16.
Cohn, B.S. (1983) 'Representing authority in Victorian India', in
Hobsbawm and Ranger (1983), pp.165-209.
Collier, D. (ed.) (1979) The New Authoritarianism in Latin America
(Princeton University Press).
Connolly, W. (ed.) (1984) Legitimacy and the State (Oxford: Blackwell).
Coraggio, J.L. (1986) Nicaragua: Revolution and Democracy (London:
Allen and Unwin).
Dahrendorf, R. (1990) Reflections on the Revolution in Europe (London:
Chatto and Windus).
Denitch, B.D. (1976) The Legitimation of a Revolution: the Yugoslav
Case (Yale University Press).
Denitch, B.D. (ed.) (1979) Legitimation of Regimes (London: Sage).
De Sainte Croix, G.E.M. (1981) The Class Struggles in the Ancient
Greek World (London: Duckworth).
Djilas, M. (1958) The New Class (London: Thames and Hudson).
Dunn, J. (1972) Modern Revolutions (Cambridge University Press).
Enayat, H. (1982) Modern Islamic Political Thought (London: Mac-
millan).
Feher, F. (1982) 'Paternalism as a mode of legitimation in Soviet-type
societies', in Rigby and Feher (1982), pp.64-81.
Feher, F., Heller, A. and Markus, G. (1983) Dictatorship over Needs
(Oxford: Blackwell).
First, R. (1970) The Barrel of a Gun: Political Power in Africa and the
Coup d'Etat (London: Allen Lane).
Flynn, P. (1974) 'Class, clientelism and coercion', Journal of Common-
wealth and Comparative Politics, 12, pp.129-56.
Fox, A. (1985) History and Heritage (London: Allen and Unwin).
Freeman, M. (1972) 'Review article: theories of revolution', British
Journal of Political Science, 2, pp.339-59.
Bibliography 253
Friedl, E. (1975) Women and Men (New York: Holt, Reinhart and
Winston).
Friedman, M. (1962) Capitalism and Freedom (University of Chicago
Press).
Friedrich, C.J. (1961) 'Political leadership and the problem of charis-
matic power', Journal of Politics, 23, pp.3-24.
Friedrich, C.J. (1963) Man and Government (New York: McGraw-Hill).
Fukuyama, F. (1989) 'The End of History?', National Interest, 16,
pp.3-18.
Gellner, E. (1983) Nations and Nationalism (Oxford: Blackwell).
Genovese, E. (1970) The World the Slaveholders Made (London: Allen
Lane).
Genovese, E. (1975) Roll, Jordan, Roll (London: Andre Deutsch).
George, S. (1988) A Fate Worse than Debt (Harmondsworth: Penguin).
Giddens, A. (1968) "'Power" in the recent writings of Talcott Parsons',
Sociology, 2, pp.257-72.
Gill, G. (1982) 'Personal dominance and the collective principle: indi-
vidual legitimacy in Marxist - Leninist systems', in Rigby and Feher
(1982), pp.94-110.
Gilsenan, M. (1988) 'State and popular Islam in Egypt', in Halliday and
Alavi (1988), pp.167-90.
Grafstein, R. (1981) 'The failure of Weber's concept of legitimacy',
Journal of Politics, 43, pp.456-72.
Habermas, J. (1973) 'What does a crisis mean today? Legitimation
problems in late capitalism', Social Research, 40, pp.643-67.
Habermas, J. (1976a) 'Legitimationsprobleme im modernen Staat',
Merkur, 332, pp.37-56 (revised English version in Habermas, 1979).
Habermas, J. (1976b) Legitimation Crisis (London: Heinemann).
Habermas, J. (1979) Communication and the Evolution of Society
(London: Heinemann).
Habermas, J. (1984) The Theory of Communicative Action, vol. 1
(London: Heinemann).
Habib, J.S. (1978) Ibn Sa'ud's Warriors of Islam (Leiden: Brill).
Halliday, F. (1979) Iran: Dictatorship and Development (Harmond-
sworth: Penguin).
Halliday, F. (1988) 'The Iranian revolution: uneven development and
religious populism' in Halliday and Alavi (1988), pp.31-63.
Halliday, F. and Alavi, H. (1988) State and Ideology in the Middle East
and Pakistan (London: Macmillan).
Haraszti, M. (1977) A Worker in a Worker's State (Harmondsworth:
Penguin).
Harris, M. (1977) Cannibals and Kings (New York: Random House).
Harris, R.L. and Vilas, C.M. (1985) Nicaragua: a Revolution under
Siege (London: Zed Books).
Hart, H.L.A. (1961) The Concept of Law (Oxford University Press).
Hasek, J. (1973) The Good Soldier Svejk (London: Heinemann).
Havel, V. (1987) Living in Truth (London: Faber).
254 Bibliography

Hayek, F. (1976) Law, Legislation and Liberty, vol. 2 (London: Rout-

ledge).
Hegel, G.W.F. (1953) Reason in History (New York: Hobbs-Merrill).
Held, D. (1982) 'Crisis tendencies, legitimation and the state', in Held,
D. and Thompson, J.B. (eds.) Habermas: Critical Debates (London:
Macmillan), pp.182-95.
Held, D. (1989) Political Theory and the Modern State (Cambridge:
Polity).
Heller, A. (1982) 'Phases of legitimation in Soviet-type societies', in
Rigby and Feher (1982), pp.45-63.
Hennis, W. (1976) 'Legitimitat', Merkur, 332, pp.17-36.
Hinton, W. (1966) Fanshen (New York: Monthly Review Press).
Hiro, D. (1987) Iran under the Ayatollahs (London: Routledge).
Hobbes, T. (1960) Leviathan (Oxford: Blackwell).
Hobsbawm, E.J. and Ranger, T. (eds.) (1983) The Invention of Tra-
dition (Cambridge University Press).
Hopkins, K. (1978) Conquerors and Slaves (Cambridge University
Press).
Hopwood, D. (1982) 'The ideological basis: Ibn Abd al-Wahab's Muslim
revivalism', in Niblock (1982), pp.23-35.
Huntington, S.P. (1968) Political Order in Changing Societies (Yale
University Press).
Joseph, R.A. (1987) Democracy and Prebendal Politics in Nigeria (Cam-
bridge University Press).
Keane, J. (1984) 'Introduction', Offe (1984), pp.ll-34.
Keddie, N.R. (1981) Roots of Revolution (Yale University Press).
Keddie, N.R. (1982) 'Comments on Skocpol', Theory and Society, 11,
pp.285-92.
Keddie, N.R. (ed.) (1983) Religion and Politics in Iran (Yale University
Press).
Keddie, N.R. (1988) 'Ideology, society and state in post-colonial Muslim
societies', in Halliday and Alavi (1988), pp.9-30.
Kimmel, M.S. (1990) Revolution: a Sociological Interpretation (Cam-
bridge: Polity).
King, D.S. (1987) The New Right: Politics, Markets and Citizenship
(London: Macmillan).
Kolinsky, M. (1981) 'The nation state in Western Europe' in Tivey
(1981), pp.82-103.
Kramnick, I. (1972) 'Reflections on revolution: definition and expla-
nation in recent scholarship', History and Theory, 11, pp.26-63.
Krejci, J. (1983) Great Revolutions Compared (Brighton: Wheatsheaf).
Lane, C. (1984) 'Legitimacy and power in the Soviet Union through
Socialist ritual', British Journal of Political Science, 14, pp.207-17.
Lessnoff, M. (1986) Social Contract (London: Macmillan).
Lewis, P.G. (1982) 'Obstacles to the establishment of political legitimacy
in communist Poland', British Journal of Political Science, 12,
pp.125-47.
Bibliography 255
Lewis, P.G. (ed.) (1984) Eastern Europe: Political Crisis and Legitim-
ation (London: Croom Helm).
Linz, J.J. and Stepan, A. (eds.) (1978) The Breakdown of Democratic
Regimes (John Hopkins University Press).
Lipjhart, A. (1977) Democracy in Plural Societies (Yale University
Press).
Lipset, S.M. (1958) 'Some social requisites of democracy', American
Political Science Review, 53, pp.69-105.
Locke, J. (1967) Two Treatises of Government (Cambridge University
Press).
Lomax, B. (1984) 'Hungary- the quest for legitimacy' in Lewis (1984),
pp.68-110.
Lowenthal, R. (1979) 'Political legitimacy and cultural change in West
and East', Social Research, 49, pp.401-35.
Ludz, P.C. (1979) 'Legitimacy in a divided nation', in Denitch (1979),
pp.161-76.
Luhmann, N. (1969) Legitimation durch Verfahren (Darmstadt: Luchter-
hand).
Lukes, S. (1974) Power: a Radical View (London: Macmillan).
McFarlane, B. (1984) 'Political crisis and East European economic
reforms', in Lewis (1984), pp.176-99.
Macintyre, A. (1970) 'The idea of a social science' in Wilson (1970),
pp.112-30.
McQuail, D. (1984) Communication (London: Longmans).
Mahdavy, H. (1970) 'The patterns and problems of economic develop-
ment in rentier states: the case of Iran', in Cook, M.A. (ed.)
Studies in the Economic History of the Middle East (Oxford Univer-
sity Press), pp.428-67.
Mann, M. (1986) The Sources of Social Power (Cambridge University
Press).
Marx, K. (1966) Capital, 3 vols (Moscow: Progress Publishers).
Marx, K. and Engels, F. (1968) Selected Works in One Volume (London:
Lawrence and Wishart).
Marx, K. and Engels, F. (1970) The German Ideology (London: Lawr-
ence and Wishart).
Merelman, R. (1966) 'Learning and Legitimacy', American Political
Science Review, 60, pp.548-61.
Milgram, S. (1974) Obedience to Authority (New York: Harper and
Row).
Mill, J. (1937) An Essay on Government (Oxford: Blackwell).
Mill, J.S. (1909) Principles of Political Economy (London: Longmans).
Mill, J.S. (1964) Utilitarianism; Liberty; Representative Government
(London: Dent).
Mill, J.S. (1970) On the Subjection of Women (London: Dent).
Millett, K. (1971) Sexual Politics (London: Hart-Davis).
Moore, B. (1978) Injustice (London: Macmillan).
Morriss, P. (1987) Power: a Philosophical Analysis (Manchester Univer-
sity Press).
256 Bibliography

Mortimer, E. (1982) Faith and Power: the Politics of Islam (London:

Faber).
Mosca, G. (1939) The Ruling Class (New York: McGraw-Hill).
Mouzelis, N. (1986) Politics in the Semiperiphery (London: Macmillan).
Nagel, J.H. (1975) The Descriptive Analysis of Power (Yale University
Press).
Nelson, J.M. (1979) Access to Power: Politics and the Urban Poor in
Developing Nations (Princeton University Press).
Niblock, T. (ed.) (1982) State, Society and Economy in Saudi Arabia
(London: Croom Helm).
Niblock, T. (1987) Class and Power in Sudan (London: Macmillan).
Nordlinger, E.A. (1972) Conflict Resolution in Divided Societies (Harv-
ard University Center for International Affairs).
Nordlinger, E.A. (1977) Soldiers in Politics: Military Coups and Govern-
ments (New Jersey: Prentice Hall).
Nove, A. (1983) The Economics of Feasible Socialism (London: Allen
and Unwin).
Oakeshott, M. (1962) Rationalism in Politics and Other Essays (London:
Methuen).
Oakley, A. (1972) Sex, Gender and Society (London: Temple Smith).
O'Brien, P. and Cammack, P. (eds.) (1985) Generals in Retreat (Man-
chester University Press).
O'Donnell, G. (1973) Modernisation and Bureaucratic Authoritarianism
(University of California Institute of International Studies).
Offe, C. (1984) Contradictions of the Welfare State (London: Hutch-
inson).
Offe, C. (1985) Disorganized Capitalism (Cambridge: Polity).
Okin, S.M. (1980) Women in Western Political Thought (London:
Virago).
Onimode, B. (ed.) (1989) The IMF, the World Bank and African Debt,
2 vols, (London: Zed Books).
Pakulski, J. (1986) 'Legitimacy and mass compliance: reflections on Max
Weber and Soviet-type societies', British Journal of Political Sci-
ence, 16, pp.35-56.
Parent, W.A. (1974) 'Freedom as the non-restriction of options', Mind,
83, pp.432-4.
Pareto, V. (1966) Sociological Writings (London: Pall Mall).
Parkin, F. (1974) 'Strategies of social closure in class formation' in The
Social Analysis of Class Structure (London: Tavistock), pp.1-18.
Parkin, F. (1979) Marxism and Class Theory (London: Tavistock).
Parsons, T. (1960) Structure and Process in Modern Societies (Glencoe:
Free Press).
Pateman, C. (1980) 'Women and Consent', Political Theory, 8,
pp.149-68.
Pateman, C. (1985) The Problem of Political Obligation (Cambridge:
Polity).
Payne, G. (1987) Mobility and Change in Modern Society (London:
Macmillan).
Bibliography 257

Perlmutter, A. and Leogrande, W.M. (1982) 'The party in uniform:

towards a theory of civil - military relations in communist political
systems', American Political Science Review, 76, pp.778-89.
Philip, G. (1985) The Military in South American Politics (London:
Croom Helm).
Pitkin, H.F. (1972) Wittgenstein and Justice (University of California
Press).
Plamenatz, J.P. (1968) Consent, Freedom and Political Obligation
(Oxford University Press).
Plant, R. (1982) 'Jiirgen Habermas and the idea of legitimation crisis,'
European Journal of Political Research, 10, pp.341-52.
Poggi, G. (1978) The Development of the Modern State (London: Hutch-
inson).
Powis, J. (1984) Aristocracy (Oxford: Blackwell).
Randall, V. (1982) Women and Politics (London: Macmillan).
Randall, V. (ed.) (1988) Political Parties in the Third World (London:
Sage).
Rawls, J. (1971) A Theory of Justice (Harvard University Press).
Reeve, A. (1986) Property (London: Macmillan).
Richter, M. (1982) 'Toward a concept of political illegitimacy: Bonapar-
tist dictatorship and democratic legitimacy', Political Theory, 10,
pp.185-214.
Rigby, T.H. (1980) 'A conceptual approach to authority, power and
policy in the Soviet Union', in Rigby, Brown and Reddaway (1980),
pp.9-31.
Rigby, T.H. (1982) 'Political legitimacy, Weber and communist mono-
organisational systems' in Rigby and Feher (1982), pp.1-26.
Rigby, T.H., Brown, A. and Reddaway, P. (1980) Authority, Power
and Policy in the USSR (London: Macmillan).
Rigby, T.H. and Feher, F. (eds.) (1982) Political Legitimation in Com-
munist States (London: Macmillan).
Rosaldo, M. and Lamphere, L. (eds.) (1974) Woman, Culture and
Society (Stanford University Press).
Rose, G. (1983) 'Velayat-e Faqih and the recovery of Islamic identity
in the thought of Ayatollah Khomeini' in Keddie (1983), ch.9.
Rothschild, J. (1977) 'Observations on political legitimacy in contempor-
ary Europe', Political Science Quarterly, 92, pp.487-501.
Rousseau, J-J. (1963) The Social Contract and Discourses (London:
Dent).
Runciman, W.G. (1965) Social Science and Political Theory (Cambridge
University Press).
Ryan, A. (1987) Property (Open University Press).
Said, E.W. (1981) Covering Islam (London: Routledge).
Saint-Simon, C.H. (1975) Selected Writings (London: Croom Helm).
Sanday, P.R. (1981) Female Power and Male Dominance (Cambridge
University Press).
Schaar, J.H. (1969) 'Legitimacy in the modern state' in Green, P. and
258 Bibliography

Levinson S. (eds.) Power and Community (New York: Random

House), pp.276-327.
Schmidt, S.W. et al. {eds.) (1977) Friends, Followers and Factions (Cali-
fornia University Press).
Scott, J. (1982) The Upper Classes (London: Macmillan).
Scott, J.C. (1969) 'The analysis of corruption in developing nations',
Comparative Studies in Society and History, 11, pp.315-41.
Scott, J.C. (1976) The Moral Economy of the Peasant (Yale University
Press).
Scott, J.C. and Kerkvliet, B.J. (1977) 'How traditional rural patrons
lose legitimacy', in Schmidt (1977) pp.439-58.
Semmel, B. (1960) Imperialism and Social Reform (London: Allen and
Unwin).
Skocpol, T. (1979) States and Social Revolutions (Cambridge University
Press).
Skocpol, T. (1982) 'Rentier state and Shi'a Islam in the Iranian revo-
lution', Theory and Society, 11, pp.239-62.
Smith, A. (1976) An Inquiry into the Nature and Causes of the Wealth
of Nations (University of Chicago Press).
Spender, D. (1980) Man Made Language (London: Routledge).
Steiner, H. (1974) 'Individual liberty', Proceedings of the Aristotelian
Society, 75, pp.33-50.
Stewart, J.H. (1951) A Documentary Survey of the French Revolution
(New York: Macmillan).
Stone, L. (1965) 'Theories of revolution', World Politics, 18, pp.159-76.
Szeftel, M. (1983) 'Corruption and the spoils system in Zambia' in
Clarke, M. (ed.) (1983) Corruption (London: Pinter), pp.163-89.
Szeftel, M. {1987) 'The crisis in the Third World', in Bush (1987),
pp.87-140.
Szentes, T. (1983) The Political Economy of Underdevelopment (Budap-
est: Akademiai Kiado).
Taylor, M. (1988) Rationality and Revolution (Cambridge University
Press).
Therborn, G. (1980) The Ideology of Power and the Power of Ideology
(London: Verso).
Thomas, C.Y. (1984) The Rise of the Authoritmian State in Peripheral
Societies (New York: Monthly Review Press).
Thompson, E.P. (1975) Whigs and Hunters (London: Allen Lane).
Thorpe, F.N. (ed.) (1909) The Federal and State Constitutions of the
USA (Washington: Government Printing Office).
Tilly, C. (ed.) (1975) The Formation of National States in Western
Europe (Princeton University Press).
Tivey, L. (ed.) (1981) The Nation State (Oxford: Martin Robertson).
Tordoff, W. (1984) Government and Politics in Africa (London: Mac-
millan).
Weber, M. (1956) Wirtschaft und Gesellschaft, 4th ed. (Tubingen; J.C.B.
Mohr).
Bibliography 259

Weber, M. (1958) Gesammelte Politische Schriften, 2nd ed. (Tubingen:

J.C.B. Mohr).
Weber, M. (1968) Economy and Society (University of California Press).
Weingrod, A. (1968) 'Patrons, patronage and political parties', Com-
parative Studies in Society and History, 10, pp.377-400.
Whitelegg, E. et al. (eds.) (1982) The Changing Experience of Women
(Oxford: Blackwell).
Wilber, C.K. (ed.) (1984) The Political Economy of Development and
Under-development, 3rd ed. (New York: Random House).
Willis, P. (1977) Learning to Labour (Aldershot: Gower).
Wilson, B.R. (ed.) (1970) Rationality (Oxford: Blackwell).
Winch, P. (1958) The Idea of a Social Science (London: Routledge).
Winch, P. (1964) 'Understanding a primitive society', American Philo-
sophical Quarterly, 1, pp.307-24; reprinted in Wilson (1970),
pp.78-lll.
Wolin, S.S. (1981) 'Max Weber: legitimation, method and the politics
of theory', Political Theory, 9, pp.401-24.
Wollstonecraft, M. (1970) A Vindication of the Rights of Women
(London: Dent).
Wolpe, H. (1968) 'A critical analysis of some aspects of charisma',
Sociological Review, 16, pp.305-18.
Wrong, D. (1979) Power: its Forms, Bases and Uses (Oxford:
Blackwell).
Young, M. (1961) The Rise of the Meritocracy (Harmondsworth: Pen-
guin).
Zagorin, P. (1973) 'Theories of revolution in contemporary historiogra-
phy', Political Science Quarterly, 88, pp.23-52.
Zubaida, S. (1982) 'The ideological conditions for Khomeini's doctrine
of government', Economy and Society, 11, pp.138-72.
Index

Abdul-Aziz 195 214-15, 219-20, 230, 233-4,

acclamation 12, 93-4 236, 238-9, 240, 245
achievement principle 21, 80-1,
167 belief(s) 6, 8, 10-12, 16-17,
agreement, normative 17, 86, 21-3, 69, 70, 75-6, 89-90,
169 104-6, 109-10, 117, 125,
128-9, 149, 156-7, 164, 185,
agreements 12, 13, 18, 93, 102
207' 225' 233' 238-9
Amin, I. 235 belief in legitimacy 6, 8-13,
anarchism 6, 138 23-5, 34, 38, 91, 108, 234
Angola 120 Bentham, J. 140
Arab-Israeli conflict 192, 196, birth 4, 17, 77-80, 113, 128
201 Brezhnev, L. I. 187
Argentina 119, 235 Bulgaria 217
aristocracy 52, 77-8, 80, 229 bureaucracy 82-3, 125, 185
Aristotle 30, 73, 249 Burma 223
armed forces see military
ascription, principle of 77-80, 97 capitalism 30, 32, 52, 82, 84-5,
103, 107, 140, 162, 164-70,
Australia 163
172-3, 176-8, 184
authority 16, 19, 26, 36, 49-50,
capitalist democracy see liberal
52, 82, 102, 121, 122, 125, democracy
146, 167, 169, 210-12, 231 caste 98
moral 57, 94, 132, 180, 186, Castro, F. 180
187, 210, 214, 216-19, 221, Ceausescu, N. 218
236 ceremonial 18-19, 67, 93
source of 17, 21, 70-6, 89-90, charismatic authority 10, 24-5,
109, 126, 128-35, 149, 159, 156-7, 180
161, 171, 202, 204, 207, 211, children 4, 45, 79, 87, 96

260
Index 261

China 83, 179, 218 contract 12, 13, 18, 33, 91-2, 93,
choice 18, 91-2, 96-7, 151-2, 94-7
155, 157, 236, 239 convention(s) 12, 14, 16, 18, 20,
civil disobedience see 21, 35, 54, 65-7, 92, 96, 100
disobedience coordination, social 26, 46
civil society 164, 172 corruption 142-4, 230-1, 234
civil war 120, 212, 215, 224, 227 counter-revolution 222, 224, 227
class 21, 22, 37, 51-3, 55, 59, coup d'etat 16, 20, 101, 118,
83-8, 97, 104, 144, 152, 119-20, 171, 205-7, 211-13,
154-5, 165, 173-4, 178, 231, 228-32
237 crisis of authority 35, 110, 119,
clientelism 154, 173-4 153, 165, 186, 217, 219
coercion 6, 28, 30, 40, 44, 53, 58, crisis tendencies 120, 168
86, 88, 91, 121, 138-9, 179 Cuba 179
collapse of regime 33, 119, 187, Cyprus 235
199-200, 218, 224 Czechoslovakia 119, 180, 186-7,
collective bargaining 30, 93 217, 223
colonialism 59
command economy 32, 52, 85-6, delegitimation 19, 20, 110, 153,
141, 183-6 183, 186, 197, 200, 201, 205,
206, 209-10, 211, 212, 216-20
common interest see general
interest democracy, democratisation 11,
73, 75, 113-4, 119, 129, 165,
communist model 179-90, 203,
188-9, 199, 211, 212, 214,
204, 210, 227, 238-9
222-3, 228, 231-2, 241, 244,
communist regimes 9, 28, 32, 95, 247, 248
119, 124, 129, 161
see also liberal democracy
comparative analysis 21-2, 246 dependency 45, 48, 53-4, 87-9,
conquest 20, 30, 55, 93, 98, 101, 96-7, 141
206 developing countries 85, 133-4,
consensus theories 104, 220-1 140-2, 171-8, 241
consent 3, 5, 12, 13, 16, 18-19, developmental psychology 15,
21, 22, 24, 31, 37, 60, 61-2, 248
63, 90-7, 98, 100, 104, 107-8, developmental stages 98-9,
125, 138, 148, 150-60, 171-3, 240-1, 248-30
194, 205, 209, 234, 235, 237, dictatorship 20, 1i7, 119, 162,
238, 240, 245 189, 199, 211, 226, 227, 236,
consent 239-40
electoral mode of 131, 151-6, differentiation, principles of 59,
188, 200-1, 223, 226, 227, 239 63, 76-82, 98, 109, 245
ideal conditions for 14, 38, dismissal from office 136, 146,
111, 113, 246 169
liberal theory of 18, 91-2, disobedience 7, 8, 19, 26, 33,
96-7 209-11, 216
mobilisation mode of 131-2, disorder 139-40
151, 155-8, 181-3, 185, 188, dissent 183, 186, 209, 215
200-1, 221, 225-6, 239 divine authority 21, 36, 71, 89,
consociational democracy 148 129
262 Index

division of labour 31, 49, 50-4, family, the 53-4, 66, 80-1, 87,
57, 66, 76, 82, 102, 106-7, 113, 128, 167
154, 243 fascism 238-9
dominance 4, 47-56, 58 feudalism 82, 92
dominant groups 17, 19, 34-6, fiscal crisis 167
59, 67, 82, 104-6 France 128, 129-30, 131, 153,
duress 20, 44, 62 224, 226
freedom 19, 43, 58, 60, 63, 91,
97
East Germany 190, 217 freedoms, political 152, 155,
Eastern Europe 28-9, 119, 176, 164, 168, 171, 174, 176, 177,
179-80, 186-8, 210, 223 189, 203, 210, 212, 230, 236
education 28, 61, 167, 225, 227,
Fukuyama, F. 189-90
229 functionalism 34, 61
Edwards, B. 86
Egypt 192, 196 gender 21, 31, 37, 39, 53-4, 55,
election(s) 12, 18, 92, 93, 96,
59, 76, 78-80, 90, 97, 101,
119, 127, 129, 146-8, 151-7,
105, 152, 154
163-4, 170, 172-5, 182, 187,
general interest 17, 27, 46-7,
188, 207, 212, 215, 222, 224,
59-60, 61, 63, 77, 82-90, 98,
227, 230, 235-6, 239
106-7, 113, 127, 135-6,
electoral system, British 11 147
142-50, 164, 175, 177, 212,
elites 33, 105, 180, 182, 186 198
218, 220 , , 214, 216, 231, 237, 239, 245
Genovese, E. 87-8
ends of government 136-42, 149,
Gorbachev, M. 28, 130, 161,
159, 207, 233, 234
179, 181, 187, 188-9
epistemology 110-11, 248 government performance 145-8
equality 66, 80, 113, 154, 203, 184-5, 207-9, 211, 216, ,
239, 246, 247
231-3, 235
of opportunity 112-13, 128, Grafstein, R. 9
189, 238, 240 Greece 119, 120, 235
erosion of legitimacy 20, 23, 28,
32-3, 109-10, 118, 128, 157,
Habermas, J. 14-15, 165-70, 248
163, 229, 239
Hart, H. L. A. 65
Ethiopia 120
Hegel, G. W. F. 8, 248-9
ethnic groups 55, 133, 144, 148
175, 237 , heredity 4, 17, 21, 75, 78, 126,
127, 129, 132, 193-4 229 238
explanation Hinton, W. 83 ' '
of behaviour 6, 23, 26, 100,
historical
246
of beliefs 9-10 22 of legitimacy 15 , 18 , 21 ,
forms
of legitimacy 101-10, 246
origins 102-3
exploitation 58, 83
Hitler, A. 156-7
expropriation 16
Hobbes, T. 8, 80, 138-9, 247
external intervention 122, 202-4,
Holland 129
228, 236
human nature 102, 247
human rights 128, 236
Falkland Islands 235 Hungary 119, 180, 186
Index 263

Huntington, S. 229-30 Latin America 174, 229

Hussein, S. 235 law 4, 7, 16, 40, 65-7, 70, 121,
143, 200, 237, 245
idealism 37, 249 law, rule of 68, 122-6, 158
ideology 34, 62, 78, 95, 104-7, lawyers 4, 13, 14
225, 227, 229 leadership 156-7, 201, 236
illegality 117, 124, 142-3, 205, legality, legal validity 4, 5, 12,
218, 229 13, 15-16, 56-7, 64-9, 121-6,
illegitimacy 4, 9, 16, 20, 76, 179, 158-9, 164, 171, 205, 206-7,
205, 206-7, 232-6 211, 218, 224, 227, 229
incentives 27, 28, 38, 44-5, 87 legitimacy deficit 18, 20, 75, 120,
India 175 149, 205, 207-9, 211, 214
individualism 18-9, 91-2, 151-2, legitimation 19, 63, 81, 90-7,
169-70 150-60, 181-3, 200-1, 205,
industrialisation 141, 163, 172-3, 206, 221-8, 234-6, 238-9, 246
195-6 crisis 165-71, 178, 179, 181,
inequality 43-5, 47, 52, 54, 58, 186
60, 81, 102, 141, 145, 153, legitimations 9, 19, 30-1, 39,
154, 165, 172, 177, 198-9, 234, 236, 241
228, 237, 239, 246 Leipzig 189
inferiority, ideas of 55, 59, 60, Lenin, V. I. 156
77-8 liberal democracy 92, 97, 149,
influence 8, 34-5, 43-4, 62, 152, 156, 161, 163-78, 182,
105-6 183, 203, 209, 227, 230-3,
injustice 15, 22, 170, 231 239, 240-1, 248-50
intafada 210 liberalism 18, 91-2, 96-7, 124,
intellectuals 215, 217, 220 151-2
international system 122-3 Lipset, S.M. 9, 136-7, 234
Iran 33, 119, 129, 130-1, 193, Locke, J. 8, 12, 153, 228
196-202, 214, 219
Iraq 162, 195, 196, 201
Islam 71, 129, 162, 191-204, 225, majority principle 145, 164, 230,
238-9 247-8
Makarios, Archbishop 235
Japan 163 male power 17, 53-4, 71, 88
J aruzelski, W. 29, 187 Mao Zedong 156, 180
Jordan 129, 238 Marcos, F. 223
judiciary 68, 123-4 market 84-6, 138, 140, 141,
justice 5, 83-4, 145, 167, 170 164-9, 176-8, 184, 187
marriage 4, 91, 97
Keynesianism 169 Marx, K., Marxism 22, 51-2, 55,
Khomeini, R. M. 6, 130-1, 156, 61, 103, 104-5, 107-8, 138,
199-201, 225, 226 174, 218
Kurds 134 Marxism-Leninism 73, 89, 129,
Kuwait 162, 196, 235 130, 132, 180-6, 190, 225
means of production 47-8, 51-3,
Labour regime 30-1, 83-8 55
labour theory of value 52 Merelman, R. 9
264 Index

merit, meritocracy 36, 77, 80-2, obligation 5, 7, 12, 13, 19, 26,
97, 112-13 33, 36, 92, 138, 244
Mexico 175 Offe, C. 165-70
military, the 33, 92-3, 124, 138, oil industry 193, 195-6, 197-8
158, 182, 198, 212, 217, 220, opposition 119, 158, 209, 218
228-36 order, social and political 6, 29,
military regimes 149, 213, 33-4, 138-9, 212, 234
228-36, 239-40 organisation 33, 50, 218
Mill, J. 140
Mill, J. S. 145
mobilisation regime 201, 211, Pahlavi, dynasty 196-9
213, 225-7, 238-9, 240-1 parasitism 58
monarchy 78, 89, 126, 129-30, Pareto, V. 104-5
147, 193-9, 202-3, 229 Parliament 129, 207, 224, 227
Moore, Barrington 22 Parsons, T. 220-1
moral dilemmas 5, 7, 26 particularism 136, 150, 250
moral sense 22, 27, 29, 33, 38, paternalism 45, 87-90, 95, 172,
59, 245 180, 182, 190, 240-1
Morocco 129, 238 patronage 143, 146, 173-4, 198
Mosca, G. 104-5 peasants 66, 82, 83-4, 226
Mossadeq, M. 197, 198 people, the 21, 37, 75, 89, 128,
myths 8, 21, 105 132, 134-5, 182, 189, 211-12,
214, 224, 239, 250
Nasser, G. A. 192 people power 200, 223
nation, national autonomy 75, Philippines 119, 223
127, 132-4, 179-80, 186-7, philosophers 5, 9, 14, 73
188, 191-2, 196, 199, 215, physical force 47-8, 50, 52, 53-4,
233, 235, 250 62, 102, 103, 121, 123, 206
natural law 58, 71-3, 76 Pitkin, H. 9
nature, arguments from 30-1, planning, central 85-6, 163, 166,
58, 62, 78-80, 107, 109, 183-6
113-14 plebiscite 12, 131, 236
Nazism 213 Poland 119, 180, 186
neo-liberalism 138, 169-70 political community, membership
New Zealand 163 of 19, 60, 94, 151, 189, 237
Nicaragua 214, 227-8 political economy 30, 73, 84-5,
Nkrumah, K. 142 103
norms 20-2, 32, 38, 223, 245 political parties 124, 130, 146,
Northern Ireland 120, 134 148, 151-7, 164, 170, 173-5,
Norway 129 177-8, 181, 182-3, 187, 201,
Numeiri, J. F. 193 210, 218, 222, 226, 238
political philosophy 5-9, 13-14,
oath of allegiance 12, 18, 33, 23, 37, 39, 40-1, 111-14, 137,
92-3 243-50
obedience 5, 8, 16, 20, 28, 33, political power 21, 37, 39-40, 54,
38, 47, 91-2 55, 70, 76, 104-5
grounds for 23, 26-8, 101 political science 8, 120, 126, 180,
psychological theories of 26 243-50
Index 265

political systems, types of 9, 11, religion, religious belief 6, 36,

122, 161-204, 209, 237-42 55, 71, 75, 76, 191-204
poll-tax 12 rentier-state 195, 198
popular sovereignty 89-90, 94, representation, political 11, 36,
113, 117, 127, 128-35, 75, 89, 129, 181, 183, 192,
149-50, 151, 163, 168, 182, 197, 210, 215, 216, 233, 247
185, 189, 194, 211, 215, repression 110, 157, 174, 183,
233-4, 238-41, 250 211, 216, 217-18, 220, 239
Portugal 119 resources 27, 33, 38, 43-4, 47-8,
power 50-2, 56, 59, 61, 77, 80, 82,
acquisition of 12, 16, 39, 121, 106, 195
212-13, 218, 222, 229-31 revivalism 193-5, 201
definition of 38, 42-6 revolt, rebellion 6, 32, 213, 215
effectiveness of 28, 29, 32-4, revolution 6, 20, 94, 110, 118,
119, 136-7 132, 179, 203, 205-7, 211,
limits to 35-6 212, 213-28, 233
means of 47-51, 55, 57, 80-1, Bolshevik 224, 226
106 Chinese 224, 226
need for legitimation of 56-63 East European 9, 179, 187,
relations 4, 5, 6, 8, 11, 12, 189-90, 223
17-9, 22, 25, 27, 30, 32, 35, French 127, 224, 226
37-8, 45, 49, 52, 56, 59, 61, Iranian 196, 197-201, 219-20,
88-9, 100, 102, 104, 106-11, 224-5
121, 246 Nicaraguan 227-8
social organisation of 46-56 right(s) 5, 56, 65, 66-7, 137, 153
powerlessness 45, 62, 97, 147 Riyadh 195
Prague 189 roles 78-80
principles, moral or political 4, Romania 217-18
5, 37, 100, 125, 126, 222, 244 Rousseau, J. J. 8, 246-7
private sphere 53-4, 167, 185 rule-conformity 16, 19, 22, 23,
propaganda 9, 19 27, 35, 64-9, 205
property 4, 37, 39, 48, 51-2, 55, rules 3, 5, 11, 12, 16, 17, 21,
65, 66, 76, 84, 101, 102, 137, 35-7, 39-40, 56-7, 69, 74,
153, 164, 166, 173, 174 101, 102, 121, 244
public interest see general breaches of 20, 35, 40, 57, 67,
interest 76, 121, 124, 206, 211, 218,
public opinion 36, 164, 183, 210 229
justifiability of 4, 5, 11, 12, 13,
16, 17, 19, 21, 22, 57, 63,
quality of performance 28-33
69-90, 91, 158-9, 205, 237
rules of exclusion 48-51, 54, 65
race 30, 55, 59, 88, 90, 153 rules of power 16, 20, 51, 54, 58,
rational choice 27 62-3, 67-8, 75, 96-7, 102,
rational-legal authority 10, 24, 108, 113-14, 117, 131, 139,
125 168, 179, 204, 212, 216, 221,
Rawls, J. 15, 247 240, 246, 247
realist theories of power 27, 29,
38, 218, 243 Sadat, A. 193
266 Index

Saint-Simon, C. H. 74 26, 29-34, 50, 56, 59, 66-7,

sanctions 27, 28, 38, 45 82, 91, 105, 109-11, 173
Saudi Arabia 193-6, 197, 201-2, subordination 5, 47-56, 58, 60,
238 96-7, 152, 154, 165
science 21, 73, 76, 89, 114 Sudan 120, 193
Schaar, J. H. 9 suffrage 128, 153, 154, 163, 165,
Scott, James 83-4 173, 203, 215
secularisation 71, 75, 191, 199, superiority, ideas of 17, 55, 59,
203 60, 77-8
security 127, 138-40, 237 symbolism 12, 18
self-interest 5, 12, 18, 23, 27, 95, Syria 196
167, 177
separation of powers 68, 123-4, Tanzania 235
129, 148, 158 technocracy 74, 167
serfdom 98 Tehran 198
shari' a 192, 200, 202 theocracy 75, 114, 127, 129. 132,
Shi'ism 200, 203 162, 193
Sieyes, Abbe, 153 Tiananmen Square 179, 210, 218
single-party regimes 148, 171, Tito, J. 180
175, 236 trade unions 165, 174, 178
Skocpol, T. 218-20 tradition 21, 74-5, 77, 89, 114,
slavery 17, 20, 30-1, 55, 59, 128-9
86-8, 98 traditional authority 10, 24, 93,
social science 5-15, 21-7, 37-41, 94, 147, 150-1, 180, 189,
100-1, 110-14, 243-50 194-7, 203, 213, 215, 227,
socialisation 8, 46, 78-·9, 81, 107, 229, 238, 240-1
125, 182 Turkey 120
socialism 85-6, 184, 199
ulama 192, 195, 200, 202, 225
Somoza, A. 227
underdevelopment 141-2,
South Africa, Republic of 153,
171-3, 176, 178
210 Union of Soviet Socialist
sovereignity 122-3, 125-6, Republics 28, 134, 179, 181,
203-4, 206, 237 186, 188-9, 191, 226, 241
Soviet-type systems 52, 73, 85-6, United Kingdom 129, 147-8
156, 161, 179-90 United States of America 87-8
see also communist model 140, 171, 187, 191, 195, 196:
Spain 119 198, 202, 227-8
Sri Lanka 120 universality 5, 6, 8, 13, 14, 21,
stability 33-4, 120, 208-9, 219, 71, 75, 250
230-1 usurpation 16, 57, 102
Stalin, J. 161, 180, 226 utility 5, 137
state, the 6, 32, 40, 117-18,
121-2, 164, 166-70, 177-8, values 6, 8, 11, 12, 21, 44, 117,
183, 205, 219, 221, 226, 237 220-1, 243
state apparatus 32-3, 118, 185, Vietnam 169
187, 198, 218, 219, 223
subordinate groups 17, 18, 19, wage labour 54, 96
Index 267
Wahabism 194-5 women 51, 53-4, 66, 75, 78-80,
Weber, M., Weberian 6-14, 96-7, 105, 107, 153, 170
23-5, 34, 38, 52, 108, 125, working class 30-1, 36, 52, 75,
156, 180 81, 84-5, 129, 130, 153, 181,
welfare 85, 127, 138, 165, 183-4, 186, 188, 196, 210
169
Yugoslavia 179
Westernisation 198-9
Winch, P. 101, 245 Zia Ul-Haq 193