WILL ROBOTS AUTOMATE

YOUR JOB AWAY? FULL

EMPLOYMENT, BASIC INCOME,
AND ECONOMIC DEMOCRACY

Ewan McGaughey

WP 496
March 2018
 WILL ROBOTS AUTOMATE YOUR JOB AWAY? FULL
EMPLOYMENT, BASIC INCOME, AND ECONOMIC DEMOCRACY

Centre for Business Research, University of Cambridge, Working Paper no. 496

Ewan McGaughey
King’s College, London
and
Centre for Business Research, University of Cambridge
[email protected] or @ewanmcg

March 2018
Abstract

Will the internet, robotics and artificial intelligence mean a ‘jobless future’? A
recent narrative says tomorrow’s technology will fundamentally differ from
cotton mills, steam engines, or washing machines. Automation will be less like
post-WW2 demobilisation for soldiers, and more like the car for horses.
Driverless vehicles will oust truckers and taxi drivers. Hyper-intelligent clouds
will oust financial advisers, doctors, and journalists. We face more ‘natural’ or
‘technological’ unemployment than ever. Government, it is said, must enact a
basic income, because so many jobs will vanish. Also, maybe robots should
become ‘electronic persons’, the subjects of rights and duties, so they can be
taxed. This narrative is endorsed by prominent tech-billionaires, but it is flawed.
Everything depends on social policy. Instead of mass unemployment and a basic
income, the law can achieve full employment and fair incomes. This article
explains three views of the causes of unemployment: as ‘natural’, as stemming
from irrationality or technology, or as caused by laws that let people restrict the
supply of capital to the job market. Only the third view has any credible
evidence to support it. After WW2, 42% of UK jobs were redundant (actually,
not hypothetically) but social policy maintained full employment, and it can be
done again. Unemployment is driven by inequality of wealth and of votes in the
economy. Democratic governments should reprogramme the law: for full
employment and universal fair incomes. The owners of the robots will not
automate your job away, if we defend economic democracy.

Key words: Robots, automation, inequality, democracy, unemployment, basic

income, NAIRU, sheep, Luddites, washing machines, flying skateboards

JEL Codes: E62, E6, E52, E51, E50, E32, E12, E00, E02, D6, J01, K1, J20,
K31, J23, J32, K22, J41, J51, J58, J6

Acknowledgements

I am very grateful for discussion with Jason Hickel, Christopher Markou, Adam
Coutts, Simon Deakin, Liv Jores, Gavin Kitching, James Hurrell, Andrea
Bertolini, Julian Huppert and Shaun Hargreaves Heap. No robots or electric
sheep were harmed in writing this article. All charts are free to copy.

Further information about the Centre for Business Research can be found at:
www.cbr.cam.ac.uk
 1.Introduction

Tomorrow’s technology will let us make a paradise on earth, where scarcity and
poverty will be forgotten, and yet we remain consumed by visions of dystopia.
As inequality has become more extreme, a busload of billionaires own more
wealth than half the planet,1 and many of them are saying new technology (that
they aim to own) will mean mass unemployment. The solution they propose is a
basic income. Facebook owner, Mark Zuckerberg said ‘technology and
automation are eliminating many jobs’ and we ‘should explore ideas like
universal basic income to give everyone a cushion to try new things.’ 2 Tesla
owner, Elon Musk said ‘Twenty years is a short period of time to have
something like 12-15 percent of the workforce be unemployed’ and a basic
income is ‘going to be necessary’ because there ‘will be fewer and fewer jobs
that a robot cannot do better.’3 The foundation of eBay’s Pierre Omidyar, says
‘automation is replacing traditional jobs,’ and the gig economy ‘may make
employment far less stable and reliable for supporting a livelihood.’ 4 Advocates
of this narrative respond sharply to criticism: a contributor to the Washington
Post, owned by Amazon billionaire Jeff Bezos, wrote ‘Sorry, but the jobless
future isn’t a luddite fallacy’.5

There is intellectual support in economics and sociology for the narrative. 6 A

basic income (if not its tax source) enjoys cross-ideological support from Milton
Friedman to Guy Standing.7 Basic income can be seen as a way to decrease
welfare cost, or social stigma. Some think it could eliminate the collective
welfare state, others that it completes the welfare state. But once the shape of a
basic income plan is agreed, we still must ask is ‘automation’ the right reason
for it? In 2013, Frey and Osborne dramatically stated ‘47 per cent of total US
employment’ is ‘potentially’ at risk from automation ‘over some unspecified
number of years’.8 They developed this argument with Citibank Research. 9
Another paper in 2017 (part-funded by a Facebook billionaire) says robots will
replace most human functions and write a best-selling book by 2049. Like Frey
and Osborne, its method was surveying 352 ‘AI experts’. 10 These figures (if not
the reasoning) have reached viral proportions in social media, and academia,
with citations mounting as if ‘likes’ or ‘retweets’. The problem is predictions of
the future, ‘over some unspecified number of years’, are unfalsifiable. Like
saying there is an ‘end of history’,11 or ‘skateboards might fly’,12 this
‘technological forecasting’ has all the shiny appeal, but also the intellectual
integrity, of a crystal ball.

This article explores how legal policy can create full employment and fair
incomes, not just mass unemployment and basic incomes. Part 2 discusses three
1
main views of the causes of unemployment. One says unemployment is
‘natural’, and full employment will accelerate inflation. A second says human
irrationality drives unemployment, and if not technology may. A third view, less
discussed outside times of crisis,13 says unemployment is a policy choice. Law
can prevent or enable restrictions on the supply of capital to the job market. The
law of contracts, property, corporations, labour and more, determines
employment. Law can guarantee full employment, and advance social welfare
for everyone. In its simplest dimensions, and as the owners of Animal Farm
eventually knew,14 technology will mean social prosperity, if the gains are not
trottered away by a few.

We cannot know the future, but we can learn from the past. Part 3 explores five
periods of history where technology was changing fast. First, enclosures for
sheep (maybe the first disruptive ‘app’15) and the textile industry drove Luddite
and Swing riots in Britain. If people are excluded from technology’s benefits,
paid for by their labour, they revolt. Second, after WW2, demobilisation in the
UK meant 42 per cent of all workers, from armies to munitions factories, moved
to civilian production. The change was immediate, huge, and far more complex
than automation. Chaos after WW1 encouraged careful thought in WW2, for
Full Employment in a Free Society.16 Third, mass produced motors did mean
mass redundancies: of horses. In the US, a staggering 88 per cent of horses lost
their jobs, over 45 years. But human beings will not share the fate of horses.
Horses do not write law. People do. Fourth, the humble washing machine was a
revolutionary technology, saving labour, improving gender equality. If people
internalise the gains of automation, human development advances rapidly. Fifth,
while artificial intelligence is remarkable, the evidence suggests computers will
not develop cognitive functions.17 While the basis of a computer’s ‘intelligence’
is ‘0’s and ‘1’s, Cartesian self-awareness seems far away. As most teachers
know, rote-learning is not critical thinking. Data is not knowledge. Doing is not
understanding. Until science fiction becomes fact, and AI hype moves beyond
clickbait, technology’s value will remain what it always has been: a way both to
empower and accentuate the uniqueness of the human mind.

History is clear that change can mean redundancies. So part 4 asks, does the law
enable everyone to share fairly in technology’s gains? The answer is ‘not yet’.
We can and should ‘reprogramme’ the law. First, full employment must be
guaranteed by law, while regional investment and training must advance. The
votes in the economy must be distributed fairly, to prevent artificial restriction
of the supply of capital to the job market. Regular investment, regardless of
conflicting interests of private or public actors, must be maintained by funds
where contributors have a vote. Second, social security must be universal. A
2
‘universal basic income’ is a powerful idea, but it must not, like the disastrous
Speenhamland system in the 19th century, subsidise poverty-pay employers.
Third, people who benefit from robot ownership must bear all the risks of their
enterprise. Robots are products, not persons. The notion of ‘electronic
personality’ could frustrate stricter enterprise liability. To advance universal fair
incomes, asset owners must pay their fair share of tax. ‘Taxing the robots’ is fine
as a metaphor for taxing robot owners, alongside all corporate profits. Part 5
concludes: robots will not automate your job away, if we defend economic
democracy.

2. Three views on unemployment’s causes

Before discussing technology’s impact on jobs and how to respond to

automation, it makes sense to engage the contest of ideas on unemployment’s
causes in the 20th century. The breadth of policy charges the debate. Labour law,
banking, corporations, property, contracts, tax and constitutional law all interact
with administrative discretion in employment policy. Political discourse often
does the subject little credit. In one corner it is common to hear the unemployed
are responsible for unemployment, because they are lazy. 18 In another, it is often
said there is a conscious plan to manufacture a reserve army of the unemployed,
to press wages down.19 Credible theories tend to focus on objective factors: how
social institutions and behaviour affect employment. They also acknowledge the
desirability of full employment, and social security, as in the Universal
Declaration of Human Rights of 1948.20 Three main theories are that (1)
unemployment has a ‘natural’ rate, and attempts to go below accelerate
inflation, (2) unemployment results from irrational human behaviour or maybe
technology, and (3) law determines employment by altering the supply of capital
to the job market. Only the third view has any credible evidential support.

2.1 A ‘natural’ rate of unemployment

A first main view, popular from the late 20th century, is unemployment has a
‘natural’ rate, worsened by labour or social rights. In 1967, as President of the
American Economic Association, Milton Friedman argued the ‘natural’ rate was
something that no government’s monetary or fiscal policy could sustainably
reduce. Apparently unemployment ‘can be kept below the “natural” rate only by
accelerating inflation.’21 If monetary authorities tried to reduce unemployment
below ‘3 per cent as the target rate’, but ‘the “natural” rate is higher than 3 per
cent they will trigger inflation.’22 We ‘cannot know what the “natural” rate is’,23
said Friedman. But apparently it came from ‘legal minimum wage rates’, pro-
3
labour public procurement and ‘the strength of labor unions’, which made ‘the
natural rate of unemployment higher than it would otherwise be.’ 24 Friedman
argued that his theory was similar A.W. Phillips’ work from 1958. Phillips
showed between 1861 and 1957, UK wages tended to rise when unemployment
was low, and wages fell when unemployment was high.25 This confirmed an old
logic, starting with Adam Smith: fuller employment raises worker bargaining
power.26 Higher unemployment reduces worker power. When bargaining power
is less unequal, wages increase. But without any evidence, Friedman argued all
inflation (not just wage rises) accelerates with full employment. This justified
using monetary policy only to concentrate on price stability, apparently
unconnected to full employment, and to undo all labour rights.27

Friedman’s theory was not new.28 In 1950, Friedrich von Hayek had already
argued that governments risked inflation if they create full employment. In
Hayek’s words, if monetary policy or fiscal stimulus were used to boost enough
demand to pay for ‘the kind of services... the unemployed offer’, it would have
to be ‘of such a magnitude as to produce major inflationary effects.’ 29 That is, a
government could not choose lower unemployment at the cost of slightly higher
inflation: if unemployment was suppressed below the natural rate, the
acceleration of inflation would be out of control. Then, said Hayek, governments
would be driven ‘to control ever increasing parts of the economy.’ Hayek did
not feel the need to provide evidence for this ostensible phenomenon because,
with his bestselling Road to Serfdom in mind,30 it was ‘by now too well known
to need elaboration.’31 This was the start of trying to find a supposed natural rate
of unemployment which would not ‘accelerate’ inflation. The unknowable
‘natural’ unemployment rate became ‘NAIRU’ - the ‘Non-Accelerating Inflation
Rate of Unemployment.’

Among their followers, Friedman and Hayek seemed vindicated when the
Organization of the Petroleum Exporting Countries unilaterally raised oil prices
in 1973. This OPEC crisis had many causes, but the immediate trigger was the
Yom Kippur War, in the long and desperately sad Arab-Israeli conflict. Arab
countries had attacked Israel in retaliation for Israel’s Six Day War in 1967. The
US government sided with Israel. Arab oil producers raised their prices. As oil
cost more, oil-dependant economies had less money to invest. As the costs were
distributed through the market, there was stagnation, and few ways to avoid
inflation: ‘stagflation’. Government practice had been to offset volatility in
private and international investment with more spending. But now this policy
had a price tag nobody could avoid. Despite being inflicted by despots, 32
stagflation said to NAIRU disciples that full employment in a free society was
impossible, without accelerating inflation. Certainly it was true that
4
unemployment rose after OPEC. From 1945 to 1971, the UK had between 1.2
and 2.7 per cent. Inflation from 1952 to 1967 averaged 3.12 per cent.33 After
OPEC, inflation surpassed 24 per cent, and bad became the norm.

But why when employment began to recover in 1977, did it stay so high in the
1980s and 90s? Unemployment in the UK reached 12.9 per cent in 1983, and by
2017 still never went lower than 4 per cent. The days of 1 to 2 per cent, and full
employment (and without underemployment) seemed to be lost.

In 1977, Friedman advanced more thoughts at his Bank of Sweden prize

lecture,34 still without evidence. Why were unemployment and inflation rising
together? He said squarely ‘there is no stable trade-off between inflation and
unemployment’. But supposedly a natural rate of unemployment still existed,
consistent with real prices and ‘perceptions’. 35 The natural rate had ‘clearly been
rising’ because ‘women, teenagers and part-time workers’ were working, but
moved jobs more, and unemployment insurance was ‘more generous’.36
Supposedly before the OPEC crisis ‘in 1973, most countries show a clearly
marked association of rising inflation and rising unemployment’. 37 If inflation
and unemployment now rose together (not an inverse, but a positive correlation)
only a ‘modest elaboration of the natural-rate hypothesis’ was needed. This was
‘a temporary phenomenon that will disappear as economic agents adjust their
expectations to reality’, though ‘we do not know what a complete adjustment
will consist of.’38

5
The ‘reality’ was that Friedman and Hayek’s theory was always evidence-free: a
theory that held unless it did not. Clear in 1977, 39 and clear now, the charts
below show the logical correlation between fuller employment and higher wages
in the Phillips curves. But despite the assertions of NAIRU theorists that ‘wage
inflation is a crucial component of price inflation’, 40 no correlation between
fuller employment and inflation exists in the UK, US, or elsewhere. Full
employment, rising wages, and zero inflation will co-exist when production
expands. Prices do not rise if people produce more: everyone’s welfare
improves.
The curve of unemployment and wage rates from A. W. Phillips’ original charts:

The random non-correlation of UK and US unemployment and inflation:

6
Many economists opposed natural rate theory. In 1980, Shaun Hargreaves Heap
explained that on its own logic the greater danger was to overshoot the
(unknowable) natural unemployment rate: this would entrench higher
unemployment, because people who are out of a job for a long time find it
increasingly difficult to get back into work.41 Like a magnetic field lingers after
the charge is switched off, this became known as ‘hysteresis’. 42 In 1982, Kim
Clark and Lawrence Summers, also rejecting a natural rate’s existence,
explained how during WW2 in the US, an unprecedented number of women
entered the workforce whilst full employment was maintained. Instead of
blaming women or young people, the tendency to find entering work harder
after being out of work partly explained unemployment.43

But evidence did not stop NAIRU economics,44 and nor did better theory. In a
‘special’ issue of the Review of Economic Studies introduced by Richard
Layard,45 it was said natural unemployment was rising because ‘people are
trying to achieve too high real wages.’46 The ‘rise in secular unemployment in
Britain’, said Martyn Andrews and Stephen Nickell, was caused by benefits,
moving workers, ‘the introduction of employment protection legislation and the
rise in union power’.47 Backed by some formidable linear equations, they
asserted the oil shocks caused but one third of 1973-1977 inflation.48 By 1990,
Samuel Bentolila and Giuseppe Bertola composed a table of employment
protection laws in four countries (wholly unreferenced49) and were confident that
these ‘costs’ were the cause of ‘Eurosclerosis’.50 Meanwhile, the Organization
for Economic Cooperation and Development had begun regular reports to
encourage cuts to job security.51 In 1994, Paul Krugman wrote that the
‘disincentive effects of welfare state policies’ and the lack of ‘flexibility of the
labor market’ causing unemployment was ‘conventional wisdom’.52 Eliminating
social and job security might be ‘harsh’ and ‘some people end up on the scrap-
heap’, but the ‘wisdom’ was clear: ‘Any tax or transfer payment distorts
incentives.’53 In just 30 years, it seemed a generation of economists had been
turned against the Universal Declaration of Human Rights.54

NAIRU economics was close to manifesting itself in law. At work, the US

remained a relatively rights-free jurisdiction,55 but it did have the Social Security
Act of 1935, and a statutory commitment to ‘maximum employment’ in the
Federal Reserve Act of 1913.56 In other OECD countries, labour rights and
social security were robust, and countries like the UK, Japan and Australia had
maintained full employment for over 25 years, while Germany did for 15. But
written into the Bank of England Act 1998, and the Treaties underpinning the
European Central Bank, was the goal of ‘price stability’ (not exactly ‘zero’ or
‘low inflation’), which seemed to be interpreted by bankers as more important
7
than ‘full employment’.57 In 1996, Janet Yellen, future Federal Reserve chair,
wrote that NAIRU reflected ‘structural aspects of the economy’,58 while Bill
Clinton’s economic advisers were calculating the unknowable natural
unemployment rate: an endeavour in which they conspicuously failed. 59 More
than this, US and EU governments had started to place more duties and
sanctions on unemployed people to search for jobs that did not exist. Everyone
had to be ‘activated’ to get full employment, it seemed, except government
itself.

Meanwhile, in 1997, Stephen Nickell had admitted ‘there is no evidence in our

data that high labor standards overall have any impact on unemployment
whatever’.60 But many colleagues still preferred Nickell’s more confident
position from 1982. That was, said Richard Layard, cutting labour rights for ‘a
flexible system of wage differentials. Nothing else will do the trick.’ 61
Economics papers were now written arguing simply that labour rights kill jobs. 62
But even then, the better side of the evidence (and overwhelming evidence
today) does not merely refute the notion that labour rights, especially
employment protection, harm the economy. From the best legal database of 117
countries, using econometric analysis at Cambridge’s Centre for Business
Research, it appears that labour rights advance productivity, economic growth,
equality, and therefore human development. 63 Labour rights also spur
innovation.64

Critically, the original NAIRU argument (governments seeking full employment

would accelerate inflation) was quietly segregated or forgotten. 65 So were some
hard facts. Even ignoring chronic under-employment and wage insecurity, 66
unemployment was always a political choice, nowhere less than the US.

The chart below shows that from 1952-2016, Democrats always left the White
House with unemployment lower: down 12% overall. Republicans always left
the White House worse: up 14% overall. The chart also calculates that, by
including mass incarceration, the effective unemployment rate was around 1.3
per cent higher than the recorded rate in recent decades. Putting people in prison
is not a way to get employment in a free society. 67 Is the abysmal Republican
record on jobs an accident? Probably not. At the very least, a natural rate of
unemployment was a set of evidence-free conjectures, propped up by a
chimerical hand.

8
2.2 Irrationality + technology

A second view of unemployment is it results from ‘irrational psychology’, or

possibly from technology, both central to John Maynard Keynes’ theory. 68 In the
Great (and unnatural) Depression, Keynes argued the idea that unemployment
was voluntary must be wrong.69 If total spending or effective ‘aggregate
demand’ for goods, services and labour dropped away, involuntary
unemployment results.70 Poorer people have a higher ‘propensity to consume’
their income, because they need to spend everything as they get it. Wealthier
people save, or invest in the financial sector.71 Before the Wall Street Crash,
more and more affluent Americans invested in the stock market for retirement
and, said Keynes, the ‘mass psychology of a large number of... individuals is
liable to change violently’.72 Even skilled business people have an ‘extreme
precariousness’ in their ‘basis of knowledge’ to calculate investment returns in
future.73 Triggered by speculative swings, cycles of boom and bust result when
business over-invests in productive capacity, stock piles up, production is cut
back, and business makes people redundant.74 This slows the ‘velocity of
money’ flowing through the economy,75 and with it aggregate demand. Asset
owners cannot be relied upon to boost aggregate demand in a depression,
because they have a lower propensity to consume. So, active government should
balance business volatility and irrational thrift.

Keynes was careful to say we ‘should not conclude... everything depends on

waves of irrational psychology’, but human irrationality was at the fore of his
analysis. Law was almost absent.76 Even post-Keynesian economists have not
9
stressed legal reform, rather than focusing on mystic and supposedly inevitable
business ‘cycles’ and market ‘forces’.77 But legal reform has always been central
to ensure social institutions became more rational, despite our individual
irrational behaviour. First, from the world’s first stock market crash, caused by
speculation on South Sea Company shares,78 the courts of equity prohibited any
possibility of a conflict of interest by people (like stock brokers, asset managers,
trustees, or Chancellors) who manage ‘other people’s money’.79 Second, the
common law from 1766 required a positive duty of disclosure about material
facts in any ‘contract based upon speculation’.80 This general principle of
‘utmost good faith’ was confined to insurance contracts by more regressive
courts,81 but expanded once more in 1878 to all contracts for the sale of
company shares.82 Commercial goods contracts might not require positive
disclosure,83 but most contracts on financial markets did.

The trouble was, US contract, trust, and corporate law branched into 50 new
systems from 1776,84 and in every jurisdiction, the law was under-enforced. 85
Meanwhile, around WW1 the US Supreme Court had declared public pensions
or social insurance would be unconstitutional: they violated an ostensible
principle of ‘freedom of contract’.86 Americans still sought dignity in old age,
but now there was little choice except bank savings, and stocks on Wall Street,
to build up a retirement fund. This was the start of a great stock market
expansion, a mass separation of ownership and control, as company shares came
to be held many millions of individual investors.87 Unprepared, US law had few
adequate duties of disclosure on companies, brokers or promoters.88 People
thought they were sold old age dignity in the boom, but found they bought a
patch of Florida swamp after the bust. The strangling of social security,
consumer and investor protection (not just problems with money supply 89)
caused the Wall Street crash.

So when the New Deal was formulated, the Securities Act of 1933 and the
Securities and Exchange Act of 1934 mandated disclosure for all securities
contracts, forbid conflicts of interest, and required public enforcement. The
author of those laws and President Roosevelt’s historic speech on a new
economic constitution, A.A. Berle, provided a model that was rolled out around
the world.90 The Social Security Act of 1935, with the US Supreme Court
shamed into reversing its ‘freedom of contract’ dogma,91 ensured everyone had
the right to an old age pension and unemployment insurance. 92 But if law reform
worked back then, why was there a global financial crisis from 2007? The
answers are much the same. Instead of retirement, the social right speculators
monetised was the right to a home.93 Instead of company shares, the contracts
that evaded disclosure rules were derivatives on sub-prime mortgage debt.94 As
10
the Bush II administration eviscerated federal regulators,95 conflicts of interest
among rating agencies and banks became rife once more.96 Law was able to
eliminate irrational business cycles: in Europe and the Commonwealth there was
no stock market or bank (rather than political) crisis since WW2. But global
banks traded US securities. Keynes was right that individual human behaviour
could be irrational. The idea that social institutions must be irrational, that this is
‘human nature’, is evidently wrong.

What about technology? In 1930, in The Economic Possibilities of our

Grandchildren, Keynes argued society was being afflicted by ‘technological
unemployment... due to our discovery of means of economising the use of
labour outrunning the pace at which we can find new uses for labour.’97 Keynes
himself quickly added that any unemployment from technology was ‘only a
temporary phase of maladjustment’ and ‘mankind is solving its economic
problem.’ As technology continued to save labour, thought Keynes, society
could be 4 to 8 times richer by 2030, and we could endeavour to ‘make what
work there is still to be done to be as widely shared as possible’ by having a 15
hour working week.98

Keynes was certainly not alone, seeing technology as a problem but also an
opportunity. Joseph Schumpeter wrote in 1942 that new technologies were like a
‘perennial gale of creative destruction’.99 A ‘rapid change’ could disorganise
industry and ‘create avoidable unemployment’. Yet we could, said Schumpeter,
‘avoid coming down with a crash’, and make an ‘orderly retreat’ instead.100 Most
popularly, George Orwell in 1945 wrote in Animal Farm how the animals
‘listened in astonishment’ to how ‘fantastic machines... would do their work for
them while they grazed at their ease in the fields or improved their minds with
reading and conversation.’101 Once their human oppressors had been evicted, a
future of leisure, a 3 day week, was possible through technology, and collective
endeavour. In this way, Orwell, Schumpeter and Keynes all believed that
technological change could have an impact upon jobs. Did technology mean
more work, or more freedom? The answer, they said, depends on social policy.

In stark contrast, a view spread recently that ‘technological unemployment’

could be permanent. In 1995, in The End of Work, Jeremy Rifkin argued one of
the great, unspoken tragedies of the American labour force had been
mechanisation in agriculture.102 This (not absence of labour rights) led to mass
post-war redundancy of farm workers in the south, 103 and was central in
triggering Martin Luther King Jr’s ‘March on Washington for Jobs and
Freedom’ in 1963.104 And by the 1990s, computers and robots had become like a
terminator, ‘steadily moving up the office hierarchy, subsuming not only routine
11
clerical tasks but even work traditionally performed by management.’ 105 The
solution to mass unemployment, argued Rifkin, is that a new value added tax in
the US should fund a ‘social wage’ for people to do volunteer work. 106 Much
like Rifkin, in 2013 Erik Brynjolfsson and Andrew McAfee argued that US
unemployment, and workers’ stagnating incomes, result from ‘losing the race
against the machine.’107 They argued government should educate and invest
more, but continue ‘resisting efforts to regulate hiring and firing.’ 108 In their next
book, drawing on Friedman, they advocated a ‘negative income tax’ where
people out of work, and in work, receive a basic income up to a limit to act as a
‘subsidy on labor’.109

What is truly surprising, especially given the authors’ social democratic

views,110 is that the ‘ossification of American labor law’ is missing to explain
US unemployment and inequality.111 In essence, since 1976, when a disastrous
US Supreme Court case held the rich could spend unlimited money on political
campaigns as part of ‘free speech’, there was an almost total halt in American
social progress.112

This chart shows, as labour was attacked, and trade union density fell, inequality
measured by income share of the top 1% of US earners soared. The same pattern
holds in all developed systems where unions are the main channel for voice at
work.113 In the US, collective agreements made by unions are critical to regulate
hiring and firing because there is just one federal Act to protect job security. 114
This worsens unemployment. The easier it is for a conflicted, irrational authority
figure to point a little finger and bark “you’re fired”, the worse is an economic
crash:

12
This chart shows the post-financial crisis peaks in unemployment under three
legal alternatives. In the US, with employment ‘at will’, unemployment soared
by 5.3 per cent from May 2008. In the UK employees must have reasonable
notice, before a fair dismissal, and redundancy pay enforceable in a Tribunal. 115
Unemployment rose 2.8 per cent. In Germany, there is better notice and
redundancy pay, but also employee elected work councils may veto or defer
dismissals.116 Unemployment rose just 0.8 per cent, because German work
councils negotiated working time reductions and pay restraint, including
executives, to stop job losses.117 Again, if it is easy for conflicted, irrational
managers to make people unemployed, there will probably be more, not less,
unemployment.118 Cutting edge econometric analysis at the Centre for Business
Research’s shows this is right. Job security improves employment, equality and
prosperity.119

But the narratives of Rifkin, Brynjolfsson and McAfee are less troubling than
the most recent tech predictions. In 2013, Carl Frey and Michael Osborne
argued ‘47 per cent of total US employment’ was ‘potentially’ at risk from
automation in a working paper abstract.120 They also said inequality was due to
technology, without mentioning changes in labour law. 121 Their paper was not
accepted into a peer reviewed journal until 2017, but by then the ‘47 per cent’
claim was cited across the media, and soon in academia: over 900 citations on
Google Scholar by April 2017, over 1600 citations by January 2018. Only 20 of
these citations, however, appeared to read past the abstract: that change would
be ‘over some unspecified number of years’.122 Moreover their ‘method’ was
that ‘with a group of [machine learning] researchers, we subjectively hand-
labelled 70 occupations, assigning 1 if automatable, and 0 if not’ by ‘eyeballing’
13
different tasks.123 In 2015, Frey and Osborne secured funding from Citibank
‘Research’,124 and similar predictions went viral. These were not based on any
defensible academic method, but rather ‘research’ papers from McKinsey, 125
PwC,126 Deloitte (working ‘with researchers at Oxford University’), 127 and other
consultancy businesses trying to boost demand for themselves by stoking social
anxiety. More academics have learned the game. The Oxford ‘Future of
Humanity Institute’ released a 2017 paper (part-funded by a Facebook
billionaire) saying robots will replace most human functions, and write a best-
selling book by 2049. It reached this conclusion by surveying 352 ‘AI
experts’.128
It is questionable whether technology predictions can be classed as academic
enquiry. They have certainly received withering criticism by academic
economists without corporate funding,129 and even the Obama White House.130 If
the job of making predictions about future employment could be automated,
what would the robots say? Would they reject the ‘research’ by McKinsey,
PwC, or Deloitte as little better than a crystal ball, tarot, or tea leaves? Given the
involvement of corporate
finance, all such predictions are
highly vulnerable to conflicts of
interest. But also, if mass job
loss is predicted ‘over some
unspecified number of years’,
how can this really mean
anything? For instance, from
The Canterbury Tales in 1387,
it would appear there has been a
mere 46 per cent redundancy
rate for Geoffrey Chaucer’s characters. That is, in 631 years, a rate one per cent
under Frey and Osborne’s prediction. Now, it might be objected this Canterbury
Tales study has an unscientific method, uses a questionable categorisation of
jobs, with lazy application of statistics. But if all this is true, it has one thing
which the new methods of ‘research’ do not: evidence.

2.3 Law + capital supply

A third main view is that, far from being natural, irrational or technological,
unemployment results when law enables the restriction of capital supplied to the
job market. Capital, said Adam Smith, is the part of one’s stock that is used to
make ‘revenue’, but is not for ‘immediate consumption’. 131 That is, capital is
property used for production.132 The distinction was central to New Deal and
post-war UK government policy. First, in 1951 according to M.S. Eccles, Chair
14
of the Federal Reserve, Wall Street crashed and depression had ensued because
wealth accumulated as ‘idle or hoarded funds’. Speculation and consumer debt
maintained production and jobs for a while. But like ‘a poker game where the
chips were concentrated in fewer and fewer hands’, when most people’s ‘credit
ran out, the game stopped.’133 If there had ‘been a better distribution of the
current income’ with ‘lower prices or higher wages and with less profits to the
corporations and the well-to-do, it would have prevented or greatly moderated
the economic collapse.’134 Property carries responsibility,135 and people holding
productive property had more.

Second, the 1944 UK White Paper, Employment Policy saw swings in

investment as the barrier to full employment. Consumer spending is relatively
stable, and government could rationalise its spending. But private and
international investment were volatile. Government should balance both,
shifting projected spending forward or back in five year terms. 136 At the same
time, large employers acquired duties to hire disabled people up to a quota, 137
and government ensured investment into under-developed regions, so the
country did not languish behind cities.138 The Board of Trade would create
development areas if ‘there is or is likely to be a special danger of
unemployment in that area’.139 John Wardlaw-Milne, a backbench MP,
proclaimed this was the ‘very antithesis of private enterprise... bureaucracy and
Socialism carried to the last limit.’ 140 And with Conservative, Labour and
Liberal consensus, it worked for 30 years.

The theory behind full employment was not all that new. In 1912, Sidney Webb
had advocated government spending to counteract shifts in foreign trade and
private consumption. He also argued for public labour exchanges to match
workers to job vacancies efficiently, for people to be guaranteed a minimum
number of hours a week instead of being ‘on call’ (with ‘zero hour contracts’),
and for progressive reduction of everyone’s hours of labour over the long
term.141 For the relatively few “won’t works”, those people should be given
training, support and discipline, but not crime and punishment.142 By contrast, in
1943 Michal Kalecki argued that discipline was in fact the very reason
unemployment existed. If there were full employment, ‘workers would ‘get out
of hand’ and the ‘captains of industry’ would be anxious to ‘teach them a
lesson’.’143 Conscious or not, Kalecki was saying, employers have an incentive
to restrict the supply of capital for jobs, to reduce the bargaining power of
workers. If unemployment is high, employers as a group have few reasons to do
anything to reduce it.144
But after World War Two, full employment was maintained by political
consensus in the UK, as it was in Australia, Japan and Germany. 145 In 1962,
15
A.A. Berle added that a legal duty for full employment could extend to
corporations. Just as government would be responsible to maintain spending,
companies ‘ought not to hoard when expenditure is needed’ and they might need
‘to slow down under some conditions.’146 The most fascinating fact was full
employment cost government very little. According to Robin Matthews, far
from a ‘Keynesian revolution’ with persistently or cyclically high government
spending, the UK government post-war had a constant current account surplus.
This indicated spending restraint.147 In other words, once government had made
the initial investments, and then committed to do ‘whatever it takes’, this
maintained business confidence, economic stability, and full employment.148
After the OPEC price hikes, once confidence resumed, full employment would
have been possible again.

The weakness in theories that have emphasised law’s role in the supply of
capital for jobs is the prominence given to government to guarantee investment.
Indeed, Kalecki already feared a government hostile to the effect of full
employment on labour relations would scrap it all.149 Why should democratic
society expose its prosperity to interest groups that would undermine its very
basis? After OPEC, full employment was abandoned. An era of recurrent crisis
took hold. The financial sector inflated. Corporations began hoarding more
assets.150 Economic theories were chosen to justify it. The return of corporate
cash hoarding since the 1990s has been nothing less than extraordinary, 151 and
nowhere more than among tech-corporations. By 2016, US non-financial firms
held $1.68 trillion in cash, $1.2 trillion overseas. Apple, Microsoft, Alphabet,
Cisco and Oracle alone held $504 billion in cash. 152 That money, $504 billion
hoarded in just five tech companies, was enough in 2017 to give every
unemployed person in the US a full-time job on $12 an hour for a four year
presidential term.153 This cannot be regarded as an outcome that tech
corporations desire: it is the socially irrational product of what seem to be
individually rational firm decisions to save. This concentration of wealth,
withdrawn from economic production, challenges the very legitimacy of the
modern corporation and private property.154 The law is enabling an artificial
restriction of the supply of capital to the job market, and damaging human
development. Unemployment is caused by law, not individuals or their
technology. History affirms more this is true.

3. Technology and jobs: five lessons from history

We cannot know the future of technology’s impact on jobs, but we can learn
from the past. The best guides for what to do (not best guesses of what might
ensue) are times of significant historical change. Five are instructive: (1) the
16
shifts from agriculture to wool and mechanised textiles, (2) the massive post-war
demobilisations, (3) the impact of motor vehicles on horses, (4) the washing
machine, (5) computerisation and AI. History’s syllabus shows technology
improves lives, if people are organised.

3.2 Sheep + Luddites

A first lesson starts with the displacement of agricultural labour in Reformation

England. It shows how people react if they cannot share in technology’s gains.
By 1516, a disruptive, even devastating ‘technology’ had multiplied the value of
land as never seen before. ‘Sheep,’ wrote Sir Thomas More,155

have now apparently developed a raging appetite, and turned into man-
eaters.... Each greedy individual preys on his native land like a malignant
growth, absorbing field after field, and enclosing thousands of acres with
a single fence. Result - hundreds of farmers are evicted. They’re either
cheated or bullied into giving up their property, or systematically ill-
treated until they’re finally forced to sell... men and women, husbands and
wives, widows and orphans, mothers and tiny children.... you can’t run a
farm without plenty of manpower.

The sheep – or more specifically, wool – market, wrote More, was ‘almost
entirely under the control of a few rich men, who don’t need to sell unless they
feel like it, and never do feel like it until they can get the price they want.’
Worse, this led to a ‘great army of unemployed’, and turned people ‘into
beggars or thieves.’156 More recommended reviving employment for ‘plenty of
honest and useful work’.157 But the response of the state was ever stiffer
penalties. Under the Vagabonds Act 1530, it was decided that unjustified
begging should be punished with a good whipping. This did not seem to work,
and meanwhile in 1535 More was executed. So, the Vagrancy Act 1547 placed
beggars in servitude, or slavery for a second ‘offence’. Kett’s Rebellion in 1549
led to a repeal,158 but criminalisation of the poor and enclosure went on.

With wool market expansion, in 1589 an inventor named William Lee created
the first mechanical knitting machine. Queen Elizabeth I refused Lee a patent,
fearing stocking workers would become redundant. She asked him to invent
something for silk instead.159 Whether this slowed mechanisation or not, steam
power sped development up. In 1769 Samuel Wise, a clockmaker, attached a
cog to stocking machines that turned with steam. 160 Steam and machine meant
unprecedented speed. Workers were made redundant, and in 1779 an apprentice
named Ned Ludd was said to have smashed machines in Leicester after being
17
whipped for vagrancy. Even if this was a myth, Ludd became legend, as
sporadic ‘Luddite’ and ‘swing’ riots broke out. The Protection of Stocking
Frames, etc. Act 1788 section 4 required 7 to 14 years’ transportation for
destroying machines, just as the ‘First’ Fleet of Arthur Phillips anchored in
Australia. The Destruction of Stocking Frames, etc. Act 1812 imposed a death
penalty.161 Lord Byron opposed it in his maiden speech.162 The machines
‘superseded the necessity of employing’ workmen, said Byron, ‘who were left in
consequence to starve’. ‘It is the mob that labour in your fields, and serve in
your houses... and can also deny you, when neglect and calamity have driven
them to despair.’163

Unrest continued for decades, as real wages stalled and fell. It was not merely
that people’s incomes convulsed with the wild paroxysms of agricultural prices.
Especially from 1747, and again from 1795, long-term down-swings in wages
were punishing, recurring until the labour movement formed.164

Did wages decline because of technology in the Industrial Revolution? No. Law
caused the decline. First, the Master and Servant Act 1747 wrought a massive
extension to all classes of workers of punishments for leaving employment, and
for ‘justices of the peace’ to cap, fix and repress wages.165 There was also no
right to unionise and strike,166 but employers could form partnerships,
corporations, and dismiss workers regardless of the social cost.167 Second, on 6
May 1795, a meeting of poor law magistrates at a ‘Pelican Inn’, in
Speenhamland, Berkshire, decided unemployed people, and the working poor,
should get relief because grain prices had risen (again). 168 They rejected a
minimum wage.169 This reckless, paternalist act was a cue for employers to use

18
parish subsidies as an excuse to cut employment wages. It was not merely that
people in work became dependent on poor law relief, whatever the Malthusian
moralising.170 Nor is it relevant that Speenhamland was replaced by an even
ghastlier system of workhouses, based on concocted evidence, in the Poor Law
Amendment Act 1834.171 As working people had no vote in politics, 172 this
‘basic income’ only lined landlord pockets, the very ‘definition of a rentier’
system.173 It made, wrote Eric Hobsbawm, ‘universal pauperism of demoralised
men who could not fall below the relief scale whatever they did, who could not
rise above it... reduced to as little as the village rich thought fit for a labourer.’174

Slow change began in 1834, when a group of agricultural labourers in

Tolpuddle, Dorset, were sentenced to transportation for making ‘unlawful
oaths’. They had formed a union.175 They did not break machinery, but
according to Sidney and Beatrice Webb their trial followed the ‘specially hard
times of 1829... outbursts of machine-breaking, rick-burning and hunger riots...
put down in 1830 by the movement of troops... 1000 prisoners, several of whom
were hung and hundreds transported.’176 As the labour movement rallied in their
support, they became the ‘Tolpuddle martyrs’, all because the government
insisted on making an example of them. Far more than Luddites, this became the
real legend of the labour movement. It was not until the Trade Union Act 1871
that unions finally became lawful, and not till the Taff Vale case, the formation
and electoral success of the Labour Party, and the Trade Disputes Act 1906, that
the right to collective action was secure. But this episode of history shows, when
they do not share in the benefits of technology, people revolt and resist.
Eventually, they organise.

3.2 War + demobilisation

A second lesson of history comes from demobilisation on the First and Second
World Wars. Unlike most wars of the 18th and 19th centuries, total war, with
conscription, employed a large minority of the population directly, and
indirectly almost everyone. As the chart in part 2(1) showed, demobilisation in
the UK after World War One led to soaring unemployment, near 18 per cent by
1922. The norm lurched above 10 per cent until the rearmament for World War
Two. It was chaos, with all the ‘dimensions of a calamity’ and British labour
fought before it accepted ‘a reduction of its standard of living’. 177 While David
Lloyd George had promised ‘homes fit for heroes’, his post-war coalition had no
credible plan for jobs.178 Directly after Armistice, 30,000 workers on Tyneside
marched against unemployment, 3,000 protested rapid demobilisation on New
Year’s Day of 1919 in the Rhondda Valley, and 50,000 iron workers and
engineers struck in the autumn. Between 1920 and 1931, over twenty Acts were
19
introduced to Parliament to tinker with unemployment policy. 179 But even after
the Great Strike of 1926, when coal miner wages had been cut by 43 per cent,
none bit the bullet: the need for law to guarantee full employment.

The United States fared better, partly because of its sheer size and shorter
involvement in the war, partly because a million men were employed on public
works in 1919. But as the chart on US unemployment in part 2(1) shows, it did
not save the Democratic Party’s hold on the White House.180 One might expect
that Germany fared worse: 6 million soldiers redundant upon the Reich’s
collapse, 800 thousand people returning from captured territory,181 its economy
in ruins, its political class in infamy. Yet against the UK and US, the immediate
rise in post-WW1 unemployment was the least.

This was because Social Democrats passed laws preventing dismissal without
good reason, involving unions and elected work councils, and organised labour
exchanges.182 It all changed, of course, with the erection or tariff barriers by US
President Harding,183 with the malevolent terms of the Versailles Treaty,184 the
resulting inflation, and then the government’s crackpot policy of hyperinflation
that wrecked the German mark.185 The Germans’ roaring twenties meant endless
humiliation, until a sociopath with a silly moustache promised respect. By 1939,
as the Nazi state launched its campaign to enslave Europe it was financially
bankrupt. Nothing could have been further from full employment in a free
society.186

After WW2, both UK and US governments would not allow a repeat. As part
2(3) explained, the UK government’s White Paper, Employment Policy of 1944
committed to full employment. In June 1945, the UK had total available
manpower of 21.5 million men and women, with 5.219 million in the military,
20
and 3.837 million producing for the military.187 That is, over 42 per cent of the
total labour force produced directly for war, and so were redundant on war’s
end. The need to redeploy people was obvious: for ‘houses to be built, shops to
be filled, factories to be transformed, plant and rolling-stock to be replaced and
export trade to be renewed and extended.’188 Even if it was ‘vain to imagine that
patches of unemployment can be eliminated altogether in the transition’ the
White Paper said they ‘should not be so widespread or so persistent’.
Government would switch capacity, arrange civilian work and products, and
dispose of surplus government stock to avoid disrupting trade. 189 This was the
basis of economic recovery.

The counterpart plan in the US involved three major Acts. First, the
Servicemen’s Readjustment Act of 1944, or the ‘GI Bill’, funded retraining and
education fees for veterans.190 Second, Congress passed the Employment Act of
1946, although far more watered down than its original draft, to ‘promote
maximum employment, production, and purchasing power.’191 Once it was
passed, it enabled the Truman government to use its influence in fiscal and
monetary policy to grow employment, even if Eisenhower would refuse to
follow. Third, the Foreign Assistance Act of 1948 led to $12.7 billion in grants
and loans for Western Europe. According to this ‘Marshall Plan’, the assistance
was to ‘provide a cure rather than a mere palliative’ for ‘restoring the confidence
of the people of Europe in the economic future’.192 The results were indeed
tremendous: it was a new era of European and American prosperity.

The real lesson from demobilisation is that, with concerted action, even the
massive shock of 42 per cent unemployment in the UK, and similar figures
elsewhere, could be swiftly contained. Demobilisation was an infinitely greater
challenge, in size and social complexity, than any credible prediction about the
effect of automation. Hundreds of thousands of people were disabled as a result
of war, both physically and psychologically. Entire economies had to shift, not
over decades as technology was rolled out, but at the instant of armistice, to ‘win
the peace’. This is not to say military expeditions did not continue: they
certainly did. But there was nothing that labour law, dismissal protection,
redeployment rights, and full employment, could not accomplish. As the
Universal Declaration said, this was the true meaning of the ‘right to work’.193

3.3. Smoke + horses

Although social policy can ensure massive shocks to employment are contained,
a third lesson of history is that technology’s replacement of labour, even if total,
can be exceedingly slow. In 1885, as professional football first became legal in
21
Britain, and the Statue of Liberty arrived in New York, Gottlieb Daimler
patented an internal combustion engine for a motorbike. A few months later
Karl Benz patented the first motor car in Germany. Their engines ran on
petroleum gasoline, refined from oil, a resource whose mass extraction was
equally recent.194 Mass motor production only began in 1908, with the Model T
Ford. Motor vehicles, cars and trucks, did not merely compete with steam
engines and rail: those could not reach specific delivery points. No, this
technology meant the horses of the world were soon out of a job.

Many could not accept the inevitable. In 1908, Mr H.B. Brown wrote in the Yale
Law Journal that for ‘park and other pleasure driving’ when cars ‘cease to be a
fashionable fad, the public will probably return to carriage and horses.’ 195 This
was, said Brown, because cars were ‘attended by a cloud of dust and smoke’,
‘the emission of a noisome odor’ and the ‘cold and heartless mechanism of the
automobile’ was to many a ‘veritable terror’. Indeed, the ‘automobile lacks one
of the most attractive concomitants of pleasure driving in the companionship of
the horse.’196 Brown urged his readers to ‘pray’ against ‘the extinction or
dethronement of the noblest of all domestic animals’. But his prayers would be
brutally run down. Relentless, unforgiving, the car stomped the US population
of horses from 26 million in 1915, down to 19 million by 1930 and 3 million by
1960. Mass horse unemployment became a reality: 88 per cent of horse jobs
were slaughtered.197

Yet the lesson this gives us is that,

even though a new technology made an
old one obsolete in every respect, the
transition (at least with matters left
mainly to the market) took at least 45
years. Factors slowing transition
include the lifespan of horses, the costs
of switching once horses were bought,
the costs of upgrading related
infrastructure networks (e.g. stables to
petrol stations), and no doubt elements
of romantic attachment. What could
this mean for drivers of cars and trucks? If driverless technology becomes real
and safe for commercial use (at whatever date that may be), redundancies may
be very slow.

Driverless vehicles and jobless drivers have become a central image in

automation discourse. So it is worth asking, will people be as jobless with
22
robots, as horses were jobless with cars? The answer is almost certainly ‘no’.
First, as the chart below shows, the actual proportion of driving jobs is relatively
modest: taxis, public transport, delivery and on-site motors. In 2016, it was near
4 per cent of the UK workforce, just under 1.3 million people. Second, many
commercial aeroplanes are computerised, and there is still a need for a pilot, not
merely to supervise the robot, but to change course, take-off and land most of
the time. Much the same is true for trains. Even in supermarkets, which have
partly ‘automated’ checkouts (and quietly turned customers into workers 198),
staff monitor the machines’ constant malfunctions. But third, even if all
computing and robotic problems can be overcome, it is doubtful that human
beings will become redundant like horses. It is not merely that human beings
enjoy the ‘companionship’ of other humans, although this is certainly true: it is
the point of ‘customer service’. It is that humans, not horses, write the law, and
will protect their own welfare. Unless some people manage to treat others as
disposable commodities or beasts, jobs can be improved.

This is not to suggest rapid

transition to automation is not
both desirable and necessary.
First, in 2015 the deaths from
avoidable traffic collisions
totalled 84,589 in Europe, 34,064
in the USA, 238,562 in India,
261,367 in China, and 1.25
million across the world.199
Vehicles driven solely by humans
cannot be seen as safe at any
speed. Everything must be done to
end the misery, and secure
everyone’s freedom from harm on
the roads. Second, many drivers
suffer from stress, loneliness and monotony.200 Automation is an opportunity to
humanise work, if drivers are fully compensated and redeployed. Third, if
driverless vehicles can be introduced, with no salaries to pay it will radically
reduce the cost of production. The gains can be made green: every ‘no driver’
vehicle could be a ‘no smoking’ vehicle too. But the transition to a safe, zero
carbon transport system will only happen with foresight and smart regulation,
not by ‘spontaneous order’.

3.4. Washing machines

23
History’s fourth lesson is technology drives human development most when
people internalise the gains. If motor vehicles were the most significant
technology rolling out near WW1, the washing machine was the equivalent
rolling around near WW2. In 2010, Hans Rosling famously described the
washing machine as being like ‘magic’, a ‘miracle’ for human development. 201
Before the washing machine, all laundry had to be done by hand, labour usually
without pay, and it was mostly seen as the job of women. A look back in
washing history reveals just how all-consuming the issue was, and how strong
the desire for a solution. In 1841, Catharine Beecher and Harriet Beecher Stowe
exclaimed: ‘How would it simplify the burdens of the American housekeeper to
have washing and ironing day expunged from her calendar!’202 Without the
technology there, they called for a common system of laundries, to save on
individual labour.

Even though washing machines began to be mass produced in the 1940s,

distribution took time, and remains a question of development. From the 1950s,
Katharine French-Fuller recorded a woman in Chile saying ‘every week all year
long—every week without fail—there was washday.... The clothes would be
scrubbed ... by a woman bending over the tub.’203 In 1985, Jean Robinson
recorded a Chinese woman as saying: ‘It’s really a pain not to have plumbing.
Every day we need about 4 or 5 buckets of water ... every day it takes more than
hour just to get the water, and once we’ve used it, I have to get rid of it.’ 204 In
other words, women from Chile to China, with access to plumbing and
automated washing, were able to save a whole day’s worth of work each week.
The washing machine was almost as powerful as collective bargaining from the
1920s and 1930s, which not only spread ‘childhood’ and ‘retirement’ for most
people,205 but (as the chart shows206)
actually created the word of a two
day ‘weekend’.207 By 2010 Rosling
estimated that 2 billion people had
access to a washing machine,
though 5 billion people still did
not.208

To save labour and advance gender equality, washing machines need public
infrastructure: a piped water network, and an electricity grid. 209 In 2015, the
World Bank recorded that just 71 per cent of the world’s population used ‘safely
managed drinking water services’, a figure stagnant since 2010, and one which
does not necessarily mean that water is piped and pumped. In 2014, 85 per cent
of people had access to electricity. Public investment in roads and fuel stations
are essential for the existence of private transport, just as public communication
24
networks enable even the most giant websites to function.210 Water and
electricity networks are the same for washing machines. When everyone can
share in technology’s gains, when people can internalise their gains through
distributed benefits, human development will be rapid. But social organisation is
essential for it to happen.

3.5 Artificial ‘intelligence’

A fifth lesson of history is also one of the future. Science fiction is filled with
optimistic and apocalyptic predictions. In 1989, Back to the Future: Part II said
we would see a flying skateboard by 21 October 2015. Of course, this was
fiction. But how different is the marketing hype on driverless cars, hyperloops,
or ‘personal air vehicles’? Similarly, in 1991, a James Cameron film projected
on 29 August 1997 computers would become self-aware, launch a nuclear war
against human beings, and develop a robot army to eradicate the rest.211 Again,
this is fiction. But how different was Rifkin’s apocalyptic End of Work? Indeed,
the net effect of computers appears to have been an increase in employment, not
an ‘end’.212 The lesson is, while the only limit of imagination is itself, the limits
of physical science follow another script.

When it comes to artificial intelligence, computational power exceeds human

beings in strength and speed. But ‘intelligence’ means ‘knowledge’. 213 Data is
not knowledge. Human beings have cognitive power, and that means
‘understanding’.214 Doing is not understanding. Human beings can interpret rules
because they understand their meaning and purpose. Following rules is not
interpreting them. Computers might process regulations, but is it possible to
programme a sense of equity, let alone justice?215 In a democratic society, justice
cannot be reduced to an arithmetic or geometric equation,216 because the
parameters of society’s norms are recreated by the very act of interpretation.
Can that be programmed? Computers might play and win games like chess or
go.217 But as every teacher knows, rote learning is not critical thinking. Human
intelligence derives from at least five organic senses. It is not clear that a set of
‘0’s and ‘1’s, or even quantum computing, could logically be a basis for
Cartesian self-awareness.218

While it is both fascinating and desirable that ‘machine learning’ is developing,

the philosophy of language shows why AI may not achieve the unique qualities
of the human mind. With Bertrand Russell, one of the greatest philosophers of
the 20th century, Ludwig Wittgenstein, believed language could be reduced to its
‘atomic’ components. Once the underlying logical form of any sentence or
statement was analysed, they thought, its truth or falsehood could be
25
ascertained.219 This logical positivism was heralded as an end of philosophy,
because most philosophical problems were merely terminological
misunderstandings that could be positively analysed, clarified and resolved.
After WW1, Wittgenstein returned to teach maths to primary school children,
and later became the fiercest critic of his earlier theory. Language had no
underlying, logical structure because the meaning of a word is its use. 220 Words
never have a literal or ‘original’ meaning because meaning exists only through
its context.221 Anyone who has found their irony, double-meanings, or humour
fall flat over an email knows this to be true. Human intelligence is, in every
sense, a social phenomenon. ‘Artificial’ intelligence is literally that: fake
intelligence, not real. This is not a cause for pessimism, but instead should focus
attention on what really matters. This is that technology’s true value is to
accentuate and empower the uniqueness of the human mind.

4. How to reprogramme the law

While history shows that technology can improve jobs and people’s lives with
the right social policy, ‘technological displacement’ does exist. This leads to the
essential question: does the law do enough to guarantee full employment on fair
incomes in a free society (not mass unemployment on a basic income in a
billionaire’s dystopia)? Though legal systems differ, it cannot yet be said the law
lets everyone share fairly in technology’s gains. But as technology develops, we
can ‘reprogramme’ the law, and prepare for the future of work and leisure. We
can (1) guarantee full employment with economic democracy, (2) ensure
universal social security for fair incomes, and (3) ensure strict enterprise liability
for robots, like the products (not persons) they are.

4.1 Economic democracy

First, to make mass unemployment from robots impossible, government must

reprogramme the law for full employment: revitalising fiscal, monetary,
financial securities, working time, and job security policy. Full employment
means everyone has the hours they need ‘at fair wages’. 222 It does not mean
underemployment, minimal wages, or soaring inequality, because all undermine
aggregate demand and investment in jobs. Part 2(3) showed full employment’s
cost is low if business is confident government will do ‘whatever it takes’. But
corporations and asset owners must still pay their fair share of tax, to stop
capital being hoarded and withheld from the job market.223 Every central bank’s
objective must be crystal clear: full employment, with real wage growth, and
zero price inflation. Two per cent inflation, permanent wage attrition, and

26
soaring inequality is not ‘price stability’.224 Securities law must ensure full
disclosure and eliminate conflicts of interest. Good faith must become the
organising principle of contract law, to uphold the reasonable expectations of
honest people.225 Working time must be progressively reduced: next with a three
day weekend.226 Job security must mean a right to notice, a fair hearing,
redundancy pay, and rights of elected work councils to defer dismissals, to stop
the conflicts of interest in management decisions.227

But full employment needs more than mid-20 th century strategies, however
successful they were. Government did give up its commitment to balance private
and international investment volatility. Even when commitment is re-
established, that could happen again. The sources of volatility and inequality
must be uprooted: the unaccountable concentration of votes in the economy into
the hands of an irrational few. In 2015, Richard Freeman, one of the most
important economists in recent history, wrote that if we ‘own the robots’ we
could live in a more equal society, and the risk of mass unemployment would
go. His central proposal was to advance ‘employee ownership’ through trusts,
stock options and profit sharing schemes, apparently like John Lewis,
Mondragon, or Google.228 Freeman mentioned pension and mutual funds in
which people diversify their investments, but advocated share schemes where
firm employees do not. The problem is, all employee share schemes break the
cardinal rule of prudent investment: to diversify.229 Successful cooperatives like
John Lewis or Mondragon do not allow trading with shares: and so are
accurately described as employee partnerships, not ownership. What makes
them successful is not property, but the vote. In any case, the whole point of
automation is there may be very few employees left in a firm to own its shares:
this is true of Alphabet (which controls Google). But most of all, there is no
justification for concentrating employees’ risks into shares: widespread share
schemes would produce an ‘Enron economy’.230 This is why the great architect
of economic constitutionalism in the 20th century, A.A. Berle, moved from being
an advocate of share schemes,231 to seeking diversification by pensions.232 The
problem has always been, not merely to spread individual ownership, but to
democratise social power.

Democracy in the economy must be defended, as much as it is in politics. People

must have votes at work. People’s retirement savings, controlled by democratic
pension funds (not asset managers or banks), must become a progressively
larger share in global capital. The most successful, long-term corporations are
usually ones that embrace democratic voice. It is no accident that in UK
universities, staff and students have rights to vote. 233 Investment of capital in
enterprise justifies votes in business much of the time: this money is usually
27
employees’ retirement savings. This is why countries with large occupational
pension systems (not income linked public pensions) have the largest stock
markets. The ultimate investors should have a voice. Investment of labour in
enterprise always justifies the vote.234 In network enterprises or ‘natural
monopolies’, long-term investment of custom does too. There is no one size fits
all theory of enterprise,235 but a corporation can be seen as a commons where
different people stake different claims, to basic rights, standards of service, and
the all-important right to vote.236

The progressive democratisation of the economy will mean both a reduction in

the extremes of income inequality, and an end to investment volatility. With
votes at work,237 staff will no longer tolerate directors who keep paying
themselves more,238 as everyone else’s income declines. With votes on capital,
through democratic (and diversified) pension funds, it will no longer be possible
for asset managers and banks to use their conflicted positions to inflate their
salaries on ‘other people’s money’.239 With votes as a citizen, the long-term
customers of networked services will no longer be exploited for their private
data or held like hostages with no other train to catch. Most crucially, as the
votes in the economy come to reflect people’s real preferences (not what a tiny
handful of billionaires, board members and banks prefer) the patterns of
investment will reflect long-term sustainable objectives (not quarterly profits
and a golden parachute before a slump). People will want to see their savings
used to maintain full employment, social justice and a living planet. The private
sector’s short-term volatility will disappear. As countries around the world
develop their own models of economic democracy, trade will become regular,
stable and just.

4.2 Universal fair incomes

Second, universal social security must be the basis for fair incomes. ‘Social
security’, as enshrined in the Universal Declaration of Human Rights, is a
collective term that includes replacing income when people are not in
employment, especially by unemployment insurance, old age pensions,
disability and child care.240 It opposes the idea that mass unemployment is
acceptable. First, its purpose is to internalise unemployment’s social cost. It
impels government to create jobs, educate and retrain, not to allow
dispossession.241 Second, it mitigates workers’ inequality of bargaining power
against employing entities, to get fair (not basic) wages. Countries range from
income replacement models like in Germany’s Social Code, 242 to minimum floor
systems like the US Social Security Act of 1935, to hybrids like in Sweden.

28
A ‘basic income’ is such a powerful idea because it promises to universalise
social security, which underpinned full employment. It is also capable of being
hijacked: it means different things to different people, if tax plans are unspoken.
Most proposals are correct to eliminate bureaucracy and stigma, to get rid of
means testing. Social security is not charity we should beg for, but a right we
pay for. But no universal income plan must act, as the Speenhamland disaster
was shown to in part 3(1), to subsidise exploitative employers. It cannot be
necessary for universality to make transfer payments to people who already pay
income tax. To be taxed only to receive a sum back makes no sense. But more
than this, anything like Speenhamland that subsidises poverty-paying employers
diminishes bargaining power, and pay, of all workers. The minimum wage,
collective bargaining, and votes at work are fundamental for human autonomy.
Particularly in countries, like the US, where political democracy is under
sustained attack,243 handing power over incomes to a captured government is a
morbid threat to human development. ‘Basic income’ programmes, like in
Alaska, which sell commodities outside the system, not funded by internal
redistribution, differ fundamentally.244 Any basic income model that could
subsidise employers is an attack on universal human rights. So are proposals to
replace social housing, health, or care, with money. For this reason, the debate
has been shifting towards universal basic services.245

The credible forms of basic income proposal (indeed, the only credible
proposals) understand in detail behavioural psychology, tax and welfare law. 246
In fact, introducing a ‘personal allowance’ in social security taxes (so that
contributions are only made after a threshold) has the same economic effect as a
basic income transfer for an employed person, but without the subsidy effect for
employers.247 It also negates the bureaucracy. An essential principle, which
upholds the universal right to ‘just and favourable remuneration’,248 is that tax is
levied proportionately to the means to pay. For everyone not paying income tax
the legal right to income replacement, by auto-enrolment, would mean a
tremendous advance in social prosperity. Most importantly, it would help ensure
incomes are not just basic, but fair.

29
4.3 Robots as products

Third, as automation and artificial intelligence become more advanced, the

people or entities that benefit from robot ownership must be fully responsible for
them. This points to speculation in a European Parliament paper about creating
‘electronic personality’ for robots, and partly related to this, the idea of
Microsoft’s Bill Gates of ‘taxing the robots’ to pay for more human social
services. Both uphold the idea, which must be right, that robot enterprises bear
more responsibility. But the reasons why need to be clear.

First, enterprises must bear all costs of their production. This principle is
fundamental for a sound economy. If businesses can harm others without
paying, this ‘negative externality’ creates an artificial economic subsidy: it
distorts every market price. Tort victims, 249 and people with unequal bargaining
power (or ‘non-adjusting creditors’250) have no genuine ability to contract for
compensation. So the law’s role is to internalise those social costs. 251 But more
‘personalities’ in law complicates liability. In 1966, in Walkovszky v Carlton,
Carlton created ten corporations to separately own ten taxis.252 Each of these had
separate legal personality, and limited liability for Carlton. A taxi hit a
pedestrian, Walkovszky, who tried to hold Carlton personally liable, because the
corporation did not have enough money for his medical fees.253 Over a powerful
dissent by Keating J, the majority held that even though the taxi corporation was
undercapitalised, it was a separate legal person, and alone was liable for the
injuries. This was a bad judgment, because tort victims have no ability to
contract around limited liability as a secured creditor might.254 But it merely
illustrates the point that more electronic personalities (e.g. for a robot or
driverless car) could lead robot enterprises to argue they have no liability for
accidents. Unlike the risks, there are no benefits in creating electronic
personality.255 Fortunately, the EU proposals seem to have dropped electronic
personality, to strengthen the law holding enterprises as strictly liable for their
robot products.256

Second, in an interview with Quartz in 2017, Bill Gates stated that if a human
job, where income tax and social security had been paid, were suddenly replaced
by a robot, ‘you’d think that we’d tax the robot at a similar level.’257 Gates
argued revenue is needed to care for ‘the elderly, having smaller class sizes,
helping kids with special needs... where human empathy and understanding are
still very, very unique.’ If taxing ‘the robots’ is used as a metaphor, this is a
powerful argument. Social services, especially in old age and early childhood
care, are chronically underfunded or non-existent in even the richest countries.
Though personifying robots is tempting, this should be seen as a call for
30
corporations and the rich to pay a fair share of tax. If a future Bill or Melinda
Gates profits from labour displacing technology, they should pay more in
income and capital gains tax. A future Microsoft should pay more in corporate
tax, because their profits bear some relation to their work, more to luck, and
most to legal support.258 A basic principle of tax policy is to pay tax on wealth in
whatever form that wealth is held. So taxes should not be hypothecated to
specific computers or robots, instead of all asset owners paying their fair share.
If everyone contributes proportionately to their means, there will be a truly
‘sharing economy.’

5. Conclusion

The promises of technology are astounding, and deliver humankind the capacity
to live in a way that nobody could have once imagined. The industrial revolution
of the 19th century brought people past subsistence agriculture. It became
possible to live, not just from hand to mouth, bonded to lords and masters, but to
win freedom from servitude through solidarity. The corporate revolution of the
20th century enabled mass production and social distribution of wealth, for
human and democratic development across the globe. A third economic
revolution has often been pronounced or predicted, but it will not only be one of
technology. The next revolution will be social. It must be universal. Universal
prosperity with democracy and social justice, on a living planet, is achievable
not in centuries, but in years and decades. It did not begin with technology, but
with education. Once people can see and understand the institutions that shape
their lives, and vote in shaping them too, the robots will not automate your job
away. There will be full employment, fair incomes, and a thriving economy
democracy.

31
Notes

Oxfam, An Economy for the 1% (2016) suggested 62 people own more

assets than half the globe’s population, but studies vary.

2 M Zuckerberg, ‘Commencement address at Harvard’ (25 March 2017) at

the opening of his speech, Zuckerberg points out, perhaps self-deprecating,
perhaps patronising, that his audience had graduated while he did not.

3 D Galeon, ‘Elon Musk: Automation Will Force Governments to Introduce

Universal Basic Income’ (14 Feb 2017) Futurism

4 M Kubanzky and T Williams, ‘Why We Invested: GiveDirectly’ (7

February 2017) omidyar.com

5 V Wadhwa, ‘Sorry, but the jobless future isn’t a luddite fallacy’ (7 July
2015) Washington Post, adding ‘most people can’t adapt to such dramatic
change.’ Presumably the owner of Amazon, but maybe not the article’s author,
can ‘adapt’.

6 The literature rejecting the ‘jobless future’ narrative is vast: see for
example, DH Autor, ‘Why Are There Still So Many Jobs? The History and
Future of Workplace Automation’ (2015) 29(3) Journal of Economic
Perspectives 3, 23.

7 See part 4(3) below and M Friedman, ‘The Case for a Negative Income
Tax: A View from the Right’ in JH Bunzel, Issues in American Public Policy
(1968) ch 2. G Standing, The precariat: The new dangerous class (2016) 199-
212

8 CB Frey and MA Osborne, ‘The future of employment: How susceptible

are jobs to computerisation?’ (2017) 114 Technological Forecasting and Social
Change 254.

9 CB Frey, M Osborne, With Contributions from Citi Research,

‘Technology at Work: The Future of Innovation and Employment’ (February
2015) Citi GPS

32
10 K Grace et al, ‘When Will AI Exceed Human Performance? Evidence
from AI Experts’ (30 May 2017) Working Paper.

11 See F Fukyama, The End of History and the Last Man (1992). H
Hansmann and R Kraakman, ‘The End of History for Corporate Law’ (2000) 89
Georgetown Law Journal 439. Contrast D Kershaw, ‘No end in sight for the
history of corporate law: the case of employee participation in corporate
governance’ (2002) 2(1) JCLS 34 and E McGaughey, ‘Votes at work in Britain’
(2017) 46(4) ILJ, summarised on the LSE BPP blog.

12 See Back to the Future Part II (1989) starring MJ Fox and C Lloyd.

13 e.g. T More, Utopia (1516) Book I. H Sinzheimer, Grundzüge des

Arbeitsrechts o r Foundations of Labour Law (1920) ch 2, 22-27 (in German).
AA Berle and GC Means, The Modern Corporation and Private Property
(1932) Part III.

14 G Orwell, Animal Farm (1945) ch 5

15 n.b. ‘App’ can be short for both an ‘application’ and in the US, according
to the Oxford English Dictionary, an ‘appetizer’.

16 W Beveridge, Full Employment in a Free Society (1944) see below at part

4(1)

17 n.b. ‘Cognition’ means the ‘process of acquiring knowledge and

understanding through thought, experience, and the senses.’

18 e.g. ‘Octuplets’ Mom On Welfare, Spokesman Confirms’ (10 February

2009) Fox News. A Crick and K Cain, ‘Benefit Mum Rap: ‘Queen of Benefits’’
(21 February 2017) The Sun. S Markson, ‘Taxpayer works for 14 years to pay
welfare bill of just ONE dole bludger’ (22 June 2017) Daily Telegraph. Each
‘news’ outlet is controlled by Victorian mogul, Rupert Murdoch.

19 cf Alan Budd, Chief Economic Advisor to HM Treasury under Margaret

33
Thatcher: ‘The nightmare I sometimes have... there may have been people
making the actual policy decisions or people behind them or people behind them
who never believed for a moment that this was the correct way to bring down
inflation. They did however see that it would be a very, very good way to raise
unemployment, and raising unemployment was an extremely desirable way of,
of reducing the strength of the working classes, if you like. That what was
engineered there – in Marxist terms – was a crisis of capitalism which re-created
reserve army of labour, and has allowed the capitalist to make high profits ever
since. Now again, I’d not say I believe that story, but when I really worry about
all this I worry whether indeed that was really what was going on.’ Pandora’s
Box, Episode 3: The League of Gentlemen (Thursday, 18 June 1992) BBC2 at
42:55. Note, the reserve army of labour is not originally a Marxist concept, but
from Sir or Saint Thomas More, the Lord Chancellor executed by Henry VIII in
1530. See part 3(1) below.

20 UDHR 1948 arts 22-23, ‘Everyone, as a member of society, has the right
to social security’. ‘Everyone has the right to work, to free choice of
employment’ and ‘the right to just and favourable remuneration ensuring for
himself and his family an existence worthy of human dignity’.

21 M Friedman, ‘The Role of Monetary Policy’ (1968) 58(1) AER 1, 10.

22 Friedman (1968) 58(1) AER 1, 9. Note, inflation typically diminishes real

wages, but enriches owners and sellers of assets.

23 Friedman (1968) 58(1) AER 1, 11

24 Friedman (1968) 58(1) AER 1, 9

25 AW Phillips, ‘The Relation between Unemployment and the Rate of

Change of Money Wage Rates in the United Kingdom 1861–1957’ (1958) 25
Economica 283, referring to unemployment and the ‘change of money wage
rates’.

26 Unequal bargaining power is consequent on three main factors: (1)

inequality in resources or wealth: A Smith, The Wealth of Nations (1776) Book
I, ch 8, §12, (2) collective organisation: JS Mill, Principles of Political Economy
(1848) Book V, ch XI, §12, and (3) asymmetry of information: WS Jevons,
Theory of Political Economy (3rd edn 1888) ch 4, §74.
27 See also M Friedman, ‘The Methodology of Positive Economics’ in
Essays in Positive Economics (1953) ch 1, 4-7
34
28 n.b. D Hume, Essays Moral, Political, Literary (1777) Part II, Essay III,
Of Money, describes an inflation process after gold returns from Cadiz, which he
suggested temporarily increases productivity. This story has doubtful application
to anything today.

29 FA Hayek, ‘Full Employment, Planning and Inflation’ (1950) 4(6)

Institute of Public Affairs Review 174

30 FA Hayek, The Road to Serfdom (1944) described in the 1950 tract as ‘the
first comprehensive top-level intellectual assault on the socialist position and
perhaps created more stir than any politico-economic publication of the day.’

31 Hayek (1950) 4(6) IPAR 174, 181

32 See BP, Statistical Review of World Energy (June 2017) 20, roughly
tripling of oil prices in 1973-74, and again by 1979-80.

33 The only better 16 years in recent UK history was 1991-2006, at 2.76%.

For 1968-1972, around the Industrial Relations Act 1971, inflation ranged from
6.4-9.4%. P Davies and M Freedland, Labour Legislation and Public Policy
(1992) chs 6 and 7, 365.

34 Friedman won the ‘Bank of Sweden Prize in Memorial of Alfred Nobel’.

Though administered now by the Nobel Committee, Alfred Nobel never
consented to an economics prize or bank funding. Nobel’s name therefore seems
to be misrepresented.

35 M Friedman, ‘Inflation and Unemployment’ (1977) 85 Journal of Political

Economy 451, 458

36 Friedman (1977) 85 JPE 451, 458

37 They do not: inflation was falling from 1970-1973 in 5 out of 7 countries

in Friedman’s own charts (1977) 85 JPE 451, 462-3

38 Friedman (1977) 85 JPE 451, 464-5

39 See shortly after, JA Tatom, ‘What ever happened to the Phillips Curve?’
(1981) Federal Reserve Bank of St Louis Working Paper 1981-008A, 3,
‘experience in the past decade makes the NAIRU corollary highly doubtful.’
35
40 DT Coe, ‘Nominal Wages. The NAIRU and Wage Flexibility’ (1985) 5
OECD Economic Studies 87, 88

41 SP Hargreaves Heap, ‘Choosing the Wrong ‘Natural’ Rate: Accelerating

Inflation or Decelerating Employment and Growth?’ (1980) 90(359) Economic
Journal 611. Or, if government keeps unemployment below the ‘natural’ rate,
the tendency to cause inflation would reduce with less long-term unemployment,
meaning a lower (and eventually zero) natural rate over time.

42 See DJ Burningham ‘Economics for Technical Students’ (1960) 2(1)

Education + Training 11, 12-13 (saying many economic lags are analogous to
hysteresis in magnetics). ES Phelps, Inflation Policy and Unemployment
Theory: The Cost-Benefit Approach to Monetary Planning (1972) xxiii
(applying the analogy, but sadly accepting a ‘natural’ rate) and Hargreaves Heap
(1980) 90(359) Economic Journal 611 (probably the first author to use the
analogy without accepting the idea of a natural rate).

43 K Clark and L Summers, ‘Labour Force Participation: Timing and

Persistence’ (1982) 49(5) Review of Economic Studies 825. See also the better
known OJ Blanchard and LH Summers, ‘Hysteresis and the European
Unemployment Problem’ (1986) 1 NBER Macroeconomics Annual 16. Further,
‘Hysterisis in Unemployment’ (1987) 31 European Economic Review 288.
‘Beyond the Natural Rate Hypothesis’ (1988) 78(2) American Economic
Review 182, 186, ‘the longer unemployment stays away from equilibrium, the
more slowly it will eventually-if ever-return to it.’ But this should not be viewed
as a ‘natural’ law.

36
44 e.g. RM Solow, ‘On Theories of Unemployment’ (1980) 70(1) AER 1,
‘Ever since Adam Smith, economists have been distinguished from lesser
mortals by their understanding of... decentralized competitive markets... The
profession’s reservations about rent control, interest rate ceilings, and minimum
wage laws [...reflect...] a feeling that nonprofessionals simply do not understand
fully the consequences, often unexpected and undesired, of messing around with
the market mechanism.’ Note that (1) Adam Smith was a lawyer and moral
philosopher, (2) Smith argued rents were exploitative, (3) Smith approved caps
on extortionate interest for loans, and (4) he acknowledged the basis for the
minimum wage, namely unequal bargaining power: see The Wealth of Nations
(1776) Bk I, ch 11 (rents), Bk II, ch 4, §14 (interest caps), Bk I, ch 8, §12
(bargaining power). But why would ‘higher mortals’ like Solow actually bother
to read this carefully? Solow’s investigation into ‘history’ in the same article
meant going ‘back to reread’ Pigou, from 1933, a full 47 years old.

45 R Layard et al, ‘Special Issue on Unemployment’ (1982) 49(5) RES 675

46 D Grubb, R Jackman and R Layard, ‘Causes of the Current Stagflation’

(1982) 49(5) RES 707

47 M Andrews and S Nickell, ‘Unemployment in the United Kingdom Since

the War’ (1982) 49(5) RES 731, 751

48 Andrews and Nickell (1982) 49(5) RES 731, 741. Further, on a theory of
shocks, see DT Mortensen and CA Pissarides, ‘Job Creation and Job Destruction
in the Theory of Unemployment’ (1994) 61(3) Review of Economic Studies
397.

49 S Bentolila and G Bertola, ‘Firing Costs and Labour Demand: How Bad Is
Eurosclerosis?’ (1990) 57 Review of Economic Studies 381, 395-6 and 400,
with an Appendix showing five categories of laws, where notice, dismissal and
redundancy compensation are averaged, and remarkably they state ‘We do not
have data on the actual rise of [firing costs] in the seventies; we set the increase
to 33% when solving the model.’

37
50 Bentolila and Bertola (1990) 57 Review of Economic Studies 381, 394,
while the authors talk about ‘costs’ to an employer if an unfairly dismissed
employee goes to or succeeds at a Tribunal, they make no mention of the
enterprise costs of unfair dismissal: the damage where managers fire good
workers irrationally, where exuberant redundancies damage long-term good
will, or downward spirals in depression. This makes the authors’ conclusions on
‘costs’ very difficult to defend.

51 See the ambitiously entitled two part OECD, Jobs Study, Evidence and
Explanations (1994). See also the brilliant dissection in D Ashiagbor, The
European Employment Strategy: Labour Market Regulation and New
Governance (2005) ch 2, 38-48

52 P Krugman, ‘Past and prospective causes of high unemployment’ (1994)

79(4) Economic Review - Federal Reserve Bank of Kansas City 23, 50, 57 and
74

53 Krugman (1994) 79(4) Economic Review 23, 74. Also at 76, ‘the apparent
unraveling of that model makes it difficult to argue for implementation of
Swedish-style labor market policies.’ It is not entirely apparent what Krugman
means.
54 UDHR 1948 arts 22-23, ‘the right to social security’ and the ‘right to
work’. The ‘30 years’ starts at Friedman’s 1967 address.

55 Although state law can improve, a typical common law position is that
irrational employer dismissals are not reviewed, unless the Civil Rights Act of
1964 is violated, or a similar statute, or there is a collective dismissal which
requires a mere 90 days’ notice under the WARN Act of 1988. This has long, in
effect, been the ‘Donald Trump model of workplace relations’.

56 Federal Reserve Act of 1913, 12 USC §225a.

57 Bank of England Act 1998 s 11. Treaty on the Functioning of the

European Union 2007 art 127, ex TEC 1996 art 105. M Roth, ‘Employment as a
Goal of Monetary Policy of the European Central Bank’ (2015) ssrn.com makes
a convincing argument that the ECB is bound to prioritise full employment,
given TEU art 3.

38
58 JL Yellen, ‘Monetary policy: goals and strategy’ (13 March 1996) NABE
Policy Seminar, Washington, DC, a euphemism which appears to refer to labour
and social rights, as one supposes Yellen is implying they adversely affect ‘the
efficiency of the labor market in matching workers to job vacancies, the
geographic mobility... quality of skills... [and] demographics of the labor force,
and the extent of structural mismatch between job vacancies and unemployed
workers.’

59 D Staiger, JH Stock and MW Watson, ‘The NAIRU, Unemployment and

Monetary Policy’ (1997) 11(1) Journal of Economic Perspectives 33, the CBO
estimated 5.8% and Council of Economic Advisers 5.7%. On the actual rate, see
below.

60 S Nickell, ‘Unemployment and Labor Market Rigidities: Europe versus

North America’ (1997) 11 Journal of Economic Perspectives 55, 66. Nickell,
drawing upon flawed OECD protection indexes also said ‘unions are bad for
jobs, but these bad effects can be nullified’ by coordinated bargaining. Benefits
are ‘bad’, unless coupled with activation policies. Minimum wages are ‘bad’ for
young people. None of these conclusions have a credible basis, as they derive
from flawed OECD data.

61 R Layard, ‘Getting people back to work’ (1998) Autumn, CentrePiece 24,

27

62 e.g. O Blanchard and J Wolfers, ‘The Role of Shocks and Institutions in

the Rise of European Unemployment: The Aggregate Evidence’ (2000) 110
Economic Journal 1. DH Autor, JJ Donohue III and SJ Schwab, ‘The Costs of
Wrongful-Discharge Laws’ (2006) 88(2) RES 211. LH Summers,
‘Unemployment’ in The Concise Encyclopedia of Economics (2nd edn 2007)
econlib.org, ‘Another cause of long-term unemployment is unionization....
Between 1970 and 1985, for example, a state with a 20 percent unionization
rate, approximately the average for the fifty states and the District of Columbia,
experienced an unemployment rate that was 1.2 percentage points higher than
that of a hypothetical state that had no unions.’ As explained above, based on the
CBR data the evidence suggests this idea is flatly mistaken, but is also
outweighed by 1.3% of the US workers in prison.

39
63 Z Adams, L Bishop, S Deakin, C Fenwick, S Martinsson and G Rusconi,
‘The Economic Effects of Laws Relating to Employment Protection and
Different Forms of Employment: Analysis of a Panel of 117 Countries, 1990-
2013’ (2018) CBR Working Paper, forthcoming. This is based on the CBR,
Labour Regulation Index (2016)

64 VV Acharya, RP Baghai and KV Subramanian, ‘Labor Laws and

Innovation’ (2013) 56(4) Journal of Law and Economics 997

65 But contrast L Ball and NG Mankiw, ‘The NAIRU in Theory and

Practice’ (2002) 16(4) Journal of Economic Perspectives 115 and JK Galbraith,
‘Time to Ditch the NAIRU’ (1997) 11(1) JEP 93

66 JP Morgan, Paychecks, Paydays, and the Online Platform Economy (2016)

55% of 6m customers had 30% volatility in monthly wages

67 See Ashiagbor (2005) ch 2, 65-66. The answer to R Shimer, ‘Why Is the

U.S. Unemployment Rate So Much Lower?’ (1998) 13 NBER Macroeconomics
Annual 11, is ‘it is not’. Unemployment is not reduced by putting potential
workers in jail.

68 JM Keynes, The General Theory of Employment, Interest and Money

(1936) ch 12 and The Economic Possibilities of our Grandchildren (1930).
Keynes became textbook in a post-WW2 consensus. cf M Friedman (4 February
1966) Time, ‘In one sense, we are all Keynesians now; in another, nobody is any
longer a Keynesian.’ cf B Bernanke, ‘On Milton Friedman’s Ninetieth Birthday’
(8 November 2002) ‘You’re right, we did it. We’re very sorry. But thanks to
you, we won’t do it again.’ This proved unwise.
69 e.g. C Pigou, The Theory of Unemployment (1933) or JS Mill, Principles
of Political Economy (1848) Book II, ch XII, §1. The latter emerged on
Massachusetts Supreme Court in Plant v Woods, 57 NE 1011 (1900) per Holmes
J, ‘Organization and strikes may get a larger share for the members of an
organization, but, if they do, they get it at the expense of the less organized and
less powerful portion of the laboring mass. They do not create something out of
nothing.’ This is false. Holmes J was complicit in Loewe v Lawlor, 208 US 274
(1908) imposing antitrust damages on unions, reversed by the Clayton Act of
1914 §§6 and 20.

70 Keynes (1936) chs 2 and 3, arguing workers cannot bargain for higher
real wages because this also depended on price levels.
71 Keynes (1936) ch 9
40
72 Keynes (1936) ch 12, V, 149

73 Keynes (1936) ch 12, III, 145

74 Keynes (1936) ch 22, quietly adding ‘social practices and... a distribution

of wealth... result in a propensity to consume which is unduly low’. This
discloses a concern (if somewhat suppressed) that unjustified inequality harms
economic development.

75 Keynes (1936) ch 15

76 cf Keynes (1936) ch 12, VI where Keynes did highlight the ‘high

brokerage charges and the heavy transfer tax’ to access London Stock Exchange
trading, which ensured fewer ‘sins’ than at the New York Stock Exchange. This
is reductive at best.

77 e.g. N Kaldor, ‘A Model of the Trade Cycle’ (1940) 50(197) Economic

Journal 78, and ‘What is Wrong With Economic Theory’ (1975) 89(3) Quarterly
Journal of Economics 347. This is footed on variations on the notion that
‘economic activity always tends towards a level where Savings and Investment
are equal’. This is wrong because law creates every ‘tendency’.

78 See R Dale, The First Crash: Lessons from the South Sea Bubble (2004)
ch 6

79 Keech v Sandford (1726) 25 ER 223, Lord King LC holding that even a

trustee of a marketplace had to avoid any possibility of a conflict of interest. See
J Getzler, ‘Rumford Market and the Genesis of Fiduciary Obligations’ in A
Burrows and Lord Rodger of Earlsferry (eds) Mapping the Law: Essays in
Honour of Peter Birks (OUP 2006) 586.

80 Carter v Boehm (1766) 3 Burr 1905, per Lord Mansfield. For background,
see S Watterson, ‘Carter v Boehm (1766)’ in C Mitchell and P Mitchell,
Landmark Cases in the Law of Contract (Hart 2008) ch 3.

81 Affirmed in Pawson v Watson (1778) 2 Cowp 786, but see Manifest

Shipping Co Ltd v Uni-Polaris Shipping Co Ltd [2001] UKHL 1.
82 Erlanger v New Sombrero Phosphate Co (1878) 3 App Cas 1218, per
Lord Blackburn held that the sale of shares in a phosphate mining company,
41
when the company owners did not disclosure the price they paid (exactly half)
for the land, meant the share purchasers could rescind the contract and get their
money back.

83 See Smith v Hughes (1871) LR 6 QB 597, per Blackburn J on the sale of

(green not old) oats, for horse food.

84 One of the most significant, under-appreciated causes was another of Lord

Mansfield’s decisions: see AW Blumrosen, ‘The Profound Influence in America
of Lord Mansfield’s Decision in Somerset v Stuart’ (2007) 13 Texas Wesleyan
Law Review 645.

85 Berle and Means (1932) Part III

86 See Lochner v New York 198 US 45 (1905) applied in State Board of

Control v Buckstegge, 158 Pac 837, 842 (1916), Busser v Snyder, 282 Pa 440
(1925) and Railroad Retirement Board v Alton Railroad Co 295 US 330 (1935).

87 See GC Means, ‘The separation of ownership and control in American

industry’ (1931) 46(1) QJE 68

88 Explained by AA Berle and GC Means, The Modern Corporation and

Private Property (1932) Part III

89 cf M Friedman and AJ Schwarz, A Monetary History of the United States,

1867–1960 (1963) arguing the Great Depression’s cause was that government
allowed the monetary supply to shrink, and should have kept all banks going as
the lender of last resort.

90 e.g. Financial Services and Markets Act 2000 (c 8) or Markets in

Financial Instruments Directive 2014/65/EC

91 See Steward Machine Company v Davis 301 US 548 (1937) following

shortly after West Coast Hotel Co v Parrish 300 US 379 (1937)

92 Social Security Act of 1935, 42 USC §402

93 UDHR 1948 art 25, ‘Everyone has the right to... housing... and the right to
security in the event of... old age...’’

42
94 cf the notorious Opinion of Robin Potts QC to the International Swaps
and Derivatives Association, Inc (24 June 1997)

95 e.g. S Labaton, ‘Bush S.E.C. Pick Is Seen as Friend to Corporations’ (3

June 2005) NY Times

96 See JC Coffee, ‘What Went Wrong? An Initial Inquiry into the Causes of
the 2008 Financial Crisis’ (2009) 9(1) Journal of Corporate Law Studies 1. The
third major cause was no effective consumer regulation in the US. See E
Warren, ‘Product Safety Regulation as a Model for Financial Services
Regulation’ (2008) 43(2) Journal of Consumer Affairs 452, and contrast the
Consumer Credit Act 1974 or the Unfair Terms in Consumer Contracts
Directive 93/13/EEC arts 3-6.
97 JM Keynes, The Economic Possibilities of our Grandchildren (1930)

98 See also E Skidelsky and R Skidelsky, How Much is Enough?: Money

and the Good Life (2013) Preface and ch 1

99 J Schumpeter, Capitalism, Socialism and Democracy (1942) chs VII and

VIII

100 Schumpeter (1942) ch VIII, 90

101 G Orwell, Animal Farm (1945) ch 5

102 J Rifkin, The End of Work: The Decline of the Global Labor Force and
the Dawn of the Post-Market Era (1995)

103 Rifkin (1995) ch 5, 71-73, charting mechanical harvester use going from
6% to 78% in 15 years.

43
104 Contrast King’s own views, seeing absence of labour rights as the central
problem: ML King, Speech to the Fourth Constitutional Convention AFL-CIO,
Miami, Florida (11 December 1961) ‘the Wagner Act... tended merely to
declare rights but did not deliver them. Labor had to bring the law to life... we
shall bring into full realization the American dream—a dream yet unfilled. A
dream of equality of opportunity, of privilege and property widely distributed; a
dream of a land where men will not take necessities from the many to give
luxuries to the few; a dream of a land where men will not argue that the color of
a man’s skin determines the content of his character...’

105 Rifkin (1995) ch 10, 149

106 Rifkin (1995) ch 7, 257-265. See further B Dorini, Review (1996) 9(1)
Harvard Journal of Law & Technology 231.

107 E Brynjolfsson and A McAffee, Race Against the Machine: How the
Digital Revolution is Accelerating Innovation, Driving Productivity, and
Irreversibly Transforming Employment and the Economy (2011) ch 1, on
unemployment. This erroneously suggests Keynes believed that technology
produces unemployment. Ch 3, on median income decline, says nothing about
labour law.

108 Brynjolfsson and McAffee (2011) ch 5. They also call Friedrich von
Hayek ‘the greatest theorist of markets’. This is debatable.

109 E Brynjolfsson and A McAfee, The Second Machine Age: Work,

Progress, and Prosperity in a Time of Brilliant Technologies (2014) ch 14.

110 See campaignmoney.org, on Brynjolfsson contributing to ActBlue and

Bernie 2016 and McAfee to the Hillary Victory Fund.

111 See CL Estlund, ‘The ossification of American labor law’ (2002) 102(6)
Columbia Law Review 1527

112 See E McGaughey, ‘Fascism-lite in America (or the social ideal of

Donald Trump)’ (2018) BJALS (forthcoming)

44
113 See E McGaughey, ‘Votes at work in Britain: shareholder monopolisation
and the ‘single channel’’ (2017) 46(4) Industrial Law Journal, summarised on
the LSE BPP blog. Germany or the Netherlands, with strong codetermination
rights, differ.

114 The Worker Adjustment and Retraining Notification Act of 1988 (29 USC
§§ 2101-9) requires just 60 days notice for collective redundancies of more than
50 employees, or over 33% of the workforce, whichever is lower.

115 Employment Rights Act 1996 ss 86 (notice), 94 (right to a fair dismissal)

and 135 (redundancy pay).

116 Work Constitution Act 1972 (Betriebsverfassungsgesetz) §§111-113.

Dismissal Protection Act 1965 (Kündigungschutzgesetz) §1.

117 e.g. ‘Buying Itself Out of a Recession’ (4 November 2009) Spiegel

118 On employers’ potential conflicts of interest see Braganza v BP Shipping

Ltd [2015] UKSC 17, [18] per Lady Hale

119 Z Adams, L Bishop, S Deakin, C Fenwick, S Martinsson and G Rusconi,

‘The Economic Effects of Laws Relating to Employment Protection and
Different Forms of Employment: Analysis of a Panel of 117 Countries, 1990-
2013’ (2018) CBR Working Paper, forthcoming. This is based on the CBR,
Labour Regulation Index (2016)

120 CB Frey and MA Osborne, ‘The future of employment: How susceptible

are jobs to computerisation?’ (2017) 114 Technological Forecasting and Social
Change 254.

121 Frey and Osborne (2017) 114 TFSC 254, 257

122 Frey and Osborne (2017) 114 TFSC 254, 268

123 Frey and Osborne (2017) 114 TFSC 254, 263

124 CB Frey, M Osborne, With Contributions from Citi Research,

‘Technology at Work: The Future of Innovation and Employment’ (February
2015) Citi GPS
125 McKinsey Global Institute, A Future That Works: Automation,
45
Employment, and Productivity (2017). This ‘predicts’ 46 per cent of time spent
working ‘is technically automatable’. The authors may include partners in the
Preface who ‘led’ the ‘research’.

126 PwC, UK Economic Outlook March 2017 (2017) ch 4, ‘based on our own
preferred methodology’ (which is not disclosed) the authors (also not disclosed)
say ‘around 30% of jobs in the UK are at potential high risk of automation’.

127 Deloitte, Citizens, government and business: The State of the State 2017-
18 (2017) 10, with a ‘forecast’ that ‘35 per cent of jobs in the UK are highly
susceptible to automation’. The authors are unknown, but a ‘leader’ and
‘director’ at Deloitte feature themselves on page 1, and may or may not have
made a contribution.
128 K Grace et al, ‘When Will AI Exceed Human Performance? Evidence
from AI Experts’ (30 May 2017) Working Paper.

129 e.g. DH Autor, ‘Why Are There Still So Many Jobs? The History and
Future of Workplace Automation’ (2015) 29(3) Journal of Economic
Perspectives 3

130 Executive Office of the President, Artificial Intelligence, Automation and

the Economy (December 2016) 2 and 13-19.

131 A Smith, The Wealth of Nations (1776) Book II, ch 1.

132 See AA Berle and GC Means, The Modern Corporation and Private
Property (1932) Book I, ch 1. AA Berle, ‘Property, Production and Revolution’
(1965) 65 Columbia Law Review 1. O Kahn-Freund, ‘Hugo Sinzheimer’ in
Labour Law and Politics in the Weimar Republic (1981) 103. H Sinzheimer,
Grundzüge des Arbeitsrechts (1920) ch 2, 22-27 (in English, Foundations of
Labour Law).

133 MS Eccles, Beckoning Frontiers: Public and Personal Recollections

(1951) 75-76

134 Eccles (1951) 77

46
135 See the German Constitution 1949 (Grundgesetz 1949) §14(2) ‘Eigentum
verpflichtet.’

136 White Paper, Employment Policy (May 1944) Cmd 6527, [43]-[45] and
[57]-[63]

137 Disabled Persons (Employment) Act 1944 s 9, a duty of ‘substantial’

employers to hire disabled people up to a quota.

138 Distribution of Industry Act 1945

139 Distribution of Industry (1948) Cmd 7540, 26.

140 Hansard HC Debs (21 March 1945) vol 409, col 859. John Wardlaw-
Milne MP was never very popular in his party.

141 S Webb, How the Government Can Prevent Unemployment (1912) 6-9.
See also A Briggs, Social Thought and Social Action: A Study of the Work of
Seebohm Rowntree 1871– 1954 (1961)

142 Webb (1912) 12. See part 3(1) below.

143 M Kalecki, ‘Political aspects of full employment’ (1943) 14(4) Political

Quarterly 322

144 C Shapiro and JE Stiglitz, ‘Equilibrium Unemployment as a Worker

Discipline Device’ (1984) 74(3) American Economic Review 433 formulated a
similar idea: ‘equilibrium unemployment’ at a rate ‘sufficiently large that it pays
workers to work rather than to take the risk of being caught shirking’ can occur
because employers want to maintain fear. An employee who is dismissed, but
can find a job without any cost, ‘pays no penalty for his misdemeanor’. So
rational firms might restrict the number of people they hire to ensure a penalty
does result. At 435-440, the ‘market failure’ is ‘information asymmetries’
(employers do not know when workers will ‘shirk’ or not). With six pages of
linear equations this ‘can explain’ unemployment.

145 This is not to suggest there were no major problems: the worst aspect was
a sexist assumption of men being entitled to jobs.

47
146 AA Berle, ‘A New look at management responsibility’ (1962) 2 Human
Resource Management 3

147 RCO Matthews, ‘Why has Britain had Full Employment since the War?’
(1968) 78(311) Economic Journal 555

148 See also B Hansen, WW Snyder and OECD, ‘Fiscal policy in seven
countries 1955-1965: Belgium, France, Germany, Italy, Sweden, United
Kingdom, United States’ (1969) 0501/QL 200 H249

149 Kalecki (1943) 14(4) Political Quarterly 322, 330.

150 T Opler, L Pinkowitz, R Stulz and R Williamson, ‘The determinants and

implications of corporate cash holdings’ (1999) 52 Journal of Financial
Economics 3, 18 (showing decline in corporate hoarding from the 1950s till the
1990s).

151 X Gao, ‘Corporate Cash Hoarding: The Role of Just-in-Time Adoption’

(2018) Management Science (forthcoming), 6

152 P Seitz, ‘Apple Has Largest Cash Stockpile, Twice That Of Microsoft’s’
(20 May 2016) Investor’s Business Daily

153 In November 2017, the number of unemployed was 6.6m. A $10 an hour
job, for 38 hours a week, for 50 weeks in a year means $22,914 in pre-tax
($19,090.90 post-tax) annual income. That means 26.4m jobs could be created
with $504bn. This calculation vastly overestimates the expense of full
employment: the multiplier effect of such investment would be colossal.

154 cf AA Berle and GC Means, The Modern Corporation and Private

Property (1932) Book I, ch 1.

155 T More, Utopia (1516, translated by P Turner, 2004) Book I, 25. More’s
own views were probably those of Raphael Nonsenso.

156 More (1516) Book I, 27

157 The Vagabond Act 1495 (11 Hen VII c 2) said vagabonds must be ‘sette
48
in stokkes, ther to remayne by the space of 3 daies and 3 nyghtes and ther to
have noon other sustenaunce but brede and water’. R Pashley, Pauperism and
Poor Laws (2011) ch 5.

158 See further EH Shagan, ‘Protector Somerset and the 1549 Rebellions:
New Sources and New Perspectives’ (1999) 114(455) English Historical Review
34. D MacCulloch, ‘Kett’s Rebellion in Context’ (1979) 84 Past & Present 36.

159 See PJ Federico, ‘Origin and Early History of Patents’ (1929) 11 Journal
of the Patent Office Society 292, 298 and M Grass and A Grass, Stockings for a
Queen: the life of the Rev. William Lee, the Elizabethan inventor (1967) 122.

160 SD Chapman, ‘Enterprise and Innovation in the British Hosiery Industry,

1750–1850’ (1974) 5(1) Textile History 14, 24.

161 See S Webb and B Webb, The History of Trade Unionism (1920) ch II, 88
ff

162 Lord Byron, Hansard HL Debs (27 February 1812) vol 21, col 966. See
also C Brontë, Shirley (1849)

163 Hansard HL Debs (27 February 1812) vol 21, col 969, ‘the sword is the
worst argument... so should it be the last: in this instance it has been the first,
but... as yet, only in the scabbard. The present measure will... pluck it from the
sheath...’

164 See G Clark, ‘The long march of history: farm wages, population and
economic growth, England 1209 – 1869’ (2007) 60(1) Economic History
Review 97. NFR Crafts and TC Mills, ‘Trends in Real Wages in Britain, 1750-
1913’ (1994) 31(2) Explorations in Economic History 176. Wage declines did
recur after TUA 1871, but were less prolonged and severe, until 2010. See TUC,
‘UK workers suffering the most severe squeeze in real earnings since Victorian
times’ (12 October 2014).

165 See S Deakin and F Wilkinson, The law of the labour market:
Industrialization, employment and legal evolution (2005) 63-67

166 See R v Journeymen-Taylors of Cambridge (1721) 88 ER 9.

167 A Smith, The Wealth of Nations (1776) Book I, ch 8, §12

49
168 K Polayni, The Great Transformation (1944) chs 7-9, especially 82

169 S Deakin and F Wilkinson, The law of the labour market:

Industrialization, employment and legal evolution (2005) ch 3, 125-7

170 cf TR Malthus, An Essay on the Principle of Population (1798)

171 cf R Bregman, Utopia for Realists: How We Can Build the Ideal World
(2017) arguing without evidence that technology caused wage decline, and that
Speenhamland was a success because the Royal Commission Reports
misrepresented its operation. This does not follow. Just because the Royal
Commission was a farce, it does not mean Speenhamland was not a disaster.

172 The Great Reform Act 1832 extended the vote in Parliament, yet still to a
tiny percentage of the male population.

173 cf R Bregman, ‘No, wealth isn’t created at the top. It is merely devoured
there’ (30 March 2017) Guardian

174 E Hobsbawm and G Rudé, Captain Swing (1969) 50-51

175 R v Lovelass (1834) 172 ER 1380

176 S Webb and B Webb, The History of Trade Unionism (1920) ch III, 144-
153

177 H Feis, ‘The Industrial Situation in Great Britain: From the Armistice to
the Beginning of 1921’ (1921) 11(2) American Economic Review 265-6,
predicted exactly this would happen.

178 T Wilson, ‘Notes on Labour Unrest’ (1919) 29(115) Economic Journal

367, 372

179 S Ward, Unemployment and the State in Britain: The Means Test and
Protest in 1930s (2013) ch 1, 34-36, and The Times (2 January 1919) page 3 on
the Rhondda Valley strike. On interwar legislation, see Deakin and Wilkinson
(2005) ch 3, 160-3

180 OT Mallery, ‘Preventing periods of unemployment by expanding public

works’ (1921) American Labor Legislation Review 47

50
181 See N Ferguson, ‘Germany after the First World War by Richard Bessel’
(1995) 110(439) English Historical Review 1224

182 R Bessel, Germany After the First World War (1993) ch 5.

183 The Emergency Tariff of 1921 and the Fordney–McCumber Tariff of

1922 probably improved US employment, but significantly less than it cost
everyone else.

184 JM Keynes, The Economic Consequences of the Peace (1919)

185 AR Seipp, The Ordeal of Peace: Demobilization and the Urban

Experience in Britain and Germany, 1917-1921 (2nd edn 2016) ch 7

186 E McGaughey, ‘The codetermination bargains: the history of German

corporate and labor law’ (2016) 23(1) Columbia Journal of European Law 135,
163-4, citing K Robert, Hitler’s Counterfeit Reich (1941) 121 and F Neumann,
‘Labor Mobilization in the National Socialist New Order’ (1942) 9 Law &
Contemporary Problems 544, 546.

187 G Miller, ‘The Economic Policy of British Labor’ (1946) 6(2) Antioch
Review 179, 183

188 White Paper, Employment Policy (1944) Cmd 6527, 7, §11

189 (1944) Cmd 6527, 7-8, §14. Also the Reinstatement in Civil Employment
Act 1944 placed a duty on a last employer before the war began to reinstate a
soldier.
190 Servicemen’s Readjustment Act of 1944. See also the Veterans’
Preference Act of 1944, updating the government’s commitment to prefer
veterans in its own hiring policies.

191 Employment Act of 1946, 15 USC §1021. See KVW Stone, ‘A Right to
Work in the United States’ in V Mantouvalou (ed) The Right to Work (2015) ch
15. A commitment to ‘continuing full employment’ was scrapped by the
opposition of people like Robert Taft, who soon succeeded in the passing the
disastrous Taft-Hartley Act of 1947.
192 G Marshall, The Marshall Plan Speech (5 June 1947) Harvard

193 UDHR 1948 art 23(1) ‘Everyone has the right to work, to free choice of
employment, to just and favourable conditions of work and to protection against
51
unemployment.’

194 T Wheelwright, Oil and World Politics (1991) ch 1

195 HB Brown, ‘The Status of the Automobile’ (1908) 17(4) Yale Law
Journal 223, 227

196 Brown (1908) 17(4) Yale Law Journal 223, 225-227

197 ER Kilby, ‘The Demographics of the U.S. Equine Population’ in The

State of America’s Animals IV (2007) ch 10, 176, drawing data from EM
Ensminger, Horses and Horsemanship (1969)

198 E Albin, ‘Labour Law in a Service World’ (2010) 73(6) MLR 959

199 World Health Organization, Global Status Report on Road Safety 2015
(2015) Table A2

200 J Langan-Fox et al (ed), Handbook of Stress in the Occupations (2011) ch

21

201 H Rosling, The Magic Washing Machine (2010) youtube.com

202 CE Beecher and H Beecher Stowe, American Woman’s Home: Or,

Principles of Domestic Science (1842) ch XXV

203 K French-Fuller, ‘Gendered Invisibility, Respectable Cleanliness: The

Impact of the Washing Machine on Daily Living in Post-1950 Santiago, Chile’
(2006) 18(4) Journal of Women’s History 79

204 JC Robinson, ‘Of women and washing machines’ (1985) 101 China
Quarterly 32

205 L Hannah, Inventing Retirement: The development of occupational

pensions in Britain (CUP 1986) 13

206 Chart from Google’s ngram, showing the incidence of words in books.

207 e.g. ILO Hours of Work (Industry) Convention 1919 (c 1) and then FT
Vyver, ‘Regulation of Wages and Hours Prior to 1938’ (1939) 6(3) Law and
Contemporary Problems 323, or M Seidman, ‘The Birth of the Weekend and the
52
Revolts against Work: The Workers of the Paris Region during the Popular
Front’ (1981) 12(2) French Historical Studies 249.
208 H Rosling, The Magic Washing Machine (2010) youtube.com

209 S Strasser, Never Done: A History of American Housework (1982) ch 5,

on piped water in Philadelphia, increasing consumption x 4

210 One might also point to public education and health, which are the basis
of ideas for all technology.

211 Terminator 2: Judgment Day (1991) starring L Hamilton and A

Schwarzenegger

212 J Bessen, ‘How Computer Automation Affects Occupations: Technology,

jobs, and skills’ (2016) Boston University, Law & Economics WP No. 15-49

213 n.b. ‘Intelligence’ means ‘the ability to acquire and apply knowledge and
skills.’

214 n.b. ‘Cognition’ means the ‘process of acquiring knowledge and

understanding through thought, experience, and the senses.’

215 See Q Skinner, Visions of Politics (2002) vol 1, ch 2, ‘The practice of

history and the cult of fact’

216 cf Aristotle, Nicomachean Ethics, Book V, Pts 3-4, who argued

‘corrective’ justice is arithmetic; ‘distributive’ justice is geometric.

217 See D Silver, ‘Mastering the game of Go without human knowledge’ (19
October 2017) 550 Nature 354 and cf G Kasparov and M Greengard, Deep
Thinking: Where Machine Intelligence Ends and Human Creativity Begins
(2017)

218 See R Descartes, Discourse on the Method (1637) Part 4. It is likely his
famous ‘eureka’ moment was simmering for some time.

219 L Wittgenstein, Tractatus Logico-Philosophicus (1921)

220 L Wittgenstein, Philosophical Investigations (1953) §43. cf G Kitching,

Wittgenstein and Society: Essays in Conceptual Puzzlement (2003)
53
221 See also E Durkheim, Division of Labour in Society (1893) 165-6

222 W Beveridge, Full Employment in a Free Society (1944) 18

223 For clarity, nobody has an inalienable right to demand society funds
courts to protect productive property rights when those rights are not being
utilised, and instead squandered: JS Mill, On Liberty (1859) ch 5, ‘trade is a
social act’.

224 TFEU arts 9, 119-150 and TEU art 3(3), ‘aiming at full employment’.
Bank of England Act 1998 ss 1, 11-13 and 19. Central banks which pay no
regard to rising inequality in monetary policy, cannot be seen as complying with
duties for price stability.

225 Bhasin v Hrynew 2014 SCC 71 and J Steyn, ‘Contract Law: Fulfilling the
Reasonable Expectations of Honest Men’ (1997) 113 LQR 443

226 European Social Charter 1961 art 2(1) requires ‘the working week to be
progressively reduced’ with ‘increase of productivity’

227 See further E McGaughey, ‘A Twelve Point Plan for Labour, and A
Manifesto for Labour Law’ (2017) 46(1) ILJ 169

228 RB Freeman, ‘Who owns the robots rules the world’ (May 2015) IZA
World of Labor. Freeman has advanced the same idea before, most recently in
JR Blasi, RB Freeman and DL Kruse, The Citizen’s Share: Putting Ownership
Back into Democracy (2013)

229 E McGaughey, ‘Votes at work in Britain: shareholder monopolisation and

the ‘single channel’’ (2017) 46(4) Industrial Law Journal (summary on the LSE
British Politics and Policy Blog).
230 PJ Purcell, ‘The Enron Bankruptcy and Employer Stock in Retirement
Plans’ (11 March 2002) CRS Report for Congress

231 AA Berle, ‘How Labor Could Control’ (1921) 28 New Republic 37

232 AA Berle, ‘Property, Production and Revolution’ (1965) 65 Columbia Law

Review 1, 17.

54
233 e.g. Oxford University Act 1854 ss 16-21 and Cambridge University Act
1 8 5 6 ss 5-51. E McGaughey, ‘Votes at Work in Britain: Shareholder
Monopolisation and the ‘Single Channel’ (2017) 46(4) Industrial Law Journal.
Also, ‘The codetermination bargains: the history of German corporate and
labour law’ (2016) 23(1) Columbia Journal of European Law 135.

234 See further E McGaughey, ‘Ideals of the Corporation and the Nexus of
Contracts’ (2015) 78(6) Modern LR 1057, 1061-1070

235 See T Raiser, ‘The Theory of Enterprise Law in the Federal Republic of
Germany’ (1988) 36(1) American Journal of Comparative Law 111

236 S Deakin, ‘The Corporation as Commons: Rethinking Property Rights,

Governance and Sustainability in the Business Enterprise’ (2012) 37(2) Queen’s
Law Journal 339. cf I Ferreras, Firms as Political Entities (2017) 1-11 and chs
5-6.

237 e.g. in the Model Articles for Public Companies, Sch 3, a new para 34(2)
could state: ‘The company’s workers shall be registered as members and entitled
to 30 per cent of the total votes in the general meeting. Each person shall have
one vote.’

238 In the US, a norm first found in the Delaware General Corporation Law of
1969 §141(h). See S Arsht and WK Stapleton, ‘Delaware General Corporation
Law: 1969’ (1969) 25 Business Lawyer 287. In the UK, the astonishing Model
Articles, Sch 3, para 23(2) read ‘Directors are entitled to such remuneration as
the directors determine...’

55
239 E McGaughey, ‘Does Corporate Governance Exclude the Ultimate
Investor?’ (2016) 16(1) Journal of Corporate Law Studies 221, 237-8 (summary
on the Oxford Business Law Blog). LD Brandeis, Other People’s Money and
How the Bankers Use It (1914)

240 UDHR 1948 art 25, ‘Everyone has the right to a standard of living
adequate for the health and well-being of himself and of his family, including
food, clothing, housing and medical care and necessary social services, and the
right to security in the event of unemployment, sickness, disability, widowhood,
old age or other lack of livelihood... beyond his control.’

241 This is the view taken by T Paine, Agrarian Justice (1797) on a ‘national
fund’ to pay ‘compensation in part, for the loss of his or her natural inheritance,
by the introduction of the system of landed property.’

242 The Sozialgesetzbuch has twelve volumes: SGB II concerns

unemployment, SGB VI pensions.

243 On the US today see E McGaughey, ‘Fascism-lite in America (or the

Social Ideal of Donald Trump)’ (2018) BJALS

244 cf J Hickel, The Divide (2017) ch 9 and ‘Basic income isn’t just a nice
idea. It’s a birthright’ (4 March 2017) Guardian

245 J Portes, H Moore et al, Social prosperity for the future: A proposal for
universal basic services (2017) Institute for Global Prosperity

246 See also G Standing, Basic Income: And How We Can Make It Happen
(2017) ch 7, summarising proposals. Standing favours abolishing personal
allowances to encourage voting. But auto voting enrolment, or even nudges in
pay-slips, seem easier.

247 Basic income transfers to employed people (but not tax exemptions) risk
subsidising employers partly because of the ‘endowment effect’ (workers see
pay as theirs, and hang on, but a transfer payment from government less so) and
partly because of employers’ potential political influence. See also S Deakin and
F Wilkinson, ‘Labour law, social security and economic inequality’ (1991) 15
Cambridge Journal of Economics 125, 144-5

248 UDHR 1948 art 23(3)

56
249 cf H Hansmann and R Kraakman, ‘Towards Unlimited Liability for
Corporate Torts’ (1991) 100 Yale LJ 1879

2 5 0 See LA Bebchuk and JM Fried, ‘The Uneasy Case for the Priority of
Secured Claims in Bankruptcy’ (1996) 105 Yale LJ 857, 881-890. ‘Non-
adjusting’ creditors include tort victims, employees, consumers and small
businesses, but a bank is ‘adjusting’.

251 cf RA Coase, ‘The Problem of Social Cost’ (1960) 3 JLE 1. Coase refused
to recognise bargaining power’s relevance.

252 Walkovszky v Carlton, 223 NE2d 6 (NY 1966)

253 n.b. in the UK, where the National Health Service is free, there were no
cases until Chandler on the corporate veil and personal injuries (save one
doubtful conflict of laws case). See C Mitchell, ‘Lifting the corporate veil in the
English courts: an empirical study’ [1999] Company, Financial and Insolvency
Law Review 15. This suggests tort law was needed less with the NHS.

254 PL Davies, Company Law (2010) ch 4. There is no inherent reason why

tort law cannot hold one person (human or legal) responsible for what another
does. See the pathbreaking cases of Chandler v Cape plc [2012] EWCA Civ 525
and Lungowe v Vedanta Resources plc [2017] EWCA Civ 1528. The problem is
rather that judicial paths have been historically hard to break.

255 n.b. E Palmerini et al, ‘Regulating Emerging Robotic Technologies in

Europe: Robotics facing Law and Ethics’ (2014) Robolaw, EU Grant number
289092, suggested ‘care robots will become more sophisticated and may’ assist
‘in purchasing food, drugs, newspapers, or bus tickets.... it could be useful if
robots would have the capacity to perform legal transactions.’ This is, with
respect, inaccurate: if I write a cheque, does my pen need legal personality for
the transaction to take effect?

256 This would mean updating the Product Liability Directive 85/374/EEC.
On possibly the first thorough discussion of these issues, see A Bertolini,
‘Robots as Products: The Case for a Realistic Analysis of Robotic Applications
and Liability Rules’ (2013) 5(2) Law, Innovation and Technology 214.
257 JL Delaney, ‘The robot that takes your job should pay taxes, says Bill
Gates’ (17 February 2017) Quartz
57
258 cf B Gates, The Road Ahead (1996) 19, ‘luck has played a role, but I think
the most important element was our original vision.’ See further, P Piff, ‘Does
money make you mean?’ (2013) TEDxMarin, showing Monopoly game players
who got twice the money, more for passing go, and double dice rolls, viewed
their (inevitable) victory as deserved, despite knowing it was rigged.

58