alberto alemanno

Electronic copy available at: http://ssrn.com/abstract=2005672Electronic copy available at: http://ssrn.com/abstract=2005672

EUROPEAN JOURNAL OF RISK REGULATION – 1/2012 – Symposium on Nudge

EUROPEAN JOURNAL OF RISK REGULATION – 1/2012

Symposium on Nudge

Nudging’ Healthy Lifestyles: The UK Experiments with the Behavioural Alternative to Regulation and the Market

Adam Burgess

Liberalism and Lifestyle: Informing Regulatory Governance with Behavioural Research

On Amir and Orly Lobel

Nudging Cannot Solve Complex Policy Problems Evan Selinger and Kyle Powys Whyte

Nudging Smokers: The Behavioural Turn of Tobacco Risk Regulation Alberto Alemanno

Real Nudge Luc Bovens

Electronic copy available at: http://ssrn.com/abstract=2005672Electronic copy available at: http://ssrn.com/abstract=2005672


Introduction

The EJRR starts the new year by hosting a pioneering symposium devoted to one of the latest policy innovations that is currently experimented in the United Kingdom and the United States: the ubiquitous, yet controversial, Nudge. This idea originates from the homonymous, 2008 best-selling book published by the economist Richard Thaler and the legal scholar Cass Sunstein. By building upon the findings of behavioural research, they refute the classic economic assumption that “each of us thinks and chooses unfailingly well”1 and they advocate the need for public authorities to nudge people to make decisions that serve their own long-term interests without however removing their right to choose.

At a time in which governments are taking considerable interest in the use of nudging, we have asked some of the leading authors who have already contributed to the literature surrounding the regulatory innovations, generally referred as New Governance, to share their ideas on this appealing regulatory approach.

In his opening essay, Nudging Healthy Lifestyles, Adam Burgess provides a critical assessment of the introduction of behavioural, nudging approaches to correct lifestyle behaviours in the UK. His thought-provoking analysis triggered a lively debate that has been framed along the subsequent essays signed by On Amir and Orly Lobel, Evan Selinger and Kyle Powys White, Alberto Alemanno and Luc Bovens.

Alberto Alemanno

Editor-in-Chief, European Journal of Risk Regulation

Jean Monnet Professor of EU Law & Risk Regulation, HEC Paris

1 Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth and Happiness, (London: Yale University Press, 2008), p. 6, ∗ Reader in Social Risk Research, University of Kent. Thanks to the anonymous reviewer who provided detailed and very useful criticism of the original article. Also to Tracey Brown, Jerry Busby and Helen Reece for helping think through some of the issues. 2 See speech by David Cameron, “David Cameron attacks UK 'moral neutrality”, Daily Telegraph, 7 July 2008, available on the

Electronic copy available at: http://ssrn.com/abstract=2005672


‘Nudging’ Healthy Lifestyles: The UK Experiments with the Behavioural Alternative to

Regulation and the Market

Adam Burgess∗

This article critically reflects upon the introduction of behavioural, ‘nudging’ approaches into UK policy making, the latest in

a series of regulatory innovations. Initiatives have focused particularly upon correcting lifestyle risk behaviours, marking a

significant continuity with previous ‘nannying’ policy. On the other hand, nudging represents a departure, even inversion of

previous approaches that involved the overstating of risk, being based partly upon establishing a norm that bad behaviours are

less, rather than more common than supposed. Despite substantive similarities, its attraction lies in the reaction against the

former approach but must also be understood in the context of the economic crisis and a diminished sense of liberty and

autonomy that makes intimate managerial intervention seem unproblematic. Problems are, in fact, substantial, as nudging is

caught between the utility of unconscious disguised direction and the need to allow some transparency, thereby choice. Further,

it assumes clear, fixed ‘better outcomes’ but encourages no development of capacity to manage problems, contradicting a wider

policy intent to build a more responsible and active citizenry. More practically, nudging faces considerable barriers to becoming

a successfully implemented programme, in the context of severe, Conservative-led austerity with which it is now associated.

I. ‘Nudge’ crosses the Atlantic

A new Conservative-led Coalition government came to power in May 2010 in the UK. They have initiated

a programme of ‘nudging’ individuals into making better choices through manipulating their environment,

as part of a radical programme of transformation. Alongside wellbeing, transparency and decentralisation

∗ Reader in Social Risk Research, University of Kent. Thanks to the anonymous reviewer who provided detailed and very useful criticism of the original article. Also to Tracey Brown, Jerry Busby and Helen Reece for helping think through some of the issues.


this experiment with behavioural economics is one of the new emphases in government thinking. All this

is underpinned with the promotion of an ethos of promoting greater personal responsibility.2

Conservative Party leader, David Cameron, has been engaged in a long term project of creating a modern,

liberal conservatism. Now in power, and despite the limited mandate of a minority government, the

Coalition have quickly moved to attempt a transformation of institutions, and the individuals’ relationship

to them. Power and responsibility are to be devolved to the local level, in an oddly entitled project to

establish a ‘big society’.3 Even critics are agreed that one thing the programme does not lack is ambition; it

is widely seen as nothing less than a cultural revolution in a country historically defined by its gradualism

and stability.4 Thus university funding has been virtually abandoned, with costs passed on to individual

students, and much of the health budget is to be devolved to the control of local doctors. Alongside this

large scale cost-cutting, the government is encouraging a philosophy of smart, cheap and local solutions to

a range of problems in society, inspired by modern behavioural economics. The Coalition’s message is to

pose the classical question: ‘Ask not what society can do for you, but what you can do for society’. Their

role is to make that more possible, be that through making it easier to do voluntary work, or reduce the

burden on the health service by encouraging healthier lifestyles.

The dramatic change of political direction in the UK has been accompanied by a shift in the sources of

intellectual inspiration. The previous Labour government expanded the contribution of social research to

its ‘evidence based’ programmes, encouraging a language of risk avoidance that Cameron made a focus of

2 See speech by David Cameron, “David Cameron attacks UK 'moral neutrality”, Daily Telegraph, 7 July 2008, available on the Internet at

<http://www.telegraph.co.uk/news/newstopics/politics/conservative/2263705/David-Cameron-attacks-UK-moral-neutrality---full-text.html> (last accessed on 11 January 2012).

3 For a mapping of the Conservatives’ agenda see Nick Boles, Which Way’s Up? (London: Biteback Books, 2010) by a leading moderniser and Cameron advisor. 4 Astonishingly, their programme has been frequently described as ‘Maoist’ in character. See, for example, Ed Rooksby, “Vince Cable is right: in some ways the coalition is a bit like Maoism”, The Guardian, 23 December 2010, available on the Internet at <http://www.guardian.co.uk/commentisfree/2010/dec/23/vince-cable-mao-coalition-marxist-capitalism?INTCMP=SRCH> (last accessed on 11 January 2012).


attack during the election.5 The Conservatives, by contrast, are drawing upon the disciplines of psychology

and economics, with their starting points in the abstract individual, rather than social structure or context.

More specifically, the new government is drawing upon behavioral rather than neoclassical economics,

which is based on the more real and imperfect ways in which the individual makes choices, rather than the

purely rational actor of neoclassical theory. This is combined with insights from social psychology -

nowhere more clearly and attractively than in the bestselling account, Nudge by Richard Thaler, often

described as the founder of behavioural economics, and Cass Sunstein, similarly described as the father of

behavioural law and economics, and now ‘nudger in chief’ as head of the Office of Regulatory Affairs in

Obama’s American administration.6 The book is a practical manual for applying the principles of

behavioural economics; how institutions can configure ‘choice architecture’ to encourage beneficial

decision making amongst the population. The Conservative modernisers were drawn to the practical, non-

ideological solutions suggested by Nudge, and Cameron made it obligatory reading for his colleagues

before the election. By mid 2008 the Conservatives were consulting with Thaler over specific schemes

such as to reduce the carrying of knives by young people and increasing recycling.7 An example of the

kind of solutions being explored are American schemes that attempt to reduce electricity consumption by

providing feedback to households on their own and neighbours’ usage. The intention is that individuals

will try and adjust – hopefully reduce – their consumption in line with the norm (and thereby reinforcing

the norm itself).8

Behavioural economics, now some 40 years old, is not the same as the ‘nudging’ with which it is now

associated, but lends itself to all manner of approaches. The boundaries between behavioural economics,

5 See Cameron, David Cameron attacks UK’ moral neutrality, supra note 1.

6 Richard H. Thaler and Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth and Happiness, (London: Yale University Press, 2008).

7 As an indication of the extent, and public nature of collaboration, Cameron’s second-in-command, Chancellor George Osborne, co-wrote an article with Thaler – with the unfortunate title, “We can make you behave”, The Guardian, 28 January 2010, available on the Internet at <http://www.guardian.co.uk/commentisfree/2010/jan/28/we-can-make-you-behave?INTCMP=SRCH> (last accessed on 11 January 2012).

8 See Dora Costa and Matthew Kahn, Energy Conservation “Nudges” and Environmentalist Ideology: Evidence from a Randomized Residential Electricity Field Experiment. Presentation at 2010 Power Conference, Princeton University, available on the Internet at <http://academics.hamilton.edu/economics/home/kahn_hamilton_paper.pdf> (last accessed on 11 January 2012).


various forms of ‘nudging’, social marketing and psychology are not clear, but they all inform the policy

mix that is now generally known as ‘nudging’.9 The language of ‘nudges’ and ‘shoves’ has been a relatively

familiar one in the US for more than a decade.10 There is also a significant crossover between behavioural

and ‘new governance’ approaches in regulation.11

Behavioural approaches proceed from the recognition that we misjudge decisions systematically because

of our inherent biases and rules of thumb for making sense of information. These ‘heuristics’ were

elaborated most famously in the experiments of Kahnemann and Tversky, and are laid out in Nudge.

Among other insights, the distinction between short term pleasure and long term benefit is a key one in

the behavioural armoury, which explains a range of our (at least formally) irrational decision making. A

wide range of behavioural studies continue.12 Whilst some seem obvious, even banal, other studies are

more interesting and useful, such as observation of how we choose to pay back the small ones first when

dealing with multiple loans – despite a higher rate of interest on the larger one.13 Another example

underlines the potentially large impact behavioural modification can make, despite the typically modest

nature of ‘nudges’ in themselves. A field experiment in China on the effect of incentives upon productivity

suggests the difference between those framed positively and negatively amounted to over 1 % per annum;

9 Alongside Nudge, the two key texts now usually cited are Robert Cialdini, Influence: The Psychology of Persuasion (New York: Harper Business, 2007), and Ariely, Predictably Irrational, supra note.7.

10 Dan Kahan, “Gentle Nudges versus Hard Shoves: Solving the Sticky Norms Problem”, 67(3) University of Chicago Law Review (2000), at pp. 607–646.

11 On Amir and Orly Lobel, “Stumble, Predict, Nudge: How Behavioral Economics Informs Law and Policy”, 108 Columbia Law Review (2008), at pp. 2098–2139.

12 For a useful review see, Amir and Lobel, Stumble, Predict, Nudge, supra note 10, at pp. 2127–2132. 13 For example, findings that we are more prepared to have unprotected sex when sexually aroused, or that we are more likely to steal stationery from work than the financial equivalent. See these and other examples in Dan Ariely, Predictably Irrational: the Hidden Forces that Shape Our Decisions, (New York: Harper Collins 2008). Results are perhaps less surprising than why it was thought these issues were thought worthy of investigation in the first place.


a ‘hugely significant’ difference.14 The collective insights from behavioural studies have been drawn

together in simple terms through the UK government’s MINDSPACE framework.15

These approaches are addressed to demonstrated patterns of behaviour that are deeply embedded, and

realistically draws upon them to produce better outcomes. The contrast here would be with policy making

that imagines that all human behaviour is readily changed, and that it is only a matter of finding the right

combination of carrot and stick. Doing this, it embraces the power of social norms and pressures rather

than an assumption that simply changing laws, regulations or policies will be effective. Nudging’s strength

is its practical character. It attempts to design around our imperfections for positive social ends,

recognising that we are typically lazy about what we choose not to prioritise, but nonetheless regard as

right. An example would be ‘getting around to’ agreeing to organ donation, for example, something which

behavioural thinking sees could be changed through establishing an ‘opt out’ rather than ‘opt in’ system.16

Nudging is grounded rather than exhortative, calling into question the assumption that more information

about negative consequences will result in improved behaviours and providing the modern policy maker

with a fresh and practical perspective on a range of problems in society.

The centrality of using behavioural insights to the Coalition government is such that it introduces the

formal written agreement between the Conservative and Liberal parties, in the foreword by the two party

leaders:

“There has been the assumption that central government can only change people’s behaviour

through rules and regulations. Our government will be a much smarter one, shunning the

14 This interesting example is cited in the interview with Ariely in Matthew Taylor, “Better the devil you know”, RSA Journal, available on the Internet at <http://www.thersa.org/fellowship/journal/features/features/better-the-devil-you-know> (last accessed on 11 January 2012).

15 Cabinet Office and Institute for Government, MINDSPACE: Influencing Behaviour through Public Policy (London: Cabinet Office, 2010). 16 Ben Saunders, “Normative Consent and Opt-Out Organ Donation,” 36(2) Journal of Medical Ethics (2010), at pp.84–87.


bureaucratic levers of the past and finding intelligent ways to encourage support and enable

people to make better choices for themselves”.17

They have set up a 7-person ‘Behavioural Insight Team’ within the Prime Minister’s Cabinet Office – a

statement of priority on its own, at a time when everywhere else in the administration is experiencing

significant cuts to personnel. The country’s then leading civil servant, Gus O’Donnell, is directly involved,

charged with ensuring delivery. Behavioural solutions are here considered in the context of a wider

devolution of power to local communities, the ‘nudge unit’ looking to examples from around the world of

schemes where communities themselves appear to have solved problems without the direction of central

authorities. David Halpern, head of the new unit, claims that, in combination with greater transparency

and social network affects it can be ‘genuinely transformative’.18 They have now set out their first

initiatives, centred primarily on eliminating various ‘unhealthy behaviours’.19 The strategy is to school the

civil service in ‘smart’, behavioural solutions to social problems. It is through this mechanism that they

hope to square the awkward contrast between a new centralised directing body, and the local initiative

supposedly at the heart of the new project. The intention is that officials schooled in the new approach

will successfully diffuse into local communities the spirit rather than only the prescriptive details of what

they have learnt from the behavioural insight team.

Despite the recent association of modern Conservatism with ‘nudge’, it is worth noting that there is no

necessary political connection along these lines. Whilst there may seem to be a natural affinity between

Conservatism’s and economics’ focus upon the individual, this is by no means exclusive as ‘socialist’

17 The Coalition, Our Programme for Government (London: HM Government, 2010) at p.7, available on the Internet at <http://www.direct.gov.uk/prod_consum_dg/groups/dg_digitalassets/@dg/@en/documents/digitalasset/dg_187876.pdf> (last accessed 11 January 2012). 18 David Halpern, Institute for Government blog, “No.2 to No.10: Taking Mindspace to Downing Street”, available on the Internet at <http://www.instituteforgovernment.org.uk/blog/764/no-2-to-no-10-taking-mindspace-to-downing-street/> (last accessed 11 January 2012).

19 Their initial projects are on: smoking, organ donation, teenage pregnancy, alcohol, diet and weight, diabetes, food hygiene, physical exercise, and social care. See “Applying Behavioural Insight to Health”, (London: Cabinet Office Behavioural Insights Team, 2010), available on the Internet at <http://www.cabinetoffice.gov.uk/resource-library/applying-behavioural-insight-health> (last accessed on 11 January 2012).


politics has adapted itself in this direction in the UK, like elsewhere.20 The ‘MINDSPACE’ research

programme which shaped the Coalition’s behavioural work was commissioned by the previous Labour

government, and Halpern worked under them.21 In the American context, nudge is a Democrat

administration initiative.22 After all, nudging is precisely intended to represent an alternative to clear

interventionist approaches; an attempt at a ‘third way’ between the regulation associated with the left, and

‘leave it to the markets’ approach of the right.23 Back in the UK, had the Labour Party not been in a

period of disarray in the mid 2000s it could have been they, not the Conservatives, who adopted it as their

own.

II. Economic crisis and the reaction against ‘nannying’

Whatever the political identification, it isn’t self evident in its own terms why behavioural approaches not

only have a certain appeal, but have been adopted in such a wholesale and exclusive fashion. It is not

driven by compelling evidence that it works, especially not as a general approach to a range of problems.

For the most part it is only evidence-based in the sense that proposed schemes are informed by particular

misperceptions found in experimental settings. 24 It is in this context we can understand why the cleverly

modified urinals at Amsterdam airport are so often cited; not only is the placing of a pretend fly a funny

example, it is also one of the few clear successes.25 More often, examples cited are, actually, tentative (but

tend not to be suitably qualified). Certainly, part of the answer to its appeal can be located in its post

ideological, centrist character. Contemporary politics on both sides of the Atlantic has increasingly

20 On the fundamental continuity between recent Labour and Conservative administrations, and way in which the Conservatism of Margaret Thatcher subsequently shaped Labour policy, see Simon Jenkins, Thatcher and Sons: A Revolution in 3 Acts (London: Penguin, 2007). 21 MINDSPACE, Influencing Behaviour through Public Policy, supra note 14. 22 For a journalistic account of how nudging is being used in the Obama administration see, Michael Grunwald, “How Obama is using the science of change”, Time, 2 April 2009, available on the Internet at <http://www.time.com/time/magazine/article/0,9171,1889153,00.html> (last accessed 11 January 2012). 23 Sunstein situates ‘Nudge’ in this context in this interview, available on the Internet at <http://www.youtube.com/watch?v=DD-fUJs5t_k> (last accessed 11 January 2012). 24 Lord Alderdice quizzed civil servants applying behavioural solutions, making the useful distinction between ‘informed design’ which describes most behavioural ideas, and actually ‘evidence based policy’. See House of Lords Select Committee on Science and Technology Inquiry on Behaviour Change, Evidence Session 2 (2 November 2010), available on the Internet at <http://www.parliament.uk/documents/lords-committees/science-technology/behaviourchange/ucSTI021110ev1.pdf> (last accessed on 20 December 2010).

25 Men aim at the ‘fly’, reducing spillage.


concentrated around a middle ground to which nudge solutions seem ideally suited. But more important is

the wider context of the economic crisis that has undermined faith both in conventional economics and

the economic system itself. As Dan Ariely, one of the principal behavioural economists explains:

“Without the financial crisis, I don’t think behavioural economics would have gained the

popularity it has. Almost everyone believed that the market was the most rational place on the

planet, yet it failed in a magnificent way. This proved that people who deal with large amounts of

money are as capable of irrationality – from reckless gambling to myopia and overconfidence – as

anybody else”.26

Behavioural economics’ contemporary appeal is, in this regard, primarily a negative one; in a context of

disillusionment with conventional assumptions and solutions. As one commentator put it: ‘Nudge is

relevant in 2009 because it helps us look in the mirror’, at our own, and the economic system’s failings.27

Chancellor George Osborne, the Conservatives second-in-command, made a case for the relevance of

behavioural economics to recessionary circumstances before they came to power, on the basis that not

only do individuals behave irrationally, but so do markets.28 Its subsequent appeal within the British policy

making circles is the promise of cost effectiveness; achieving ‘more for less’, particularly in public services.

The practical emphasis is on ‘smart’ solutions that do not involve more resources; an imperative in

recessionary times with a government committed to reducing spending.

A second factor explaining the appeal of nudging is one particularly relevant in the British context, and

one that emphasizes why these developments are of interest to those concerned with risk and regulation.

26 Taylor, Better the devil you know, supra, note 13.

27 Jon Dennis, “We still need a nudge”, The Guardian, 25 March 2009, available on the Internet at <http://www.guardian.co.uk/commentisfree/2009/mar/25/nudge-economics-social-policy> (last accessed on 11 January 2012).

28 Andrew Sparrow, “Nudge Economics still relevant in recession”, Guardian blog, 8 April 2009, available on the Internet at <http://www.guardian.co.uk/politics/blog/2009/apr/08/nudge-george-osborne> (last accessed on 11 January 2012).


Not only does it promise a potentially a cheaper means of dealing with social problems, it is also distanced

from the heavy-handed regulatory approach identified with the previous administration. In this sense it

marks an important marker of identity for a new administration that otherwise shares much in common

with its similarly centrist predecessor. The Coalition promote ‘nudging’ as an alternative to the legislative

restrictions and ‘nannying’ of the Labour years, showing that government can do more than ‘leave it to the

market’ but without expensive, intrusive and unpredictable interference.29 Thaler and Sunstein describe

their approach as both ‘paternalist’ and ‘libertarian’, and it is the latter aspect that is emphasized in the UK

case, as the Coalition distinguishes itself from its predecessor.30 From this perspective the behavioural

initiative can be placed in the deregulatory thrust of recent British ‘better regulation’ initiatives: the Better

Regulation Executive, Better Regulation Commission and Risk Regulation Advisory Council.31 The

Coalition agreement foreword cited above declares that it will be a ‘smarter one’ than its Labour

predecessor, which resorted to ‘bureaucratic levers’ all too often. Health Secretary, Andrew Lansley,

counter posed Labour’s excessive legislative zeal and ‘nannying’ to his ‘new approach to public health’,

heavily framed by nudging, in July 2010.32

The backdrop here is a contemporary politics of heightened risk aversion and precaution that reached its

climax under the last Labour government, far too narrowly presented now as simply one of a bossy

‘nannying’. This is not to say that there was no ‘nannying’; sweeping, paternalistic criticism of unhealthy

behaviours became relatively routine. For example back in 2007, without embarrassment or qualification,

then Public Health Minister, Dawn Primarolo told off middle class, “everyday” drinkers who have drunk

too much for too long. This has to change’.33 Another health minister, Gillian Merron instructed us how

best to spend our Christmas holidays: “Whatever the weather, a traditional festive walk is a great way for

29 The notion of ‘nannying’ derives from the phrase the ‘nanny state’; that is an overly interventionist style of government that dictates to its citizens like a grandmother (‘nanny’). 30 Reaction to this description prompted a direct response from the authors; Cass R. Sunstein & Richard H. Thaler, Libertarian Paternalism Is Not an Oxymoron, 70 U. Chi. L. Rev. (2003), at p. 1159, and expanded on in Nudge (pp. 4–6). This has been the subject of much further debate, such as in Amir and Lobel, Stumble, Predict, Nudge, supra note 6.

31 These initiatives were, somewhat paradoxically, initiated by Labour governments. They were, practically, marginalised however, most clearly with the Risk Regulation Advisory Council created by Gordon Brown. 32 Andrew Lansley, “A New Approach to Public Health”, 7 July 2010, Speech to Faculty of Public Health conference, available on the Internet at <http://www.dh.gov.uk/en/MediaCentre/Speeches/DH_117280> (last accessed on 11 January 2012).

33 Cited in Duncan Brown, “Hazardous drinking, the middle class vice”, The Times, 16 October 2007, at p.7.


families and friends to avoid that sluggish feeling and have a more active Christmas.’34 But such

pronouncements were part of a much wider promotion of risks intended to stimulate a change of thinking

and behaviour. High alcohol and food consumption were presented as polarised and simple matters of life

and death, and as dramatic and implausibly recent and fixed epidemics.35

A wide range of prominent risks - from terrorists, paedophiles and possible disease threats, to purported

dangers from chemicals, food and alcohol – were politicised and brought to public attention. At the same

time the belief was encouraged that risk might somehow be eradicated if sufficient energy and resources

were dedicated to it, and all obstacles to this agenda eliminated.36 To an extent this was through the

classical modern form of a punitive law and order agenda, stimulated by outrageous criminal incidents in

the intensely reactive political culture of those years. During his decade in power Tony Blair presided over

more than 3000 new laws, 1000 of which carried jail terms.37 But a wider range of contemporary risks to

individual health and safety were politicised during these years. Ministers found themselves having to

promote various health anxieties that acquired a life of their own - often through the media. Thus Health

secretary John Reid appeared in a populist newspaper under the headline, ‘I salute the Sunday Express for

its hospital crusade’ (over ‘super bugs’), back in 2003. Here he rehearsed what became a well-worn ritual

of government adapting itself to what the media indicated were popular concerns and anxieties. 38

Characteristically the laws and regulations that emerged in these contexts were hastily and badly drafted,

reflecting the fact that they were substantially a means of demonstrating that government was ‘doing

something’ about whichever new risk had been thrust to the centre of attention.39 Not unfairly this has

34 See ‘Burn off Christmas with a walk, BBC News Online, 22 December 2009, available on the Internet at <http://news.bbc.co.uk/1/hi/health/8425433.stm> (last accessed 11 January 2012). 35 Heavy drinking in public is an endemic not epidemic phenomenon in the UK; a long established cultural trend. At the same time it is still subject to considerable fluctuation because of shifts in social behaviour, such as the decline in young people’s drinking in the later 2000s. See Adam Burgess, “The Politics of Health Risk Promotion. ‘Passive Drinking’: A ‘Good Lie’ Too Far?” 11(6) Health Risk and Society (2009), pp.527 et sqq. p.536. 36 For a substantial account of these developments in the UK see Michael Moran, The British Regulatory State: High Modernism and Hyper Innovation, (Oxford: Oxford University Press, 2007). An important policy response was provided by the Better Regulation Commission, Risk, Responsibility and Regulation – Whose Risk is it Anyway? (London: Better Regulation Commission, 2006). This provided the basis for developing the Risk Regulation Advisory Council.

37 Philip Johnston, Bad Laws (London: Constable, 2010), p.10 et sqq. 38 See Adam Burgess, “Media Risk Campaigning: From Mobiles Phones to Baby P”, 13(1) Journal of Risk Research (2010), pp.59 et sqq., at p. 69.

39 For example, the Safeguarding Vulnerable Persons Act - which subjected all adults coming into contact with children on any regular basis to regulation – was so badly drafted that it required 250 amendments by the end of its parliamentary progress. The


been described as: ‘truckloads of legislation simply to send out signals, make a point or obtain a

headline’.40

Health became a particular focus for risk concerns in a wider landscape of risks that required management

and vigilance. As public anxieties were seen to centre upon risks to health, it became natural to engage

with, and expand these concerns. It is striking that Gordon Brown chose the promise of universal health

screening as a last ditch effort to save his Premiership, in his re-launch, in late 2009, for example. Health

risks even became the villains in a curious new populism, such as when Brown called on manufacturers

and the European Union to take action against food additives, following the publication of a single study

suggesting they might affect children’s behaviour in 2007.41 The appeal of promises to remove risk

required that it first be elevated, and there was, in effect, an approach of attempting to alarm the public

into behavioural change. This is most widely recognised in relation to how the threat of terrorism was

relentlessly promoted by both Labour premiers, and draconian legislation passed to combat this allegedly

pervasive threat.42 But, again, a wider range of risks became a focus for politicisation. Alarming

representations of risk to stimulate public responses became relatively routine in the Labour years, from

dramatising and denying uncertainty around climate change, to inventing notions of ‘passive drinking’ by

the Chief Medical Officer.43 This, latter example, illustrates a process of unchecked regulatory expansion

and over confidence; it was success in banning ‘passive smoking’ that created both a need for a new

campaigning focus, and the confidence to construct a further, but this time implausible, target.

This process of risk dramatisation and expansion involved a range of characteristic techniques, among

which were the denial of any uncertainty, reliance upon the most dramatic examples, and use of worst-

Act was a response to the publicity generated by the murder of two schoolgirls by a school caretaker, Ian Huntley, and the subsequent public inquiry.

40 Johnston, Bad Laws, supra note 39, p.36. 41 Paul Webster and Dennis Sanderson, “Brown’s plea to take the additives out of children’s food”, The Times, 7 September 2007. p.5. 42 Most controversially, Labour’s anti terrorism was pursued through ‘control orders’ – virtual house arrest – and through a 28-day detention for suspects. These were major behind-the-scenes controversies for the Coalition government who were formally committed to their abolition but eventually only moderated them. 43 Burgess, The Politics of Health Risk Promotion, supra note 34.


case scenario projections as realistic models. There is some generalised recognition of, and reaction against

this approach in behavioural thinking. Cialdini’s widely-cited policy making ‘big mistake’ is to imply that

problem behaviours are relatively widespread.44 This form of risk politics centred on making problems

appear of more immediate and pervasive dimensions than they really are, is the opposite of the nudge

device, which is to underplay them instead. The idea is to make individuals believe the social norm is

different, typically lower, than they imagine in the case of consumption, stimulating them to adjust their

behaviour and eat, drink and smoke less, or consume less energy. Whilst the agenda may not be radically

different - particularly the promotion of lifestyle risk avoidance - the means of doing so certainly is.45 To

an extent, nudge represents recognition of the failure of an old risk politics perfected during the Labour

years in the UK, even whist the fundamental agenda remains unquestioned.

III. Managerial project in an era of devalued freedom?

A third factor explaining nudge’s contemporary appeal is a further negative one; that barriers to its

application have fallen away. The impact of psychologically-based behaviourism has long been the subject

of concern about its impact upon civic freedoms and rights, but these concerns are less marked in the late

twentieth and early twenty first centuries.46 Attachment to concepts of liberty and autonomy were

dismantled in the 1990s, as Wilson eloquently explores; relegated by the imperatives of risk avoidance and

security.47 The managerial regime of Tony Blair’s New Labour regarded civic freedoms as an

inconvenience, the removal of which should be of no concern in an enlightened age. They then

accelerated a longer term decline in even an understanding of the case for liberty. Individual liberties were

systematically compromised; rendered of only token value in the context of the fight against international

terrorism, and other threats.

44 Cialdini, Influence, supra note 8. 45 Lansley’s speech (A New Approach to Public Health, supra note 31) makes clear that he shares a similar agenda to his predecessors – from targeting obesity to amplifying the threat of flu epidemic. 46 Jerry Willis and Donna Giles, “Behaviourism in the Twentieth Century: What we Have Here is a Failure to Communicate”, 9 (3) Behaviour Therapy (1978), pp.15–27.

47 Ben Wilson, What Price Liberty? How Freedom was Won and is Being Lost (London: Faber and Faber 2009).


Contemplating society-wide application of behavioural solutions assumes a fundamentally pliant and

passive population that attaches limited value to individual liberty and autonomy. It is inconceivable that

nudging could be contemplated in a more contested and charged environment, such as the Britain of the

1970s, dominated as it was by class conflict. In short, behavioural solutions developed in the laboratory

seem viable when society itself can be envisaged in similar terms. In this sense the Coalition’s nudging

owes a debt to Labour’s elevation of risk and security and diminution of liberty, no matter how much they

might counter-pose their initiative, and identify itself with a libertarian agenda.

There is a wider backdrop here, of the diminishing meaning of privacy in the age of Facebook and

Google, and an ‘actuarial society’ shaped by amoral probabilistic calculation.48 There is also a greater

fluidity of social norms and assumptions in the late modern ‘risk society’.49 This is an individualized

society where each is prone to greater insecurity, and assumption that institutionalised security should be

paramount. Beck’s sense of individualization also points to the fracturing of social norms, making the

assumptions involved in everything from raising children to the etiquette of sexual encounters unclear.50

In such an environment of social uncertainty behavioural solutions can seem both necessary and viable.

Attempting to manipulate patterns of drinking and related problems of unwanted sexual encounters (in

the university context) is one of the first schemes announced by the ‘Nudge unit’, and illustrates an

attempt to influence more fluid contemporary social norms.51

It is in the context of the much blanker contemporary social canvass that we can understand the

surprising confidence of behavioural proponents – surprising when we remember that, for the most part,

they only have experimental results rather than clear, applied successes. Concluding their first report

where they set out some tentative experiments, the nudge unit declare that: ‘There is no reason why we

cannot succeed in tackling today’s rising tide of chronic lifestyle-related disease’.52 Given that unravelling

48 Wilson, What Price Liberty?, supra note 46, pp. 285–338. 49 Ulrich Beck, Risk Society: Towards a New Modernity (London: Sage 1992). 50 Ulrich Beck and Elisabeth Beck-Gernsheim, Individualization (London: Sage 1992). 51 Behavioural Insights Team, Applying Behavioural Insight to Health, supra note 18, p.13. 52 Behavioural Insight Team, Applying Behavioural Insight to Health, supra note 18, p.25.


the relationship between, for example, lifestyle factors and environmental and genetic influences on ill

health remains difficult and contested, such declarations seem unreasonably bold. The relationship

between behaviour and health is a complex one, as is the interface between what is a risk and benefit to

the individual, and society as a whole. The ‘prevention paradox’, for example, tells us that for one

individual to benefit, a far larger number have to change their behaviour—even though they gain no

benefit themselves, or even suffer, from the change.53 More generally, sustainably transforming social

norms through manipulation – even in combination with other approaches – is not a strikingly attainable

objective, yet this remains the temper of current behavioural declaration. With few apparent obstacles to

experimentation and allowed free rein, it seems behaviouralism can become an inflated project with

limited self restraint.

It can be argued that there is a relative underplaying of how behavioural solutions can be reconciled with

issues of liberty and responsibility, issues, we should recall, that are supposedly central to the Coalition

vision. Certainly by implication, behaviouralism challenges the moral autonomy at the heart of modernity.

Contesting its assumptions, one commentator thus called for:

‘...an injection of a bit of the old Enlightenment idea, that humans are autonomous agents who

can shape the world. Indeed, Locke and Mill argued specifically that people would be able to

make choices that others would consider stupid or wrong’.54

More substantially, Bovens highlights behaviouralism’s absence of concern for the development of moral

independence, instead tending to infantilize those for whom it directs choices.55 A lack of concern for

such important matters of principle is partly because ‘nudging’ is inherently technocratic; it is about

53 The paradox was first put forward by statistician Geoffrey Rose, "Strategy of prevention: lessons from cardiovascular disease", 282 British Medical Journal (1981), pp. 1847–51.

54 Alan Miller, “Let’s Banish Nudges and Bans”, Huffington Post, available on the Internet at <http://www.huffingtonpost.com/alan-miller/nudges-and-bans-lets-bani_b_781651.html> (last accessed on 11 January 2012).

55 Luc Bovens, “The Ethics of Nudge”, in T.Yanoff-Grune and S.Hansson (eds.) Preference Change (New York: Springer, 2009), available on the Internet at <http://www.bovens.org/TheEthicsFV.pdf> (last accessed on 11 January 2012).


solutions to assumed problems, and wider issues do not readily figure in its landscape.56 But relevant also

is the overconfidence of an approach whose time has apparently come.

This is not to say that normative issues are ignored. Nudge has a dedicated afterword, engaging objections

and criticisms. Explaining his approach in the House of Lords, Halpern volunteered his concern on the

‘legitimacy of who is making those choices about choices’ at the end of his contribution:

“That’s actually a pretty big deal and a profound one, not least since many of the choices we make

in the moment actually aren’t the ones that we would make on reflection. I guess it’s an argument

that applies in many areas of science, but, boy, it applies in this one. You can’t stray too far from

the legitimacy and the public permission of what you are doing. You already see, actually, some of

the reaction against this early work, that people feel worried about it, and “is it illiberal” and “is

this Orwellian?” Well, at local or national level, if you want to take these kind of approaches,

particularly some of the more controversial ones like priming, you actually just have to have that

public permission. You are going to have to have the discussion, the debate, where the public give

you permission to do the framing around the choices. And if you don’t do that I think you can

get in deep trouble. So you have to answer this agency point both at the individual level but also

at a more collective, reflective level”.57

What is striking, even as Halpern acknowledges that there are issues of legitimacy and agency, is that his

unit was immediately operationalized by the Coalition without pause for reflection, and the contradiction

between a wider libertarian agenda and behaviouralist programme not even recognised, let alone

addressed. Schemes such as to reduce the amount of sex university students may be having are explored

56 See Kyle Powys Whyte and Evan Selinger, “Competence and Trust in Choice Architecture”, 23 (3-4) Knowledge, Technology & Policy (2010), pp. 461–482.

57 House of Lords Select Committee on Science and Technology Inquiry on Behaviour Change Evidence Session 1 (9 November 2010), available on the Internet at <http://www.parliament.uk/documents/lords-committees/science-technology/behaviourchange/ucSTI021110ev1.pdf> (last accessed on 20 December 2010).


without hesitation, suggesting how marginalised concerns for privacy and liberty have become. It would

appear to confirm that a managerial approach to society is firmly established, and manipulating

information for the ‘greater good’ of avoiding risk or improving behaviour unremarkable and assumed to

be unproblematic. In this context it is important to further consider – here at least – some of the

substantive problems.

IV. Problems of transparency and learning responsibility

There are matters of competency and trust in the design of choice architecture.58 Thaler and Sunstein

address such concerns and their ultimate defence of nudging is on the grounds that it’s a process that

inevitably goes on anyway. This is an important observation, as is the fact - more generally with regulation

- that the choice is often between different forms rather than between regulating or not. But it is equally

important to also look further, to who is doing the designing, on what basis and to what effect.59 As

individuals, many of us effectively design our own choice architecture in the knowledge of foibles and

weaknesses; for example I don’t have fattening snacks available in my larder when I’m trying to lose

weight, in order to remove the temptation. This is an informed choice subject to revision, based on my

own priorities and changing knowledge. This example, of weight loss strategies, illustrates well the

necessity for continual monitoring of strategies in the light of evolving knowledge, rather than assuming it

is a matter only of technique and application. There remains uncertainty around how best to lose weight,

complicated by the possible negative effects upon health of carbohydrate-limiting diets.60 Perhaps most

importantly, my nudge is still part of an overall regime of, hopefully improving, self control which is not

neglected or precluded by the resort to manipulating (my own) unconscious biases.

Beyond the individual, there is less choice involved in other forms of nudging undertaken by authorities,

such as redesigning tax forms to prevent common mistakes, or crime prevention campaigns that aim to

58 For a useful philosophical critique, see Whyte and Selinger, Competence and Trust in Choice Architecture, supra note 54. 59 Bovens also makes this point, The Ethics of Nudge, supra note 55, p.12. 60 Recent analysis now tends to suggest that they are at least as effective, probably more so than traditional fat and calorie controlling diets. See for example, Michael L. Dansinger, Joi Augustin Gleason, John L. Griffith et al., “Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized trial”, 293(1) Journal of the American Medical Association (2005), pp. 43–53. Increasing recognition that weight loss may be better achieved through limiting carbohydrates also illustrates the often counter-intuitive nature of biological processes, as becoming less fat is not necessarily best done by eating less fat.


reduce its occurrence. Whilst such schemes are less open and less clearly involve autonomous choice, they

retain a foundation in established procedure and assumption in a democratic society, making compliance

with established rules and regulations easier. Tax must be paid and crime avoided, and nudges to facilitate

these ends have legitimacy. Nudges based upon transparent rules and regulations are ultimately based

upon a democratic consensus that is itself subject to modification. Behavioural manipulation is more

problematic when tied to commercial ends, where society recognizes a need for recourse and even

protection from advertising and marketing that is subliminal or deliberately deceptive. Similar concerns

arise when the state directs us to behavioural change that unreasonably denies choice or agency, whether

through more traditional or more behavioural means. It is because of this problem that the proponents of

nudge wish it to retain a degree of transparency, leaving it to open to, at least, theoretical exposure.

An important part of nudge’s justification is that an element of choice is preserved, unlike with more

heavy-handed regulatory responses such as legal bans. This requires – somewhat awkwardly – that nudges

remain visible rather than subliminal; awkward because externally administered nudges are likely to work

best precisely when they are not transparent; can we really all be ‘in on’ the nudge?61 On the demanding

terms of being both sufficiently masked to bypass too much conscious attention yet still recognizable to

the whistleblower who detects an infringement of liberty, few nudges can be strictly defined as such. This

awkward balance is reflected in the difficult description of being both ‘libertarian’ and ‘paternalist’.

There is another fundamental contradiction in nudging, particularly in the context of the declared intent to

increase a sense of individual responsibility outlined by the Prime Minister. Whilst there may be an

attempt to provide at least token interference transparency to preserve the possibility of exposing a ‘nudge

too far’, this also underlines how far this process is from one that encourages greater learning about

problems and how the individual might take on some responsibility for their management. Behaviouralism

directs us away from building the renewed sense of personal and social responsibility the Coalition

61 Bovens discusses this issue interestingly, suggesting that nudge needs to preserve at least ‘in principle’ rather than ‘actual’ token interference transparency; that we could, at least theoretically, be able ‘to identify the intention of the choice architecture and she could blow the whistle if she judges that the government is overstepping its mandate.’, The Ethics of Nudge, supra note 55, p. 13.


government has set out as fundamental to its mission. Its advantage lies in psychologically ‘accepting us as

we are’, utilising this to produce better outcomes. This same approach is problematic in similar terms, the

flip side being disengagement from the possibility of moral or educational improvement.

We don’t learn much in the behavioural universe. Bovens asks the important question:

“Does it increase our capacity for self control?’ The problem is that – at least theoretically – this is

unlikely; indeed the opposite appears more likely...To warrant long-term success, we should let

people make their own decisions while providing minimal aid. My point is that short-term success

of Nudge may be consistent with long-term failure. The long-term effect of Nudge may be

infantilisation, i.e. decreased responsibility in matters regarding one’s own welfare”.62

What happens when the nudging stops, with a change of government for example? There is a logical case

to say that not only will we be confronted with the problem afresh, but even less capable of dealing with

it, having grown accustomed to it being dealt with by others - with only our, primarily unconscious,

passive involvement. Or perhaps our behaviour will have been successfully reoriented towards the better

outcomes? Actually, biases may be at least partially ‘corrected’ more consciously. Self knowledge about

biases such as to clear many small loans instead of tackling the large one can lead us to try and not do that

anymore. As Bovens points out, there is no proof either way about whether genuine preference change

will result, and it remains a matter for empirical study. At the very least, however, such uncertainty and

absence of evidence suggests the need for greater caution and humility. Yet the opportunity to socially

experiment made by the diminished value attached to liberty, among other factors, has determined an

absence of appropriate restraint.

V. Better outcomes?

62 Bovens, The Ethics of Nudge, supra note 55, p.11.


In traditional democratic terms, we must consider the minority who are unhappy or unwilling to be

nudged into better or healthier choices, and to this end the process must remain one open to scrutiny.

Those who might decide against being directed towards ‘improved outcomes’ may be irrational in

economic terms, but even in economics a more social and contextual view has developed. We now

recognise that ‘happiness’ and ‘wellbeing’ are relatively independent of financial circumstances, suggesting

a wider sense of rationality and fulfilment.63

Behavioural economics, like the discipline more generally, tends toward an abstract, socially-blind sense of

‘irrational’ behaviour that takes no account of values. Yet research suggests, for example, that the success

of American schemes using nudge to reduce electricity usage are dependent upon political values;

unsurprisingly, whilst it works with liberals, it can backfire with some conservatives.64 This is a useful one

in indicating that those unwilling to be nudged might be significant, even a majority. Rejecting ‘better

outcomes’ might also be a less clear cut, political, question of values. Consider our ‘irrational’ bias towards

paying off small loans rather than tackling the larger ones, even though we are likely to pay more interest

as a consequence. Actual people – this author included – choose to limit engagement with financial affairs

on the basis that the gains are small compared to the cost of activity which we find particularly tedious. In

fact, we must all do this to some degree as there is always more energy that can be spent saving money -

always more research that could be done into switching loans and reorganising finances - so a decision

remains to be made about the point at which we stop and the balance against other aspects of our lives. In

other words, it may not be (financially) rational if we were to assume an actor focused upon maximizing

returns, but it can well be functional on the basis of particular beliefs, choices and options. Alternatively, we

can say that economics tends to work on the basis of an implausibly narrow conception of what is

rational, or not.

63 Laura Camfield, Gina Crivello and Martin Woodhead, “Wellbeing Research in Developing Countries: Reviewing the Role of Qualitative Methods,” 90(1) Social Indicators Research (2009), pp. 5–31; Bruno Frey, Happiness: A Revolution in Economics, (Cambridge, MA: MIT Press, 2008).

64 Costa and Kahn, Energy Conservation “Nudges” and Environmentalist Ideology, supra note 7.


Even in cases of relatively unambiguously better outcomes, these remain relative and implementation

requires careful weighing up on a longer time frame. All policies and actions have unintended

consequences.65 Nowhere is this a more important recognition than when thinking through risks and

benefits. Risk management is a process of complex negotiation of trade-offs and exchange.66 Risks migrate

from one site to another rather than simply disappear through policy or design.67 Consider a well

intentioned behavioural changes brought about, in this case, by legal sanction. Compulsory cycle helmet

wearing leads to a fall in the numbers of children in particular, who cycle, as it becomes a regulated rather

than more spontaneous experience.68 Initiatives can have counter intuitive effects such as how doing ‘our

little bit’ for the environment by recycling might curtail any impulse to take more significant action.

Meanwhile, any relationship of the domestic recycling ‘ritual’ to slowing, let alone stopping global

warming is, to say the least, uncertain.69 Such complexity can all too easily be ignored in behavioural

scheming focused on short term solutions.

Matters of expanded public health and its relationship to individual choice are among the most complex in

modern society. The idea of putting fluoride into the water supply, from the 1940s, to protect the teeth of

those who might not do so themselves might be considered an early behavioural nudge. But it has rightly

provoked half a decade of debate, as have more recent initiatives such as putting folic acid into flour to

reduce the number of babies born with spina bifida and neural tube defects.70 ‘Better outcomes’ –

65 Robert Merton, The Unanticipated Consequences of Purposive Social Action, 1(6) American Sociological Review (1936), at pp. 894–904.

66 John D. Graham and Jonathan Wiener, Risk vs Risk: Trade-offs in Protecting Health and the Environment (Cambridge MA: Harvard University Press 1995); W. Kip Viscusi, Fatal Tradeoffs: Public and Private Responsibilities for Risk (New York, NY: Oxford University Press, 1992).

67 Ruth Alcock and Jerry Busby, “Risk migration and scientific advance: the case of flame-retardant compounds”, 26(2) Risk Analysis (2006), pp. 369–381.

68 The most comprehensive study into the effects comes from Australia. For example, data from Western Australia shows the number of Australian children walking or riding a bicycle to school plunged from about 80 % in 1977 to the current level around 5 %. This fall directly coincided with the introduction of compulsory helmet wearing. See the compilation of research, Bike numbers in Western Australia, available on the Internet at <http://www.cycle-helmets.com/bicycle_numbers.html> (last accessed on 11 January 2012).

69 Sarah Moore and Adam Burgess, “Risk rituals?”, 14 Journal of Risk Research (2011), pp. 111–124. 70 R. Allan Freeze and Jay Lehr, The Fluoride Wars: How a Modest Public Health Measure Became America’s Longest Running Political Melodrama (London: Wiley, 2009). A David Smith, Young-In Kim and Helga Refsum, Is Folic Acid Good For Everyone? 87(3) American Journal of Clinical Nutrition (2008), pp. 517–533.


particularly in the complex world of health and human behaviour – are far from given, particularly

concerning untargeted interventions such as these. Where better outcomes are largely incontestable,

nudging only makes sense where the problem is clearly a behavioural rather than one of resources or

technology, leaving only quite particular foci for intervention. Increasing usage of oral rehydration salts to

treat diarrhoea, particularly in India, is one important example, having a potentially significant effect on

the still shocking levels of infant mortality from this treatable condition. The technological problem has

been fundamentally solved, and nor is it a basic question of resources, as salts are widely available. The

problem is to behaviourally encourage mothers to remember to medicate. Diabetes is a disease which can

be effectively treated but requires vigilant self management, and it no surprise that behavioural solutions

are also the focus of extensive research and initiative, including the UK ‘nudge unit’.71 At the same time

such examples may be more an exception than a rule; diseases which are agreed, significant and stubborn

problems with a simple solution but requiring some applied thinking. We should add that even in these

examples there is rarely such a thing as an exclusively behavioural issue, however. In more traditional

terms the problem in India also remains one of education, with many women still fundamentally

misunderstanding how diarrhoea should be treated and not recognizing the life-saving potential of the

salts.72

VI. The continuity of lifestyle risk intervention

Despite the relative absence of substantive contesting of nudging, its implementation by the Coalition has

stimulated, largely negative, commentary. One liberal columnist typically complained that: ‘Cameron’s

hijacking of Nudge theory is a classic example of how big ideas get corrupted’, affirming the sense that it

is the application rather than substance that is the subject of criticism.73 A withering tone was captured by

71 See the Behavioral Diabetes Institute, for example, available on the Internet at <http://behavioraldiabetesinstitute.org/about-Behavioral-Diabetes-Institute.html> (last accessed on 11 January 2012).

72 See Sendhil Mullainathan’s talk on these issues, available on the Internet at <http://www.ted.com/talks/sendhil_mullainathan.html> (last accessed on 11 January 2012).

73 Adita Chakkraborty, “Brain Food”, The Guardian G2, 7 December 2010, at p.5.


the title of another: ‘Oh, Mr Cameron, do stop all that annoying nudging.’74 Again, this article did not

contest the substantive, interventionist aim, however; indeed it complained that nudging will be

insufficient, as we: ‘can’t rule out a surge in obesity, hyperactivity or mass poisoning’. The problem,

particularly for critics of the liberal left, is that corporate players like food manufacturers will not now be

frozen out of regulation as they tended to be under Labour, but will also play their part in improving

health lifestyles. Those closely aligned with health promotion initiatives write dramatically of a complete

reversal in public health because industry will no longer be sidelined.75 The polarised politics of risk

consolidated under the previous government involved an at least rhetorical hostility towards corporate

influence, and this has left a considerable legacy. Other critics have attacked the British application of

nudging on the grounds that it wrongly precludes more regulation and legislation to promote health.76

Hostility towards nudge is as pronounced from the anti-regulatory right, as it is from the left. For

example, influential conservative commentator Melanie Phillips writes of the ‘happy mind benders of

No.10’77 Criticism has been such that ministers have responded in public.78

To an extent, this is a predictable problem for an approach that deliberately steers itself between left and

right. The result can be to satisfy neither side, whilst aggravating both. But the more particular aspect is

that nudge is being adopted at a time of austerity, by a Conservative-led government. Both left and right

complain that their cherished services are being cut dramatically; for the former, public services and jobs

are the focus, whilst for the right it is cuts to the defence and policing budgets that rankle. Government

experiments involving American-imported ideas, in this context, can become the focus of resentment and

ridicule. For those angered already by funding cuts, Conservative initiatives are part of an ideological

74 Catherine Bennett, “Oh, Mr Cameron, do stop all that annoying nudging,” The Observer, 5 December 2010, at p.43. 75 See, for example, Joe Millward, ‘Letting the food industry shape policy will ruin a century of progress’, The Guardian, 3 December 2010, at p.17.

76 Chris Bonell, Martin McKee, Adam Fletcher et al., “Nudge Smudge: UK Government Misrepresents ‘nudge”’, 377 The Lancet (2011), early online publication, 17 January 2011, available on the Internet at <http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(11)60063-9/fulltext> (last accessed 11 January 2012).

77 Melanie Philips, “The happy mind-benders at No.10,” Daily Mail, 29 November 2010, available on the Internet at <http://www.melaniephillips.com/the-happy-mind-benders-of-number-ten?searched=mind-benders&advsearch=oneword&highlight=ajaxSearch_highlight+ajaxSearch_highlight1> (last accessed 11 January 2012). 78 Francis Maude, The nudge is no fudge, The Guardian, 27 December 2010, available on the internet at

<http://www.guardian.co.uk/commentisfree/2010/dec/27/nudge-fudge-community-level> (last accessed 11 January 2012).


offensive to ‘roll back’ the state and make ordinary citizens and initiative pay for services themselves. The

background is here is a longstanding popular mistrust of Conservative policies as being dishonest and

hypocritical, and memories of cuts under the infamous administration of Margaret Thatcher and her

successors. There is a suspicion that the rhetoric of freedom and responsibility is only a cover for market

domination, and nudge can be seen as the latest attempt to justify reductions in state spending. This is in a

context where Coalition policy as a whole is regarded by critics as driven by an ideological hostility

towards state spending (on the poor) – rather than determined by the dire economic circumstances of the

UK economy, married with a determination to shift power towards the local and individual.79 Cameron

has assiduously sought to distance himself from traditional conservatism through adopting new ideas such

as nudge. But the imperatives of recession have confounded both his distancing from traditional

conservatism, and nudge’s independence of either left or right. In this respect, the recessionary circumstances

that have helped thrust nudging to the fore have also led to the compromising of it as a policy approach, as behavioural

economics becomes not so much ‘smart’ policy as austerity policy. Experimental conditions for

behaviouralism have turned out to be less ideal than might first appear.

Finally, it is important to recognise that the grounds upon which nudge policy making in the UK is not

being criticised; the appropriateness of government attempts to shape health lifestyle behaviours, or that this

is a self evident ‘good’. There are objections to partnership with industry in schemes to encourage ‘healthy

behaviour’, but concern is that the worthy objective of improving people’s lifestyles might be

compromised by association with commercial actors. There is little criticism of objectives themselves and

in this respect we can identify a triumph of lifestyle health interventionism. At the same time it is

significant that nudging initiatives are, thus far, concentrated primarily in this domain, precisely because

health improvement now stands as a kind of universal imperative. A rare critic from inside the medical

79 The impulse behind Coalition policy remains a debateable question. But in my view, critics tend to act as if financial restrictions simply do not exist – yet they are plainly real. Further, an example like the transformation of higher education suggests that the ideological determination to shift the focus of power is predominant. The changes in higher education are likely to cost the state more in the long term, as a large proportion of the new student loans will not be paid back. Meanwhile, the immediate consequence is to take resources from university teaching budgets and, at least theoretically, place greater power in the hands of the individual student ‘consumer’.


profession terms this the ‘tyranny of health’.80 By contrast, there is little sign of schemes in more difficult

areas where there is no such consensus.

Critical reflection allows an appreciation of the lifestyle health consensus as a distinctive, even curious

development. Basic public health has long been established. We thankfully live in an era of, remarkably,

still ever-increasing life expectancy and, at least in the developed world, the eradication of infectious

disease. Attention has shifted towards the much more uncertain world of lifestyle risk, but often as if

problems and solutions in how people choose to live their (unhealthy) lives are matters of will and

resources like public health in the past, and can be easily framed around unambiguous choices. Yet whilst

in the world of health promotion and experimental nudging the choice is a simple one between, say, the

‘good’ fruit juice and ‘bad’ fizzy drink, in reality it is not nearly so clear cut. The ‘natural’ sugar in fruit juice

will rot teeth as surely as the ‘unnatural’ in fizzy drinks, and the choice of one over the other is essentially

a lifestyle choice marking distinctions of taste and class.81 It is misleading to distinguish consumption

around ‘goods’ and ‘bads’, as diet is fundamentally a matter of balance, based around the often-confirmed

maxim that ‘all things are best in moderation’. More useful is to consider how health has become

intimately connected to morality; more a modern way of denoting ‘responsible’ behaviour than simply

reflecting upon the state of bodily functions.82 We might like our children to be more active and walk to

school, for example, but there is no clear purpose to making children ‘fitter’, and it is clear that exercise

has only the most limited impact upon weight loss – should that be the objective.

Matters become even more complex when we consider the impact of actual programmes intended to

improve health, particularly those ‘preventative’ schemes intended to identify illness in its early stages.

These are fraught with unintended consequences, such as the false positives and unnecessary treatment

that accompany the ‘good’ of mass screening programmes.83 Such problems are not resolved simply by

nudging people into more regular screening or undertaking self examinations. Meanwhile, it is, at least

80 Michael Fitzpatrick, The Tyranny of Health: Doctors and the Regulation of Lifestyle (London: Routledge, 2000). 81 For thinking socially about matters of taste, see Pierre Bourdieu, Distinction: A Social Critique of the Judgement of Taste (Cambridge MA: Harvard University Press, 1987). 82 Allan Brandt and Paul Rozin, Morality and Health: Interdisciplinary Perspectives, (London: Routledge, 1997). 83 Angela Raffle and J.A. Muir Gray, Screening, Evidence and Practice, (Oxford: Oxford University Press, 2007).


historically, uncontentious to suggest that the singular pursuit of health in its own terms and for its own

sake is not necessarily a ‘healthy’ one.84 Among the consequences are the ‘worried well’, and how the

intense scrutiny of health and consumption provides an environment conducive to health alarms.85 It

would seem to be no coincidence that the UK has witnessed so many ‘panics’ around food since the

1980s, around everything from eggs, to cheese, to meat, as the elevation of health – and risks to it – has

been particularly clear and political.86

It was the previous Conservative administration that began the politicisation of lifestyle health

subsequently consolidated under Labour. It was their ‘Health of the Nation’ White Paper back in 1992 -

with its 27 targets on issues from teenage pregnancy to taking more exercise - that signalled the new,

expanded health agenda. It was unveiled by the then Health Secretary:

Honourable Members will know that people have become more conscious about what they eat,

how much exercise they take and how they can generally improve the quality of their lives by

becoming healthier. We wish to build on that healthy trend. No responsible Government can be a

disinterested observer of an unhealthy nation. We should prevent illness wherever we

can...Prevention is better than cure.87

Food health alarms encouraged by a climate of health politicisation began under the Conservatives also,

following another health minister’s declaration that ‘all eggs have salmonella’ in 1988.88 More dramatic and

widely targeted health initiatives also began in the late 1980s, with the campaign to alarm the whole

population into believing that they were all at risk from AIDS.89 Likewise, more media-driven risk politics

was initiated during this period, with the passing of the Dangerous Dogs Act in 1991, which banned

84 Maintaining ‘good health’ in the past was not generally regarded as an end in itself – as tends to be the case today – but simply as the bodily prerequisite for moral, intellectual and practical achievement. Those unreasonably concerned were known as ‘hypochondriacs’, a term which also indicates the necessary connection with unending concern about even the most minor possible threats to health. 85 Fitzpatrick, The Tyranny of Health, supra note 79. 86 Christopher Booker and Richard North, Scared to Death (London: Continuum, 2009).

87 Virginia Bottomley, “Health of the Nation”, Hansard, 8 July 1992, Vol. 211, cc.335, available on the Internet at <http://hansard.millbanksystems.com/commons/1992/jul/08/health-of-the-nation> (last accessed 11 January 2012). 88 Booker and North, Scared to Death, supra cite 85. 89 See Burgess, The Politics of Health Risk Promotion, supra note 34, p. 537.


certain breeds of dog, following a single attack on a schoolgirl.90 As the Conservatives are now innovating

the politics of nudge, it was they also who, earlier, innovated those of risk and health ‘nannying’.

Nudge is less of an alternative than is being claimed to the ‘nannying’ of the past, particularly as it enters

an environment where lifestyle health interventionism is already well established as the norm. In this

context ‘nudge’ is only the latest addition to the portfolio of interventionist approaches, rather than an

alternative to it. Any libertarian principle is subsumed; for example in the government’s new policy

proposals on health which state, at one point, that: ‘The Government will take a less intrusive approach,

staying out of people’s everyday lives wherever possible.’91 More importantly, the document and policy are

then structured around the Nuffield Council of Bioethics Ladder of Interventions, which ‘increases in

intrusiveness’ from ‘do nothing’ to ‘eliminate choice altogether’. All behaviours seem open to intervention

in these terms, with no sense of limits or boundaries. There are here only some ‘minor’ behaviours trends

that can be left alone as they may ‘fizzle out’, and these are only one out of 8 options. In practice, a range

of new initiatives targeting lifestyle risk have been unveiled that now involve corporate as well as public

health actors.92

Nudge has been adopted in the UK in an uncritical and wholehearted way, to the exclusion, for example,

of valuable work on better regulation that better engaged with the damaging politics of risk in the

contemporary UK.93 Unveiled at the same time as the Coalition’s wider programme of change, its

attraction lay in offsetting the sense of austerity for austerity’s sake. But in the context of severe economic

cuts and widespread reaction against them, the plans to nudge behaviour have been drowned out by

criticism. Meanwhile, more substantive problems with behavioural solutions remain uncontested. The

technocratic assumption of ‘better outcomes’ ignores complex realities, and nudging crucially limits the

possibility of conscious learning and improvement. But it is not only nudging in its various forms that

90 See Johnston, Bad Laws, supra cite 34, p.62. 91 Department of Health, Healthy Lives, Healthy People: Our Strategy for Public Health in England, (London: HM Government, 2010). 92 Behavioural Insight Team, Applying Behavioural Insight, supra note 18. 93 See, for example, BRC, Whose Responsibility, supra note 33. Risk, Regulation Advisory Council, Response with Responsibility: Policy Making for Public Risk in the 21st Century, (London: BIS, 2009).


requires examination and debate, but the wisdom and consequences of the fixation on lifestyle health

issues and where the limits to all forms of direct external interference lie.


Liberalism and Lifestyle: Informing Regulatory Governance with Behavioural Research

On Amir* and Orly Lobel**

I. New Governance meets behavioural economics

Behavioural economics is helping illuminate the limits of rational individual choice. At the same time, even

as research identifies failures in rationality, policy must inquire about the possibility and legitimacy of

government intervention. In ‘Nudging’ Healthy Lifestyles: The UK Experiments with the Behavioural Alternative to

Regulation and the Market, Adam Burgess critically describes the introduction of behavioural approaches

into UK policy making. In particular, Burgess is concerned with the wholesale adoption of nudge-style

programmes to promote healthier living among citizens. Unsurprisingly, the UK developments find

equivalent developments in the United States. In January 2009, President Barack Obama suggested that

“the principles governing regulation . . . be revisited.” Envisioning a ‘behavioural dream team’ of

economists and psychologists to help lead the way, President Obama appointed Cass Sunstein, the co-

author of Nudge, as Administrator of the White House Office of Information and Regulatory Affairs.94

President Obama further ordered the US Office of Management and Budget to “clarify the role of the

behavioral sciences in formulating regulatory policy.” The parallels between the countries are striking. The

new UK leadership has consulted with Nudge’s other co-author, Richard Thaler. The new leadership has

guided top administrators to read behavioural research and a “Behavioural Insight Team” has been newly

formed within the Prime Minister’s Cabinet Office. More operationally, each of the governments has

adopted initiatives to nudge citizens to better behaviour, physically and financially.

In their rhetoric and in their newly adopted programmes, the two governments are signalling a preference

for interventions that move away from the traditional approaches of command-and-control. In doing so,

they have announced the importance of importing lessons from the field of behavioural economics to the

field of regulation. The interest in behavioural sciences by the realm of regulation and policy is widespread

and global. These developments should be understood not only as a response to the growing prominence

of behavioural economics in social science research but also as part of a rising interest by regulators in

public policy innovation. Both in practice and in scholarly inquiry, new approaches to regulation, often

collectively referred to as “New Governance,” have become a central field of experimentation.

* Associate Professor of Marketing, Rady School of Management – UC San Diego. ** Herzog Endowed Scholar and Professor of Law, University of San Diego – School of Law. 94 Michael Grunwald, “How Obama Is Using the Science of Change”, Time, 13 April 2009, at p. 28.


The starting point of New Governance theory is different from that of behavioural law and economics.

While the latter body of literature, behavioural research, begins with the challenges of individual choice,

New Governance begins with the challenges of traditional regulation, from the perspective of both

legitimacy and effectiveness. New Governance scholarship views traditional regulation, command-and-

control techniques, as a too narrow and incomplete policy framework to accomplish social goals. New

Governance scholars further challenge traditional regulation as neglectful of the active role private

individuals and organizations can and must assume in achieving these goals. In calling for more private

and individual responsibility in the implementation of policy, the two schools of thought, behavioural law

and economics and New Governance, converge on a set of understandings: private ordering and public

action are both central to policy, they operate in conjunction, and each can benefit from a better

understanding of judgment and decision-making.

Regulators around the world are attempting to reconcile the changing demands of 21st century economies

with the insights these two schools of thought, New Governance and Behavioural Economics, are

offering. The common emerging theme, expressed by the UK and American administrations as well as in

other democracies around the world, is that instead of focusing solely on substantive prohibitions and

adversarial enforcement, regulation should centre on designing processes and systems through which

private ordering and individual decision-making are directed and improved. Jason Solomon has described

New Governance approaches in this way:

Think about new governance regulation as an umbrella term covering a kind of interaction

between the state, regulated entities, and other stakeholders that has a number of desiderata--

public participation, data provision, transparency, benchmarking, sharing of best practices, fora

for deliberation on ends and means, and autonomy and flexibility for those subject to regulation.

Autonomy and flexibility for those subject to regulation, as Solomon describes, leads to the idea of choice

architecture, nudges, and default design, major tools within the New Governance regulatory toolbox.

Other tools include requiring private actors to give reasons, to set their own reflexive processes, and to

self-monitor. New Governance encourages collaborative spaces of engagement and the adoption of

preventative measures such as training, data collection, and continuous learning. These approaches to

regulation expand the ways we think about the work of governments and the relations between the public

and private sphere. The shifts are certainly political. In the United States, since the mid-1990s, Democrats

and Republicans alike seem to agree that “[t]he era of big government is over.” As with the underlying

politics in the UK enthusiasm around regulatory reforms that Burgess describes, the rhetoric in the United

States is frequently that of “neither left nor right,” but rather a call for more effective and more legitimate

policy more broadly. Indeed, as Burgess points out, the UK initiatives were originally commissioned not

by the current conservative government but by the previous Labour government. Thaler and Sunstein


themselves have marketed their vision of nudge as holding an appeal to “conservatives, moderates,

liberals, self-identified libertarians, and many others.”95

In its ideal, social science research is apolitical. Studies in psychology and economics, whether empirical or

theoretical, are neither progressive nor conservative. They are meant to improve our understandings of

how society operates. As we have discussed elsewhere, New Governance and the push for choice

architecture can be seen as challenging the left/right debates on regulation and de-regulation. But Burgess

argues that behavioural studies have been enthusiastically adopted by a conservative government because

of their political appeal. Burgess’ concerns with the behavioural enthusiasm are multiple: the concern that

the shift away from traditional regulation is in fact a shift away from regulation and public action to an era

of devolution and deregulation; the concern that the behavioural interventions that are adopted are

paternalistic in their intent to direct, or, more disturbingly, manipulate, the lifestyles of private individuals;

and the concern that we know too little both about what consists of healthy lifestyles and about the

psychology of decision-making for us to accept programmatic policies based on behavioural nudges in

field of health policy. In the following sections, we shall address each of these concerns.

II. New Governance does not imply deregulation

Burgess describes a worrisome dynamic in which enthusiasm about adopting behavioural nudges as a

mode of regulation is inextricably linked with cost cutting in government expenditure, as well as initiatives

of devolution to local communities. Burgess is correct that debates about the proper scope and form of

regulation are always intertwined with the politics of public expenditure. Yet it is important to emphasize

that a shift from traditional regulation to regulatory governance does not necessarily mean the reduction

of public action. Neoclassical economics, while generally averse to regulation, allows that under certain

conditions of imperfect markets, regulation may be an appropriate response. The insights of studies in

social psychology, rationality and behavioural economics enormously expand the world of imperfect

markets. Once one accepts that many of our decisions are based on inaccurate calculations and distorted

perception of expected benefits and costs, then it becomes clear that market failure is pervasive. This in

fact can mean more market interventions and greater government action, not less.

95 Richard H. Thaler and Cass R. Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (New Haven, CT: Yale University Press 2008), at p. 248.


In Europe, the term “soft law” has been used as shorthand for a myriad of regulatory innovations as well

as a shift from regulation to a reliance on private ordering. In the scholarly network of New Governance,

commentators frequently warn against the equation of governance approaches with merely voluntary

guidance or a shift to private and local action. As we have argued elsewhere, the newness of the new

governance model (or the innovative aspect of nudge-style directives) is not the existence of soft aspects of

law, but rather their recognition.96 In a series of articles, we explore the potential of a growing body of

regulatory tools that can be grouped under the label “new governance” and fall between a sole reliance

command-and-control regulation and a one-directional shift to deregulated markets.97 Regulation always

operates side by side with the background of markets and social interactions. The harnessing of

behavioural insights into the world of policy can therefore be conceptualized as the formalization of

informal and private practices that have always shaped policy. In fact, disclosure laws and other types of

New Governance initiatives can be more costly than their alternative command-and-control regulations.

For example, in the context of occupational safety systems, we have argued that an effective approach

must be the expansion of nudge-style approaches, requiring individual learning, internal monitoring and

reporting, and systems that make risk and remedy more salient, alongside the augmentation of targeted

enforcement of traditional regulatory rules.98 While debates about regulating occupational risk have largely

been framed as a duality between top-down standards versus market-based approaches, including

premium pay and information disclosure, behavioural lessons suggest that collaborative efforts with

regulated parties, through the ordering of self-monitoring processes, reporting systems and public granting

of carrots for beyond-compliance performance, can all have positive effects. At the same time, the smarter

use of nudge style incentives can free up public resources for the necessary continuous application of top-

down traditional regulation in cases where safety is habitually neglected and cooperative approaches have

failed. Burgess worries that “the recessionary circumstances that have helped thrust nudging to the fore

have also led to the compromising of it as a policy approach, as behavioural economics becomes not so

much ‘smart’ policy as austerity policy.” And yet, limited resources in the work of regulators are an

ongoing reality that any regulatory vision must cope with. The right questions to ask are not how can we

move away from regulation to nudges but rather: when are cooperative versus coercive styles of regulation

most effective? What is the relationship between collaborative approaches and punitive enforcement?

How can multiple approaches co-exist and dynamically relate to each other forming a comprehensive

96 Orly Lobel, “Setting the Agenda for New Governance Research”, 89 Minnesota Law Review (2004), pp. 498 et sqq., at p. 506.

97 Orly Lobel, “The Renew Deal: The Fall of Regulation and the Rise of Governance in Contemporary Legal Thought”, 89

Minnesota Law Review (2004), pp. 342 et sqq.; Orly Lobel, “Citizenship, Organizational Citizenship, and the Laws of Overlapping

Obligations”, 97 California Law Review (2009), pp. 433 et sqq.; Orly Lobel, “The Paradox of ‘Extra-Legal’ Activism: Critical Legal

Consciousness and Transformative Politics”, 120 Harvard Law Review (2007), pp. 937 et sqq.; Orly Lobel, “Interlocking Regulatory

and Industrial Relations: The Governance of Worker Safety”, 57 Administrative Law Review (2005) 57, pp. 1071 et sqq.; Lobel,

“Setting the Agenda”, supra note 96.

98 Orly Lobel, “Governing Occupational Safety in the United States”, in Grainne Dr Burca and Joanne Scott (eds), Law and New

Governance in the EU and the US (Oxford and Portland: Hart, 2006), pp. 269 et sqq.


policy system? What are the costs of programmes, whether they employ shoves or nudges, relative to their

expected benefits? The answers will be context-specific and they will almost always be an interdependent

blend between top-down regulation, nudges, and market ordering.

III. Lifestyle, the shift to prevention and the question of paternalism

When in 2010 the UK government rejected “the assumption that central government can only change

people’s behaviour through rules and regulations” it was pointing to the third way possibilities of directing

individual behaviour through regulatory nudge.99 Burgess describes a 2010 shift in the UK political

direction, from one that focused on “risk avoidance” to one that focuses on personal responsibility. The

UK government expressed an aspiration to be “a much smarter government” -- one that shuns “the

bureaucratic levers of the past” and finds “intelligent ways to encourage support and enable people to

make better choices for themselves.”100 In the process, the new government has initiated a programme to

nudge individuals to make better choices regarding their lifestyles and health. In the U.S., a similar

language of ridding red tape has become prevalent and has similarly shifted efforts to directing healthier,

preventative choices. The Obama Administration, with the energies of First Lady Michelle Obama, has

focused efforts on improving child nutrition, healthy eating, anti-obesity and active lifestyles. In 2010,

Michelle Obama announced a Let's Move! campaign to end childhood obesity within a generation. In

England, like in other countries, the recent motto has become that “prevention is better than cure.”101

Worldwide, our conception of health has evolved and in turn, so has the framework of public health

policy. The World Health Organization (WHO) long defines ‘health’ as “a state of complete physical,

mental, and social well-being and not merely the absence of disease or infirmity.”102 This broad definition

is increasingly making its way into policy, which attempts to address not only cures for diseases but also

assist and improve “physical, mental, and social wellbeing.” Just as it is relatively uncontroversial for

liberal democracies to regulate the sales and consumptions of tobacco (though it was highly controversial

99 David Cameron and Nick Clegg, “Foreward”, in The Coalition: Our Programme for Government (London: HM Government, 2010),

at p.7, available on the Internet at

<http://www.direct.gov.uk/prod_consum_dg/groups/dg_digitalassets/@dg/@en/documents/digitalasset/dg_187876.pdf>

(last accessed on 4 January 2012). 100 Ibid.

101 Adam Burgess, “‘Nudging’ Healthy Lifestyles: The UK Experiments with the Behavioural Alternative to Regulation and the

Market”, 3 European Journal of Risk Regulation (2012), this issue, citing Virginia Bottomley, “Health of the Nation”, Hansard, 8 July

1992, Vol. 211, cc.335, available on the Internet at <http://hansard.millbanksystems.com/commons/1992/jul/08/health-of-the-

nation> (last accessed on 4 January 2012).

102 Preamble to the Constitution of the World Health Organization as adopted by the

International Health Conference, New York, 19 June to 22 July 1946, 2 Official Records of the World Health Organization, at p. 100.


a few decades ago), policymakers today are justifiably asking questions about the health effects of other

lifestyle habits and consumption, including food, drinks, and drugs.

The goal of helping people make reasonable decisions about their health, risks, and environment

is a key aspect of public health policy. Perhaps more than any other policy field, health policy faces the

challenge of assisting individuals when grappling with large quantities of data with the goal of rationally

processing information pertaining to risk. Health decision-making as we understand it today largely rests

on the significance of individual responsibility for one’s decisions and choices. While in the past, some

health traditions supported a hierarchy of doctor-patient paternalism, today’s health world is based on the

widespread ethical principle of patient autonomy and informed consent.103 Consequently, perhaps more

than with some policy fields (e.g., transportation safety), health policy cannot be simply about directing

healthy behaviour but must aim for an understanding of how individuals reason and decide. For example,

rather than prohibiting liability waivers, opt-out options and choice of treatments for patients,

policymakers must ensure that that such waivers are administered and introduced in ways that ensure truly

informed consent.

Given this framework of choice and autonomy, behavioural insights can help policy improve

individual decision-making processes as well as identify limits of the corrective solutions to cognitive

failures. Behavioural research is providing new evidence that cognitive processes affect decision-making,

often leading to suboptimal conclusions. Despite the significance of this field to health policy, research in

behavioural decision making as it pertains to health and medicine is still scarce. Along with collaborators,

we have been researching decision-making and the processing of risk in the context of health-related

choices with the aim to inform public health law and policy.104 At the basis of our study are

understandings about the ability of regulatory design and policy to improve health and welfare through a

myriad of policy tools, including regulation that moves beyond traditional command-and-control and

focuses on facilitating individual judgment and choice. We seek to identify a series of mechanisms through

which individuals assess and act upon information pertaining to their immediate health and future healthy

living and to propose ways in which the law can support healthier actions. Using new experimental

techniques of impairing executive control resources, our studies investigate the interactions between

various mechanisms that affect decisions about healthy choices and health risks and the level of

processing decisions: emotion/cognition and immediate reactions/careful conscious evaluation. While we

are convinced that this line of research is highly relevant to policy, we also believe it is important to point

103 Patient Self-Determination Act, Pub. L. No. 101-508, Title IV, §§ 4206, 4751, 104 Stat. 1388-115, 1388-204 (1990).

104 Orly Lobel and On Amir, “Healthy Choices: Regulatory Design and Processing Modes of Health Decisions”, 1 July 2011,

available on the Internet at <http://ssrn.com/abstract=1876734 > (last accessed on 4 January 2012).


to the range of health challenges that cannot be addressed by New Governance approaches but rather will

continue to require more traditional regulatory approaches.

Individuals frequently make poor decisions about their health management. A large number of

recent studies show that the manner in which individuals appraise choice alternatives has an immense

influence on the outcome. Moreover, research indicates that context, framing, and surrounding signals

heavily influence choices among alternatives.105 The influence of these factors is even more pronounced in

inter-temporal decision-making and situations of risk and uncertainty, both of which are central to health

policy. For health regulators, it is of key significance to understand how individuals make decisions when

contemplating probabilities of future and long-term beneficial or detrimental outcomes. In one early

experiment by Nobel Laureate Daniel Kahneman and his collaborators, when patients were told, “Of

those who have this procedure, ninety percent are alive after five years,” they were far more likely to agree

to the procedure than when they are told, “of those who have this procedure, ten percent are dead after

five years.”106 Or take another example, highlighting our action/inaction biases. In an experimental study,

Ritov and Baron found that individuals may be reluctant to vaccinate a child when the vaccination itself

can cause death, even when this is much less likely than death from the disease prevented. DTP vaccine

cases present such a classic challenge for health policy and individual decision-making. In rare instances,

DTP may cause permanent neurological injury (1 dose out of about 300,000). This risk is far less than the

neurological injury risks from the illnesses that the vaccine prevents. In Ritov and Baron’s study, subjects

were presented with a situation in which a disease kills 10 out of 10,000 children. They informed

participants that while an available vaccine can prevent the disease in everyone, the vaccine has side effects

that kill some children. Subjects were given a table of different possible values of the risk of death from

side effects and subjects were asked the maximum level of risk that should be tolerated by government in

order to institute a compulsory vaccination programme. Risky decisions have been studied in many social

science and policy fields, yet much about the mechanisms of risk-related preferences is still to be

uncovered. For example, processing modes of risk, the state of one’s cognitive resources at the moment of

decision, have been relatively understudied.107

105 For these effects in other contexts outside the health policy arena, see On Amir, Dan Ariely and Ziv Carmon, “The

Dissociation Between Monetary Assessment and Predicted Utility”, 27 Marketing Science (2008), pp. 1055 et sqq.

106 Donald A. Redelmeier, Paul Rozin and Daniel Kahneman, “Understanding Patients’ Decisions: Cognitive and Emotional

Perspectives”, 270 Journal of the American Medical Association (1993), pp. 72 et sqq., at p. 73. 107 On Amir, Anastasiya Pocheptsova, Ravi Dhar and Roy F. Baumeister, “Deciding Without Resources: Psychological Depletion

and Choice in Context”, 46 Journal of Marketing Research (2009), pp. 344 et sqq.; Cindy Brice, George Loewenstein, Robert Arnold et

al., “Quality of Death: Assessing the Importance on end-of-life treatment in the intensive-care unit”, 42 Medical Care (2004), pp.

423 et sqq.; Lobel and Amir, “Healthy Choices”, supra note 104.


Burgess’ commentary on incorporating such behavioural insights into health policy reveals an embedded

tension. On the one hand, Burgess is worried about the shrinking of the state and the under-investment of

public resources in social problems, and, on the other, about over-“nannying”, as he and others refer to

what is perceived as paternalistic state interventions in individual choice. We believe the tension stems

from an inaccurate notion of the work of government and the role of behavioural economics in informing

this work. The tension likely also stems from the misleading term “libertarian paternalism” of which we

are critical. The distinction Burgess draws between tax and crime prevention on the one hand and other

types of state intervention on the other is misguided. Tax policy and criminal law are infused with

numerous value-driven judgments. It is no more neutral or obvious in the work of government than other

policy fields in which government must operate, including financial, health and environmental regulation.

Elsewhere, we have argued that Nudge’s concept of “libertarian paternalism” both understates and

exaggerates the jurisprudential and policy implications of behavioural economics to regulatory design.108

The term “nudge” was intended to suggest a contrast with the heavy-handed “shoves” of command and

control regulation. We believe that the distinction between gentle nudges and forceful shoves is an

important distinction for liberal democracies. We agree with Thaler and Sunstein’s notion that, under

certain conditions, regulation can be simultaneously more effective and less interventionist than traditional

command-and-control approaches. It can also be more transformational and less costly. We also caution

however against over-claiming. In fact, we view the term libertarianism as problematic precisely because it

seems to deny the inherently public aspects of regulation by design. Put differently, we view the term

“libertarian paternalism” as the wrong heuristic for the regulatory innovation of New Governance and

nudge-style experimentation. In an environment of pervasive scepticism about traditional regulation, the

term “libertarian paternalism” is a strategic choice, conveying the message that interventions are non-

intrusive while effective. And yet, any policy, however employed, inevitably entailing costs and necessarily

driven by underlying values.

And just as the term libertarian is distorting, so is the term paternalism. Burgess is concerned about the

“nanny state” but at the same time decries the reduction of public services, public funds, and public

regulation. Even without behavioural insights, liberal democracies have recognized that regulatory

interventions are needed in a myriad of context.

Burgess decries the politicisation of lifestyle health. We would argue however that policy is always

political, in the sense that we decide as a society how to employ our collective effort. We have argued

elsewhere that, “an overly broad definition of paternalism might encompass any collective effort to

improve people’s welfare and well-being. Although some extreme libertarians may subscribe to this broad

108 On Amir and Orly Lobel, “Stumble, Predict, Nudge: How Behavioral Economics Informs Law and Policy”, 108 Columbia Law

Review (2008), pp. 2098 et sqq.


definition, from a liberal perspective such definition is simply circular and meaningless.”109 Rather than

paternalistic, New Governance approaches, including nudge-style policies, can frequently be understood

by what economists would characterize as third-party externalities and the need for central coordination.

Health problems deplete public funds. Precisely because we would like to maintain universal social and

health insurance to be available to those in need, policy must work to mitigate our collective risks. Public

health programmes are not designed solely for the benefit of each individual but for the benefit of the

public. They are designed to promote good, long term health. The costs of health provision have surged in

recent years, in large part because medical treatment and the technologies of care have become far more

sophisticated. Risk taking, whether financial or physical, affects society as a whole.

There is no doubt that a focus on prevention is motivated partly by scarce resources and cost-benefit

calculations. Health issues have become increasingly costly. The UK taskforce for behavioural policy

explains that the initiatives are an answer to how government can spend public money more effectively:

“We currently spend over £2.5 billion a year on treating smoking-related illness, but less than

£150 million on encouraging smoking cessation. We spend an estimated £2.7 billion on treating

the results of excessive alcohol consumption, but only £8.7 million on promoting healthy

drinking levels (against £800 million spent on promotion by the alcohol industry).”110

Spending on prevention, rather than just remedial spending, allows regulators to allocate resources in a

more targeted and efficient manner in a regulatory world of scarcity. This is not an entirely new

development. As Julie Suk has recently shown, Europe has traditionally been more active than the United

States in preventive health in the workplace, through an infrastructure of occupational safety and health

law.111 European and national laws require employers to prevent risks to worker’s health by providing

regular checkups to all employee and monitoring the work environment to minimize risk factors. As we

stated earlier, the realities of regulation are those of scarce resources and any policy approach must be

responsive to questions about its comparative effectiveness and wisdom.

The term libertarian also conceals the redistributive effects and social purposes of nudges. Helping

individuals manages their physical and financial wellbeing affects the market. Businesses that have long

relied on risky tendency or suboptimal consumption patterns of consumers face a reduced market if

governmental interventions are effective. As behavioural research teaches us, irrationality is not universal.

While many of us experiences it frequently, sophisticated market actors, including individuals and

corporations, often find ways to overcome irrationalities. Reacting to these distributional questions,

behavioural economist Colin Camerer and others have therefore propose the term “asymmetric

109 Ibid., at p. 2124.

110 Cameron and Clegg, “Foreward”, supra note 99, at p.7.

111 Julie C. Suk, “Preventive Health at Work: A Comparative Approach”, 59 American Journal of Comparative Law (2011), pp 1089 et

sqq.


paternalism” for regulations that benefits those who err in their judgment and decision-making while

imposing little cost on others who are fully rational.112

Liberal democracies must allow individuals the freedom to make mistakes and cognitive errors. We call

into mind John Rawls’ liberal theory which explains how political principles are to be deliberately partial.

Policy must allow room for multiple conceptions of a good life, allowing people who differ to be free to

live differently and plural. The Rawlsian notion of the incompleteness of a political theory signifies the

importance of social differences to fill in the blanks. Within the realm of rationality, disagreement

pervasively exists. Take two individuals, both rational and internally consistent in respect of their own

views and behaviour: Ms. Fit and Ms. Full. Both Ms. Fit and Ms. Full have demanding careers and are

raising a family. In her spare time, Ms. Fit spends many hours a week exercising, preparing healthy meals,

and reading scores of new (and often conflicting) information about nutrition and fitness. Ms. Full prefers

to spend little time on those activities and instead in her free time chooses to read novels and perfect her

baking skills. Both know and recognize the value of fitness and weight control, but Ms. Full has made a

decision to follow her immediate and short term preferences, enjoy her free time to the fullest, and

assume the risk of any increase likelihood of health issues in the future. As a society we may still have a

reason to attempt to influence, in relatively non-coercive ways, Ms. Full’s choices. But the denial of her

rationality is simply false.

A strong example of risk taking that is based not upon irrationality but rather what we would more

accurately consider personality differences is that of extreme sports.113 The accidents from extreme

mountaineering or car racing are quite frequent. And yet, in a liberal society we allow a wide range of such

risks because we believe that extreme sports are legitimate forms of lifestyle. Smoking is a clear example of

how a society legitimately limits a pattern of harmful behaviour, not simply because among those who

smoke exists some degree of irrationality, but mostly because there is a collective cost to this choice.

These days, it is safe to assume that most smokers, if not all, know about the associated health risks of

tobacco. And despite the addictive quality of nicotine, many smokers, in particular those casual smokers

who smoke intermittently rather than regularly, are not addicted to smoking in any strong sense. They may

be perfectly rational risk takers who believe that the pleasure they derive from the occasional cigarette is

worth the health risk. And yet, it is today relatively uncontroversial that public policy should take a lead

role in regulating the sale and consumption of cigarettes. Smoking entails large externalities and high and

multiple health risks both to smokers and second-hand consumers. As stated earlier, in a world of limited

resources, the provision of affordable health care must be accompanied by a public willingness to support

healthy living.

112 Colin Camerer, Samuel Issacharoff, George Loewenstein et al., “Regulation for Conservatives: Behavioral Economics and the

Case for ‘Asymmetric Paternalism’”, 151 University of Pennsylvania Law Review (2003), pp. 1211 et sqq., at p. 1212. 113 George Loewenstein, “Because It Is There: The Challenge of Mountaineering... for Utility Theory”, 52 Kyklos (1999), pp. 315

et sqq., at p. 328.


While liberal policies must accept lifestyle differences, liberal democracies do not shy away from

normative stances. Not only traditional regulation but also New Governance and choice architecture

embody normative messages. Nudge style initiatives to promote savings through defaults is an example of

a social vision of how societies should be financially organized. Regulatory defaults convey normative

messages. This in itself is a behavioural lesson – that the packaging of information generates preferences

and norms. Preferences are adaptive and endogenous. This too underscores the inevitability of a

normative stance of a legal order. There is no reason drawing on behavioural insights in policy initiatives

would not be transformative. We disagree with Burgess’ description of behaviouralism implying that

policy is ‘accepting us as we are’ – social norms evolve and with it evolve the ways we apply decision-

making insights. Citing Bovens, Burgess worries that the long-term effect of nudging may be

infantilisation.114 This is a simplistic understanding of how people learn and behave. Assisting decision-

making does not imply unconscious, passive involvement by individuals. Rather, it takes seriously the gaps

between our revealed and stated long-term preferences and the various ways in which we stumble, as

humans, along the way.

Moreover, Burgess neglects the many instances in which nudges are inevitable. One way or another,

information is transmitted; decisions are requested. Public programmes and private markets almost always

include defaults as well as structures for disseminating and processing information. And, because of a

myriad of behavioural reasons, defaults are sticky. Choosing welfare-enhancing defaults rather than

remaining neglectfully default neutral should be a welcome focus of policymakers. The famous example of

opt-in and opt-out kidney donations is an excellent example of how design interventions can be successful

while maintaining an exit option.

IV. Scientific uncertainty and contemporary policy

There is much we still do not know about human judgment and decision-making. Burgess is therefore

correct in his call for more accurate and more humble adoption of behavioural economics into policy. The

question of whether policy can and should strive to adjust imperfect decisions depends on the prior

question of whether and how behavioural biases can be subject to corrective measures. Any analysis must

also identify the limits of the nudge-style corrective solutions to cognitive failures and the policy

challenges that cannot be addressed by new governance approaches, but rather continue to require more

traditional regulatory approaches. As we have repeated in our scholarship, the corrective solutions to

cognitive failures are limited. Identifying those challenges that cannot be addressed by new governance

approaches and will continue to require more traditional regulatory approaches is as much the work of

behavioural research as identifying the places where nudges are useful. Moreover, often it is the case that

114 Burgess, supra note 101, citing Luc Bovens, “The Ethics of Nudge”, in Till Yanoff-Grüne and Sven Hansson (eds.), Preference

Change (New York: Springer, 2009), available on the Internet at <http://www.bovens.org/TheEthicsFV.pdf> (last accessed on 4

January 2012).


regulation must be sequenced, such that nudges are employed first, followed by more top-down traditional

commands where nudges fail.115 Behavioural economics often neglects the institutional aspects of

enforcement and compliance. As Burgess points out, the starting point of behavioural economics is the

individual, rather than the social, cultural or political environment. Many of the groundbreaking

behavioural economics studies are lab experiments, which lack the rich the social and organizational

contexts of real market interactions and therefore are necessarily limited in their direct application to

concrete social policy. Integrating lessons from both lab and field experiments, as well as from the various

disciplines, including economics, psychology, sociology, and organizational behaviour, into meaningful

data is therefore a central challenge for regulatory agencies.

Importantly, building on behavioural insights, improved decision-making becomes both a goal for

regulation and a challenge for regulators, who are themselves decision-makers with human fallibility. As

we have argued previously, “if policymakers are to become consumers of the discipline of judgment and

decision-making, they must be wise consumers.” Burgess decries the uncritical adoption of nudge-style

reforms and indeed, distortion is always possible. Burgess is concerned that regulation is becoming

performative; that governments prefer to adopt shiny visible programmes rather than doing the hard work

needed for transformative change; that in the zeal to adopt certain reforms, uncertainty about what we

know may be denied. We share these concerns but argue that these are not unique to any particular mode

of regulation. Most clearly, Burgess is not in effect questioning the relevance of behavioural economics to

regulation, but rather the “wholesale and exclusive fashion” in which policymakers have taken to the field.

In debates about regulation and governance, the risk of dramatization – denial of any uncertainty, reliance

upon the most dramatic examples, making problems appear more immediate and pervasive than they

really are, and use of worst-case scenario projections as realistic models – is simply part of the nature of

politics, including, of course, risk politics. We must resist such flatness and distortion whether they appear

in the debates about the adoption of New Governance experimentations or the adoption (or reduction) of

traditional regulation.

As Burgess rightly points out, the relationship between behaviour and health is complex. In reaction to the

over-confidence of behavioural proponents, Burgess reminds us that mostly the insights arrive from

“experimental results rather than clear, applied successes.” While this is true, there have been successful

adoptions such as the aforementioned change of defaults in organ donations. One of the most profound

insights of new governance scholarship is that we must understand government itself as a web of R&D

units. Government agencies need to test, evaluate and re-evaluate regulations and programmes. In sharp

contrast to fixed agendas, a central lesson of New Governance is that strategies and policy require ongoing

learning and continual monitoring in light of new science and knowledge.116 At the same time, we must

remember that rather than be paralyzed by uncertainty, regulators must operate with the best available

information, not await perfect or complete science in order to act.

115 Lobel, “Citizenship”, supra note 97.

116 Lobel, “The Renew Deal”, supra note 97.


Burgess offers several examples of unintended effects of regulation. He asks that we consider “a well-

intentioned behavioural change brought about, in this case, by legal sanction. Compulsory cycle helmet

wearing leads to a fall in the numbers of children in particular, who cycle, as it becomes a regulated rather

than more spontaneous experience.” This is a classic example of traditional policy initiatives. Compulsory

regulation of helmets or seat belts is not a “nudge”. Over the years, such regulations have had numerous

positive effects of saving lives, and if there are empirics to suggest that some regulations have gone too

far, by being too costly or ineffective or counter-productive, this is an ongoing debate that naturally must

and will continue. Similarly, when Burgess offers as an example initiatives that seek to aid environmental

responsibility, these are, of course, as anyone in the environmental field is acutely aware of, controversial

and evolving initiatives. As Burgess has shown elsewhere, programmes may result in counter-productive

consequences, such as recycling initiatives instilling confidence of having solved the issue, while in fact

they are probably far less than what we ought to be doing as a society.117 Still, such claims and concerns

themselves require empirical examination. Are we instilling norms of caring about the environment when

we adopt recycling as a public initiative or are we instilling over confidence and crowding out other

efforts? Burgess further worries that policy-directed consumption that distinguishes between ‘goods’ and

‘bads’ will be misleading as diet is fundamentally a matter of balance and the strive for moderation. He

raises the question whether preventative schemes such as self-examination that risk being fraught with

false positives and unnecessary treatment are wise. In effect, Burgess is demanding, through the questions

he raises in his article, more, not less, lessons from behavioural studies to be brought into our public

debates and policy thinking.

The lessons embodied in New Governance approaches is that often substantive prohibitions and

adversarial enforcement are counter-productive. Command-and-control style regulation risks multiple

inaccuracies and will often lead to both over-enforcement – by focusing on the wrong and outdated issues

- and under-enforcement – by neglecting newer issues. It will also lead to over- and under-compliance,

because of the lack of input from private actors into the regulatory process. In the shift from command-

and-control regulation to more collaborative public-private problem solving, New Governance theory

suggests that individuals, groups and organizations become actively in the legal process, including the

processes of interpreting and complying with legal norms. Elsewhere, we have referred to the emphasis on

public-private cooperation and self-compliance in the shadow of the law as a “collaborative placebo,”

which triggers an ethos of voluntariness while maintaining background rules of command.118 Not only

risks, preferences, and knowledge, but also the social values and jurisprudential principles are inextricably

117 Sarah E.H. Moore and Adam Burgess, “Risk Rituals?”, 14 Journal of Risk Research (2011), pp. 111 et sqq. 118 Amir and Lobel, “Stumble, Predict, Nudge”, supra note 108, at p. 2130.


linked to the societies in which we live. As stated earlier, policy design, even when it opens space for

choice, is inherently normative and formative.

V. Conclusion

Lessons from experimental psychology and economics can inform law and policy. Predictions about

human irrationality can assist both the substance and design of regulation. Increasingly, legislators,

administrators, and courts draw on behavioural insights. Like most things in life and in law, these

developments are not risk-free. The nature of risk regulation is that information and the science-at-hand

are incomplete, evolving and subject to distorted interpretation. Yet the framework of new governance,

and the push for nudge-style experimentation, which allow choice, learning and continuous evaluation of

contemporary policy, serve to mitigate these risks and are a promising direction into the future of

regulation. One of the clear insights of psychology is that human beings value control and agency over

their lives. While there are no bright lines between “shoves” and “nudges”, there are important real

differences between policy that educates and assists better choices and policy that coerces, mandates, and

prohibits. We caution against the misleading, and ideologically-loaded, formula equating nudges, or new

governance techniques, of libertarianism meets paternalism. Policy nudges are value-driven, require debates

and resources, and are merely one approach in a fuller myriad of policy tools that regulators will have to

employ in order to be effective. Therefore, we believe that Burgess’ concerns are highly important in

efforts to encourage a more nuanced account of the range of mechanisms as well as the limits, costs, and

consequences of applying lessons from the field of behavioural economics to law.


Nudging Cannot Solve Complex Policy Problems

Evan Selinger∗ and Kyle Powys Whyte **

We deepen Adam Burgess’ insight that under current conditions nudging cannot solve complex policy problems reliably and

without controversy. We do so by integrating his concerns about nudging into Braden Allenby and Daniel Sarewitz’s three-

leveled model of the basic problems technology can address and generate. We use this model to explain why the UK

experiment with nudging has revolved around techno-fixes with limited policy potential, and conclude that nudging is best seen

as an emerging form of soft law.

I. Introduction

In theory, nudging can produce behavioral change people value. But under current conditions,

nudging cannot solve complex policy problems. Adam Burgess119 provides an excellent account of why

some advocacy of nudges tends to be overblown. We will deepen Burgess’ sociological and historical

analysis by integrating his insights into a model of the basic problems that technology can address and

generate, which we draw from systems engineering, philosophy, and science and technology studies. The

model clarifies why the UK experiment with nudging has revolved around techno-fixes with limited policy

potential. Before discussing the model, we will reconstruct some of Burgess’ insights from the 2012 paper.

1. Burgess’ concerns about nudging

a. Nudging is not the same as behavioral economics

Although nudging is associated with the 40 year-old field of behavioral economics, the two are

not identical. Behavioral economics refers to a research field that has found human decision making to be

less rational in some circumstances than classical economists typically assume. Nudges are attempts to

change human behavior by using some of the findings of studies from behavioral economics. But choice

architects must engage in the speculative and experimental process of creating the right interventions (e.g.

technologies, displays of choices, built-environments, etc.) that are adjusted for the biases and improve

undesirable behaviors on average. Burgess provides an interesting discussion of a point that has been

featured in other contexts: Behavioral economics research can show why people might make certain

∗ Associate Professor, Rochester Institute of Technology. ** Assistant Professor, Michigan State University. 119 Adam Burgess, “‘Nudging’ Healthy Lifestyles: The UK Experiments with the Behavioral Alternative to Regulation and the Market”, 3 EJRR (2012), this issue.


decisions, but it is not robust enough to cover reliable predications about how people will behave in non-

laboratory environments where variable perceptions of meaning exist.120

b. Proponents of nudging are overconfident

Nudging was branded in the United States as a “third way” between state interventionism and

market driven norms.121 David Cameron and the Conservative Party subsequently claimed nudging could

be a form of “MINDSPACE” governance that helps usher in the “cultural revolution” of the “big

society” (and its ethos of personal responsibility) and invested in the “Behavioral Insight Team” (so-called

“Nudge Unit”) “at a time when everyone else in the administration is experiencing significant cuts to

personnel within the Prime Minister’s Cabinet Office” . The Nudge Unit proclaimed that “There is no

reason why we cannot succeed in tackling today’s rising tide of chronic lifestyle-related disease.” This hype

is ironic given that health problems can involve value-laden disputes over ideals, outcomes, and even the

parameters of rationality. Indeed, health problems are often framed differently (even incommensurably),

giving rise to numerous conflicts, such as those over the best assessment metrics. They also can involve

cause-and-effect relations that are so hard to decode that seemingly reasonable attempts at solving

problems end up making things worse through unintended consequences. By themselves, behavioural

solutions like nudges can be too reductive to cope with such complexity. Some critics thus worry that

overselling the potential of nudges created opportunity costs: “Other critics have attacked the British

application of nudging on the grounds that it wrongly precludes more regulation and legislation to

promote health.” Moreover, nudging may not yield promised savings. A main reason why British citizens

initially found the idea of nudging appealing is that they were told it would help the government do more

for society than expensive alternatives. But evidence of savings is lacking, and given the limited knowledge

base, it was unrealistic to expect strong savings to occur early on.

c. Nudging is limited by conflicted ideals about freedom

Nudging tends to work best when users are unaware that their behavior is influenced by choice

architecture. In order for citizens to be comfortable with nudges, two conditions thus need to be satisfied:

first, citizens need to be comfortable having their behavior influenced at levels below the threshold of

conscious awareness; second, citizens need to perceive nudge designers as trustworthy. According to

Burgess, these conditions are difficult to meet because British culture is fundamentally conflicted about

the value of autonomy. In fact, concern about the government using nudges for Orwellian purposes

catalyzed critics to evoke Enlightenment style criticisms of manipulation. 120 See also Evan Selinger and Kyle Powys Whyte, “Competence and Trust in Choice Architecture”, 23 Knowledge, Technology & Policy (2010), pp. 461–482.; Kyle Powys Whyte, Evan Selinger, Arthur Capan et al., “Nudge, Nudge or Shove, Shove-The Right Way for Nudges to Increase the Supply of Donated Cadavor Organs”, 12 American Journal of Bioethics (2012), pp. 32–39; Evan Selinger and Kyle Whyte, “Is There a Right Way to Nudge? The Practice and Ethics of Choice Architecture”, 5 Sociology Compass (2011), pp. 923–935. 121 Anuj C. Desai, “Libertarian Paternalism, Externalities, and the ‘Spirit of Liberty’: How Thaler and Sunstein Are Nudging Us toward an ‘Overlapping Consensus’”, 36 Law & Social Inquiry (2011), pp. 263–295.


d. Nudging may infantilize

Many believe that moral maturity requires individuals who can grasp the difference between right

and wrong and self-direct towards righteous ends in the face of the temptations pulling in other directions.

Because nudging steers user behavior without teaching them anything or improving their skills, populating

policy with nudges runs the risk of infantilizing citizens—treating them like children to be controlled by

others.122

e. Nudging is technocratic

Advocates represent nudging as a means to help individuals achieve goals they set for themselves.

Nudging is thus formally differentiated from propagandizing. Nudges are in principle compatible with

democratic politics because they simply realize the will of the people. Nevertheless, nudging relies on

experts to determine “solutions to assumed problems,” while preventing “wider issues,” such as

politically based arguments against a behaviour being legitimately classified as socially undesirable, from

figuring in “its landscape.” For example, “Schemes such as to reduce the amount of sex university

students may be having are explored without hesitation, suggesting how marginalized concerns for privacy

and liberty have become. It would appear to confirm that a managerial approach to society is firmly

established, and manipulating information for the ‘greater good’ of avoiding risk or improving behavior

unremarkable and assumed to be unproblematic.” These are the sorts of decisions that people are not

always comfortable having experts and bureaucrats make for them.

2. The fundamental problem of nudging policy

Burgess’ insights revolve around a major concern: Under current conditions, nudging cannot

solve complex policy problems reliably and peacefully. In this section, we develop the insight further by

integrating Burgess’ views into a model advanced by Braden Allenby and Daniel Sarewitz (2011)123 that

clarifies the basic problems technology can address and the main problems its use can generate. We are

focusing on technology because policy applications of nudging tend to be instantiated as changes in

technological or built-environment design that adhere to the logic of “techno-fixes”—a concept

elaborated upon below.

The Allenby-Sarewitz model has three levels. Level I is “shop-floor” technology embedded in a

“simple” “physical system.” In this domain, techno-fixes are appropriate solutions for narrowly

circumscribed problems: “Most of the necessary relationships amongst goals, means, and causality have

122 Luc Bovens, “The Ethics of Nudge”, in Till Yanoff and Sven Ove Hansson (eds.), Preference Change, (New York: Springer, 2009). 123 Braden Allenby and Daniel Sarewitz, The Techno-Human Condition (Cambridge, MA: MIT Press, 2011).


already been captured in a physical system that can be used with confidence that a given input will

produce a desired output.”124 Examples include vaccines, cars, and highly developed neuropharmaceuticals

that can be used to reliably achieve goals as basic as staying protected from diseases, driving on a road in a

desired direction, and performing better at a mental task (ibid).

Level I technologies can be designed to “eliminate risk” or transform environments so that they

contain “less risky alternatives” than otherwise would be present.125 Approaching the “hierarchy of

control” in this way can differ significantly from adopting alternative approaches to risk management that

“rely on institutional and social systems.”126 Allenby and Sarewitz offer the following example as an

illustration of the divergent approaches:

Consider…the use of toxic chemicals in a manufacturing process. One way to manage such a risk

to employees is to provide appropriate protective equipment and mandate its use. But protective

equipment can break down, be misplaced, or simply not used. (Employees often dislike protective

equipment because it is uncomfortable and makes their job harder; managers may not be as strict

as they should be in requiring it as they should be, because it can impeded productivity.)

Alternatively, one can redesign the manufacturing process to use less toxic materials in the first

place. The former…is never as effective as the latter (a “vaccine approach”).127

On the basis of comparing the options available in this case as well as a host of similar ones, they

conclude that, “The more responsibility for safety you can transfer from human and institutional

decisions processes to technologies themselves, the safer the system will be, all else equal.”128 That is the

beauty of a techno-fix: means and ends are unified.

If all major problems could be solved by techno-fixes, life might be easier. But, techno-fixes have

a limited range of application precisely because social life is immensely complex.

This is where Level II comes in. It consists of “complex socio-technical systems” that are “infinitely less

predictable” than Level I technology, but which support it through delivery, coordination, and regulatory

systems.129 For example, airplanes are a Level I technology, but “airline corporations, the government

security apparatus as applied to air travel, and market capitalism in route pricing” are Level II

subsystems.130 While airplanes cannot be integrated into society without Level II support, mismatches

between Level I and Level II operations frequently arise that cause travellers to experience immense

124 Ibid., p. 107. 125 Ibid., p. 52. 126 Ibid. 127 Ibid., pp. 51–52. 128 Ibid., p. 52. 129 Ibid., p. 38. 130 Ibid.


frustration: good safety records co-exist with high incidents of delays and cancellations and poor customer

service. Frustration is not the only problem that can occur when Level II issues are at stake. In the last

section, we reviewed Burgess’s analysis of how concern over governmental development and

administration systems impede citizen enthusiasm for using nudging as a policy tool.

Let us return to the vaccine example (Level I technology), but consider it from the perspective of

Level II activity related to risk management and public health. Malaria is a serious problem in Africa and

currently no vaccine exists. The most useful preventive technology is bed netting. Despite heavy expense

and effort, by 2005 “only 3 percent of African children in malarial regions were sleeping under insecticide-

impregnated bed nets.”131 Policy consensus emerged that problems concerning local context—mismatches

between the technology’s physical properties and social/cultural norms—created the biggest obstacles to

adoption. To make progress, an approach called “Integrated Vector Management” (IVM) was embraced

that tailored interventions to local factors, including “type of mosquito and malaria parasite, climate, cost,

and available resources.”132

While this initiative proved exceptionally effective, Allenby and Sarewitz note that, “The main

selling point of IVM—its sensitivity to context—is also its weakness.”133 By weakness, they mean

adjusting interventions to context can be far less reliable and far more resource intensive than

standardized techniques that allow the same action to invariably yield the same result. For this reason,

even if a highly reliable, standardized malaria vaccine were invented, it would be unrealistic to expect that

gains in a Level I system would scale-up smoothly to the Level II action of administering to human

patients: “There would, no doubt, be controversies over appropriate policies for vaccine delivery, there

would be organizational dysfunction at various levels, and a continued lack of infrastructure for delivering

health care in parts of Africa.”134 In other words, problems would arise that are structurally similar to the

issues when employees are mandated to protect themselves by wearing safety equipment.

Level III technology is the most difficult to understand. Fortunately, not much needs to be said

about it here, as the problems with nudging mostly concern misalignment between Levels I and II. Simply

put, Level III concerns “wicked problems” occurring at a “whole earth level” in which multiple systems,

operating with different norms and at different scales, interact with one another to produce emergent

behavior that can be exceptionally difficult, if not impossible, to predict. For example, if an airplane

functions at Level I, and air traffic controllers act at Level II, then the global problem of infectious disease

travelling rapidly due to the pervasiveness of international flights can be considered a Level III effect. The

complexity at issue here strains moral imagination and requires “macro-ethical” strategies to cope with.135

Although the model is descriptive, it can be applied to further two important normative ends.

131 Ibid., p. 48. 132 Ibid., p. 48–49. 133 Ibid., p. 50. 134 Ibid., p. 49. 135 Ibid., p. 184.


1. It can help us detect invalid predictions that are made about projects revolving around artifacts

and technological systems. For example, people who believe that inventing a new solar

technology will solve the energy crisis confuse Level I and Level III systems. Likewise, people

who claim that meaningful economic development will result from transferring a vaccine to a

developing country mistake Level I and Level II systems.

2. It can help us to determine when the right and wrong scale of assessment is used to judge a

technical project or a technical system. In the transhumanism debates, for example, Allenby and

Sarewitz believe confusion arises because proponents address Level I while opponents address

Level II and Level III.

Let us apply the model to the situation of using nudges in British policy, accepting all the

information Burgess provides as accurate.

Table 1

Level I Nudging Many nudge proposals have not qualified as Level I

interventions when embodied in artifacts and built-

environments. Currently, the cause-and-effect

relations are not nearly as predictable as vaccines

and other paradigmatic Level I technologies.

Level II Nudging Deploying nudges has been a highly problematic

endeavor because, in addition to the causal

problem, the systems required to fund their design,

administer them, prime the public for widespread

use, and convince the public that nudges are

innovative tools that enhance well-being across

society, are plagued by ethical and political concerns

related to democratic governance, cognitive liberty,

fair use of public funds, and opportunity costs.

Unlike the IVM example, technical specifications

are misaligned with social and cultural

norms/values.

Level III Nudging Nudges are not really discussed in this register. For


example, although discussions abound about how

nudging can reduce energy consumption, nudge

proponents do not even attempt to discuss at all the

complex interventions that are needed to shift

diverse citizens in diverse parts of the world away

from a CO2 intensive society or adapt to the

changes that might accompany the perpetuation of

CO2 intensive industries, infrastructures, and

lifestyles.

Looking at things through Allenby and Sarewitz’s three levels, the hype surrounding the

anticipated behavioral results and savings of nudging can be seen as misguided. Not only are the robust

cause-effect relations that typify Level I absent in nudges, but the ethical and political concerns expressed

at Level II should have been foreseen and better dealt with. Having been subject to critique, the

Behavioral Insight Team is now more sensitive to these issues. In a recently published report136 they stated

that more small scale experimentation needs to take place in order to assess the feasibility of certain

nudges (and other behavioral interventions of interest). But these experiments only promise to shed light

on the Level I effects of nudges. That is, they can only help us understand whether plausible nudge ideas

could work relatively well in certain contexts. They do not appear to address the Level II problems that

Burgess identifies. Approaches borrowing from stakeholder dialog models found in participatory

governance are needed to address these concerns. But even these models may be insufficient, in-

themselves, to guide policy makers through the difficulties that can arise when gains in contextual-

sensitivity come at the expense of reliability and increase an operation’s costs.

II. Conclusion

Was solid evidence ever really offered that nudging could contribute significantly to policy? Or,

was the policy rhetoric always inflated? It should come as no surprise to the reader, but our respective

answers are: no and yes.

While Burgess focuses on UK problems, the fact is mistakes have been made from the start

concerning discussion of nudges. These mistakes began in the United States in the very text that got the

ball rolling: Nudge. One of Thaler and Sunstein’s favored examples is the “Toxic Release Inventory.”137 It

is a list of “potentially hazardous chemicals that have been stored or released into the environment”

136 BIT, “Behavioral Insights Team Annual Report”, Cabinet Office (2010-2011). 137 Richard Thaler and Cass Sunstein, Nudge: improving decisions about health, wealth, and happiness (New Haven: Yale University Press, 2008), pp.192–93.


archived on the U.S. Environmental Protection Agency’s website.138 Thaler and Sunstein depict the

information as potentially “the most unambiguous success story in all of environmental law.”139 They give

such high praise because the data created an “environmental blacklist” that generated bad publicity for

offending companies; in some cases, adverse reaction lead to stock price devaluation. 140

Unfortunately, as commentators have noted, the Toxic Release Inventory is not a nudge.141

Nudges, as per Thaler and Sunstein’s definition, cannot change behavior by significantly altering economic

incentives. The inventory, however, brings about environmental compliance by virtue of increasing

polluting costs.142 It is surprising that Thaler and Sunstein are silent on this. Indeed, their analysis actually

emphasizes the inventory’s ability to cause economic harm: “If companies are able to reduce emissions at

a low cost, they will do so, simply in order to avoid the bad publicity and resulting harms.”143 In fact, not

only do Thaler and Sunstein mistakenly describe the inventory as a nudge, but they also suggest an

analogous version—also erroneously depicted as a nudge—should be created to reduce greenhouse gas

emissions: “The government should create a Greenhouse Gas Inventory (GGI), requiring disclosure by

the most significant emitters.” 144

Another much discussed policy nudge that Thaler and Sunstein offer is the use of mandated

choice schemes to increase the supply of donated organs. This policy has recently been instituted in Great

Britain as a “required choice” policy.145 They contend that mandating people to choose whether or not to

donate their organs as a requirement for obtaining a driver’s license would increase the amount of

registered donors, while “imposing essentially no new burdens on taxpayers,” and avoiding the political

barriers of presumed consent.”146 Unfortunately, this example is not persuasive, either. It has been argued

that Thaler and Sunstein’s view of how this nudge would motivate people to embody altruistic convictions

is flawed.147 Simply put, they overlook the fact that people can be influenced by perceptions of features

found in and associated with the Department of Motor Vehicles. These features can be emotional triggers

that “shove” people looking to obtain licenses toward opting against donation. Whereas the Toxic Release

Inventory fails on definitional grounds, this one fails on empirical ones.

Now, Thaler and Sunstein do offer an example of a policy nudge that meets both definitional and

empirical criteria. This example concerns Lake Shore Drive, a U.S. roadway that has stunning views of

Chicago’s skyline. One particular segment includes a series of S curves that require drivers to slow down

to 25 mph. Distracted by scenery or failing to calculate depth properly, drivers often ignore the reduce

speed limit sign. To alter the dangerous behavior, a visual illusion, induced through a series of painted

138 Ibid., p. 192. 139 Ibid. 140 Ibid., p. 192-93. 141 Daniel M. Hausman and Brynn Welch, “Debate: To nudge or not to nudge”, 18 Journal of Political Philosophy (2010), pp. 123–136. 142 Ibid., p. 124. 143 Thaler and Sunstein, Nudge, supra note 19, at p. 193. 144 Ibid. 145 BIT, Behavioral Insights Team Annual Report, supra note 18 146 Thaler and Sunstein, Nudge, supra note 19, at p. 178. 147 Whyte, Selinger, Capan et al., “Nudge, Nudge or Shove, Shove”, supra note 2.


white stripes, has been installed; the stripes incline drivers to slow down: “When the stripes first appear,

they are evenly spaced, but as drivers reach the most dangerous portion of the curve, the stripes get closer

together, giving the sensation that driving speed is increasing. One’s natural instinct is to slow down.”148

But if we assume this example works, the result actually does not bode well for the hype about

nudges. Repainting the roadway is by no means a third alternative to regulation or the market. Road safety

is the responsibility of federal and state regulators. The white stripes are simply the product of a decision

by the traditional authorities. Second, the stripes do not actually eliminate the problem. According to a

2006 article, the Transportation Department “set aside $1.4 million for a two-year study that could result

in a redesign of the Oak Street curve and improvements in accessing the lakefront.”149 There is still much

more to be studied regarding the importance of wetness conditions, warning signs, and other

environmental features. In this case, then, a nudge provides limited gains and, at best, is still not a Level I

technology. Rather, nudging here is more like an emerging form of soft law that regulators can explore as

an ordinary part of their charges.

On the basis of Burgess’s concerns, our application of Allenby-Sarewitz model, and the problems

identified with how Nudge treats policy, our conclusion is that under current conditions nudging cannot

solve complex policy problems reliably and without controversy. The UK nudge experiment thus far

primarily has treated nudges as techno-fixes that only can have limited policy value. Whether the

Behavioral Insight Team makes the changes needed to improve Level I and II performance, or whether it

is merely rhetorically gesturing to these improvements, remains to be seen. Nudges are best seen as an

emerging form of soft law regulation.

148 Thaler and Sunstein, Nudge, supra note 19, at pp. 38–39. 149 Jon Hilkevitch, "Lake Shore Curve to Get More Alerts," Chicago Tribune, 24 July 2006.


Nudging Smokers: The Behavioural Turn of Tobacco Risk Regulation

Alberto Alemanno*

By building upon Adam Burgess’ opening essay, this contribution aims at critically examining the application of nudging approaches to the current efforts of regulating lifestyle choices, such as tobacco use, excessive use of alcohol, unhealthy diets and lack of physical exercise. In particular, it discusses the viability of nudges approaches as applied to current tobacco control policies. Tobacco regulation is a particularly contentious area because it affects the social habits of smokers by preventing them from enjoying products at the time and place and in the form they have become accustomed to. As such, it is often perceived as a symbol of the nanny-state infringing on individual liberty. After providing an account of the range of tobacco control policy tools that have developed over time, the article discusses the regulatory philosophy currently underlying anti-tobacco efforts by focusing on the mainstream concept of ‘de-normalisation’. It then illustrates how most of the policies aimed at de-normalising tobacco today rely on ‘nudging’ approaches via behavioural change rather than via the provision of information. It finally argues that – due to the actual approach towards tobacco – most of the flaws generally identified with this alternative regulatory approach seem overcome in the context of tobacco control.

I. Introduction

At a time when policy makers want to change the behaviour of citizens to tackle a broad range of social

problems, such as climate change, excessive drinking, obesity and crime, a promising new policy approach

has appeared that seems capable of escaping the liberal reservations typically associated with all forms of

regulatory action. After having relied on the assumption that governments can only change people’s

behaviour through rules and regulations, policy makers now seem ready to design policies that better

reflect how people really behave, not how they are assumed to behave as rational agents. The approach,

which stems from the increasingly ubiquitous findings of behavioural research150, is generally captured

under the evocative concept of ‘nudge’151. Inspired by ‘libertarian paternalism’152, it suggests that the goal

of public policies should be to steer citizens towards making positive decisions as individuals and for

society while preserving individual choice. Acting as ‘choice architects’ policy makers organize the context,

process and environment in which individuals make decisions. In so doing, they exploit some patterns of

irrationality, often called ‘cognitive biases’, to manipulate people’s choices153. Thus, in the famous

Cafeteria example, the school’s management might try to affect students’ diet by rearranging the display of

food to make it more likely that students will choose the healthy option. This innovative approach to

policy making is part of a broader shift from traditional regulation to regulatory governance that is

generally referred to as New Governance. As behavioural research shows the limits of human rationality,

New Governance unveils the limits of traditional regulation. In particular, by denouncing as too narrow

* Alberto Alemanno is a Jean Monnet Professor of EU Law & Risk Regulation at HEC Paris. The author would like to thank Adam Burgess and Adam Jaffee for the constructive comments received on a first draft of this piece as well as Uladzislau Belavusau for providing him the chance to test some of these ideas with his students at VU University Amsterdam in January 2012. 150 Daniel Kahneman, Thinking, Fast and Slow (New York, NY: Farrar, Straus and Giroux 2011). 151 Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth and Happiness, (London: Yale University Press 2008). 152 Cass R. Sunstein & Richard H. Thaler, “Libertarian Paternalism Is Not an Oxymoron”, 70 U. Chi. L. Rev. (2003), pp. 1159–1202. 153 See, e.g., Jonathan Baron, Thinking and deciding (Cambridge, MA: Cambridge University Press 2007); Dan Ariely, Predictably irrational: The hidden forces that shape our decisions (New York, NY: HarperCollins 2008).


and ineffective regulatory techniques such as command-and-control to manage the state’s increasing

dependence on non-state actors, New Governance promotes a diverse view of state authority and its

relationship with civil society and the business world154. If effectively reconciled, behavioural and New

Governance approaches together carry the potential to provide policy makers with a more complete

understanding not only of how people behave and make decisions but also of how they react vis-à-vis

different forms of regulatory intervention. As a result, they both encourage policy makers to experiment

with new regulatory approaches capable of internalising human irrationality as well as the inherent flaws of

traditional regulation.

By building upon Adam Burgess’ opening essay, this contribution aims at critically examining the

application of nudging approaches to the current efforts of regulating lifestyle choices, such as tobacco

use, excessive use of alcohol, unhealthy diets and lack of physical exercise. In particular, it discusses the

viability of nudges approaches as applied to current tobacco control policies. Tobacco regulation is a

particularly contentious area because it affects the social habits of smokers by preventing them from

enjoying products at the time and place and in the form they have become accustomed to. As such, it is

often perceived as a symbol of the nanny-state infringing on individual liberty155. After providing an

account of the range of tobacco control policy tools that have developed over time, the article discusses

the regulatory philosophy currently underlying anti-tobacco efforts by focusing on the mainstream

concept of ‘de-normalisation’. It then illustrates how most of the policies aimed at de-normalising tobacco

today rely on ‘nudging’ approaches via behavioural change rather than via the provision of information. It

finally argues that – due to the actual approach towards tobacco – most of the flaws generally identified

with this alternative regulatory approach seem overcome in the context of tobacco control. However,

despite its potential for providing a philosophical base justifying the current ‘permit but discourage’

approach typical of tobacco control and other lifestyle policies156, it cannot be ruled out that ‘nudging’

might encounter some of the same obstacles it faces in other less contentious areas of policy-making.

II. Tobacco regulation as a driver of social change

After centuries of being a socially accepted habit, smoking has undergone a cultural transformation in

most Western countries, which was made possible by a mix of legal controls, stigmatization, education and

cessation programmes. The shifting public attitude and social norms have been both the cause and effect

154 See, e.g., Karen Yeung, “The Regulatory State”, in Robert Baldwin, Martin Cave and Martin Lodge (eds.), Oxford Handbook of Regulation (Oxford: Oxford Universiy Press 2011), pp. 80–81. 155 See, e.g., Alberto Alemanno, “Out of Sight Out of Mind – Towards a New Tobacco Products Directive”, 18 Columbia Journal of European Law (2012), forthcoming; and Scott Crosby, “The New Tobacco Control Directive : An Illiberal and Illegal Disdain for the Law”, 27 European Law Review (2002), pp. 177–193. 156 William A. Bogart, Permit But Discourage, Regulating Excessive Consumption (Oxford: Oxford University Press, 2011).


of the legal intervention, which is generally termed ‘tobacco control’. As more citizens perceived smoking

as a filthy and unhealthy habit, support grew for regulatory intervention aimed at further decreasing

consumption. In turn, as tobacco control initiatives intensified, more citizens accepted - at both the

individual and collective level - the hazardousness of cigarettes. However, this normative shift did not

occur outside of the developed world, where tobacco prevalence has grown in recent years157. Today,

approximately 84% of the world’s smokers, around 900 million people, live in developing or emerging

economies158.

‘De-normalisation’ is the most popular concept in anti-tobacco circles today. Although its precise meaning

has changed over time159 and varies within the tobacco control advocacy community160, its underlying goal

can be defined as aiming “to change the broad social norms around using tobacco – to push tobacco use

out of the charmed circle of normal desirable practice to being an abnormal practice”161. This means de-

normalizing tobacco consumption by countering the allure of smoking created by the tobacco industry in

order to stigmatize smokers162 and thereby reduce smoking rates163.

The recently-adopted – in spite of vast resistance from the tobacco industry164 – smoking bans in public

places, such as bars and restaurants, epitomize this regulatory-induced phenomenon. In line with the de-

normalisation objective, the bans, by creating an environment where smoking becomes increasingly more

difficult, help shift social norms away from the acceptance of smoking in everyday life and promote public

rejection of cigarettes. Yet tobacco control measures represent only one part of the factors affecting the

‘de-normalisation’ of smoking that also comprehend many facets of culture, custom and social norms165.

Smoking, similar to eating and drinking, is accompanied and shaped by a myriad of norms, attitudes,

opinions and reactions. Legal regulation is but one form of regulation, albeit the most prominent

technique of social control, capable of altering this behaviour. It is as much about persuading people as it

is about commanding them. Thus, Julia Black offers a very broad definition of regulation that goes well

beyond legal norms: “the sustained and focused attempt to alter the behaviour of others according to

157 See, e.g., Prabhat Jha and Frank J. Chaloupka, “The Economics of global tobacco control”, 321 BMJ (2000), pp. 358–61 ; Campaign for Tobacco-Free Kids, China : Tobacco Burden Facts, Washington D.C. 2009 ; Roberto Iglesias, Prabhat Jha, Márcia Pinto et al., “Tobacco Control in Brazil”, Health Nutrition and Population (HNP) Discussion Papers, World Bank, Washington D.C., 2007. 158 Prabhat Jha and Frank J. Chaloupka, “The Economics of global tobacco control”, 321 BMJ (2000), pp. 358–61. 159 Eric A. Feldman and Ronald Bayer, “The Triumph and Tragedy of Tobacco Control: A Tale of Nine Nations”, 7 Annual Review of Law and Social Science (2011), at p. 88 (illustrating how the denormalisation phenomenon was at first a consequence of the public health interventions but then became an autonomous ‘discrete policy goal’). 160 Simon Chapman and Becky Freeman, “Markers of the De-normalization of Smoking and the Tobacco Industry”, 17 Tobacco Control (2008), pp. 25–31. 161 David Hammond, Geoffrey T. Fong and Mark P. Zanna, “Tobacco De-normalization and Industry Beliefs among Smokers from Four Countries”, 31 American Journal Prev. Med. (2006), pp. 225–232. 162 Erving Goffman, Stigma : Notes on the Management of Spoiled Identity (New York, NY: Penguin, 1995). 163 Sei-Hill Kim and James Shanahan, “Stigmatising Smokers : Public Sentiment Toward Cigarette Smoking and its Relationship to Smoking Behaviors”, 8 J. of Health Communication (2003), pp. 343–367. 164 Frank Jordans, “WHO chief slams tobacco firms that ‘harass’ governments to prevent anti-smoking measures”, The Washington Post, 25 November 2011. 165 On the relationship between legal regulation and social norms, see, e.g., William A. Bogart, Permit But Discourage, Regulating Excessive Consumption (Oxford: Oxford University Press, 2011), at p. 130.


defined standards or purposes with the intention of producing a broadly identified outcome or outcomes,

which may involve mechanisms of standard-setting, information gathering and behavioural

modifications”166. The convoluted relationship between the regulation of tobacco and public attitudes

towards alcohol reflects a central theme about the co-dependence of legal intervention and norms. The

relationship between law and norms, often defined as ‘normativity’, is of critical importance in regulating

lifestyle choices. Indeed, any regulatory intervention is more likely to succeed if society has recognized the

underlying issue as a problem and the regulation is aligned with other social control techniques pursuing

similar goals167. Normativity also offers a valuable lens to examine nudging and its ability to induce not

only behavioural change but also social norms. This leads to the question of the exact role played by

tobacco regulation on smoking consumption168.

The history of tobacco has been closely intertwined with social values, habits and consumption patterns169.

As a result, tobacco regulation has largely been driven by then current attitudes towards the product. In

this context the impact on social behaviour of the various regulatory mechanisms on packaging, product

contents, advertising and sales are very difficult to assess and quantify. Asserting that a law produces a

particular effect is one thing. Proving that these consequences have occurred and that they have been

triggered by that law is quite another. In assessing the impact of a law there is the challenge of perspective:

is there a neutral position from which to investigate the acts and behaviours said to be the effect of the

law? In the affirmative, how can one assess and interpret these acts and behaviours in order to draw

conclusions in terms of impact? Moreover, given the existence of a regulatory ‘mix’ in the fight against

tobacco, i.e. coexistence of different strategies targeting the same goal over the same period of time, how

can one distinguish the precise effect of any one intervention? Those questions suggest that one may

encounter a number of difficulties when called upon to show that a law actually caused any particular

result. But there is one particular complication that makes the proof of causation particularly arduous to

overcome: this is the so-called ‘plausible rival hypothesis’. To demonstrate that a law produced a particular

effect, it is necessary to prove that there were no other social, political, economic or other forces that

might explain that particular result. Only insofar as rival hypothesis can be eliminated, can conclusions be

drawn with some confidence about the causal relationship existing between the law and the specific result.

Moreover, law, like many other human endeavours, might also produce unintended consequences170.

Thus, evidence suggests that as smoking rates dropped among the most privileged segments of society,

166 Julia Black, “Regulatory Conversations”, 29 Journal of Law & Society (2002), pp.163–196, at p. 170. 167 William A. Bogart, Permit But Discourage, Regulating Excessive Consumption (Oxford: Oxford University Press, 2011), at p. 79. 168 For two different narratives of tobacco control, see, e.g. William A. Bogart, Permit But Discourage, Regulating Excessive Consumption (Oxford: Oxford University Press, 2011); Eric Feldman and Ronald Bayer, Unfiltered : Conflicts over Tobacco Policy and Public Health (Cambridge : Harvard University Press 2004). 169 Allan M. Brandt, The Cigarette Century: The Rise, Fall, and Deadly Persistence of the Product that Defined America (New York : Basic Books 2007).

170 William A. Bogart, Six Ideas about the Impact of Law – America the Outlier, in Consequences : The Impact of Law and its Complexity (Toronto: Toronto University Press 2002), at p. 111.


tobacco prevalence among the low-income classes dropped less dramatically171. As it has been argued,

“When we develop the ability to control disease and death, the benefits of this new-found ability are

distributed according to resources and knowledge, money, power, prestige, and beneficial social

connections”172. This seems to render the task of measuring the impact of regulatory intervention even

harder than originally expected and illustrates that the outcomes are not as straightforward and effective as

advocates of such interventions typically claim.

However, there seems to be a shared belief that regulation may play a role in behavioural change alongside

education, public health reform and tax policy173. In particular, it seems that even though tobacco

regulation should only contribute to make the adverse effects stemming from tobacco consumption

salient within society, it might still successfully contribute to the de-normalization phenomenon. Yet the

philosophy of tobacco regulation fundamentally lies on a self-evident incongruity: unlike any other form

of product regulation, its rules do not promote the underlying product by making it more acceptable to

the public but constrain its production, marketing and consumption without banning it. As Chapman

observes, “The paradox with tobacco is that it is so dangerous that no routine regulatory approach can

make sense of it”174. Hence, how to de-normalise a product that remains legally placed on the market?

Unsurprisingly, tobacco companies, which are always ready to challenge the control efforts, have largely

highlighted this contradiction175.

In this context, ‘libertarian paternalism’ seems to provide a promising and badly needed philosophical

justification for today’s apparently contradictory tobacco control policies. Even the most radical anti-

smoking groups within the health community concede that members of the public should have the

freedom to smoke if they want to, but the groups would rather prefer that people did not smoke and seek

to do everything they can to make it hard for smokers to do so. As a result, acting in the name of

‘libertarian paternalism’, public authorities let you to smoke but since they know that you would like to

quit - if only you had the strength of will as well as the sharpness of mind - they will keep you away from

tobacco products. Unlike paternalists, who ban some things and mandate others, ‘nudging’ – as is

currently embedded in most tobacco control efforts – aims only to skew individual decisions, without

infringing greatly on freedom of choice176.

171 On the existence of a social gradient in smoking, Eric A. Feldman and Ronald Bayer, Conclusions : lessons from the comparative study of tobacco control, in Eric Feldman and Ronald Bayer, Unfiltered : Conflicts over Tobacco Policy and Public Health, (Cambridge, MA: Harvard University Press 2004), at p. 300. 172 John Pehlan and Bruce Link, Controlling disease and creating disparities: A fundamental cause perspective, 60(b) J. Gerontology B. (2005), at p. 27. 173 Geraint Howells, The Tobacco Challenge – Legal Policy and Consumer Protection (Surrey: Ashgate 2011), at p. 7. 174 Simon Chapman, “Benefits and Risks in Ending Regulatory Exceptionalism for Tobacco”, 17 Tobacco Control (2008), pp.73–74. 175 Michael Givel, “FDA Legislation”, 16 Tobacco Control (2007), pp. 217–218. 176 Leaders, “Soft Paternalism: the State is Looking After You”, The Economist, 6 April 2006.


III. The evolution of tobacco control

Although it was not until the twentieth century that the health problems associated with tobacco use

gained public knowledge, smoking bans in public places are as old as tobacco consumption. The first

known smoking ban occurred in 1590 and was declared by Pope Urban VII during his short reign as

pope, and others were issued by the Ottoman sultan Murad IV in 1633, by the Prussians in Königsberg in

1742, and finally by the Nazi regime, under the auspices of Karl Astel’s Institute for Tobacco Hazards

Research, created in 1941 under the orders of Adolf Hitler. Yet what makes smoking bans more salient

today is their rapid spread and virtually universal character, being mandated by an International agreement,

the Framework Convention on Tobacco Control (FCTC), the world’s first global public health treaty

negotiated under the auspices of the World Health Organization (WHO)177. Under this instrument, its 174

signatories, representing more than 90% of the world’s countries – including China and India – are

expected to adopt, besides advertising and smoking bans, an entire new generation of tobacco control

measures. Interestingly enough, the FCTC “instead of treating the act of smoking as an individual

preference or a legal right, [ ] presents it as an unacceptable risk obligating the national and international

community to intervene in the name of public health”178.

The first generation of measures was implemented only during the Seventies, when taxes imposed on

tobacco products, especially cigarettes, became a principal weapon in the fight against tobacco use. A

series of studies by economists demonstrated convincingly that increases in cigarette prices, driven by

increases in cigarette taxes, reduced cigarette smoking179. In particular, several studies indicated that tax-

induced price increases were particularly effective in discouraging smoking by young people180. Although

demand for cigarettes is only modestly responsive to price changes, there is consensus that in developed

countries, for every 10 per cent increase in price, the quantity of cigarettes demanded by consumers will

fall by about 4 per cent. Since cigarette taxes constitute only a fraction of total cigarette price, large

increases in cigarette tax rates can simultaneously generate significant increases in governmental revenues

and a substantial decrease in smoking. Thus, tobacco taxation gives governments the opportunity to

177 See Framework Convention on Tobacco Control, 21 April 2003, 42 International Legal Materials (2003), at p. 503. For an analysis, see Benn McGrady, Trade and Public Health: Tobacco, Alcohol and Diet (Cambridge: Cambridge University Press 2011).

178 Eric Feldman and Ronald Bayer, Unfiltered : Conflicts over Tobacco Policy and Public Health (Cambridge : Harvard University Press 2004), at p. 81. 179 See, e.g. Frank J. Chalopka, Melanie Wakefield and Christina Czart, “Taxing Tobacco: The Impact of Tobacco Taxes on Cigarette Smoking and Other Tobacco Use”, in Robert L. Rabin and Stephen D. Sugarman (eds.), Regulating Tobacco (Oxford: Oxford University Press 2001). 180 For an overview of the literature on tax policies in tobacco control, see, e.g., Geraint Howells, The Tobacco Challenge – Legal Policy and Consumer Protection (Surrey: Ashgate 2011), pp. 208–211.


enrich their treasuries while improving public health181.

Besides fiscal measures, laws were passed to protect bystanders and youth so as to counter the pervasive

criticism of paternalism. Thus, governments prohibited tobacco sales to minors and banned smoking in

public places, notably work places. In the same period, authorities introduced restrictions on direct

advertising, leading the way, in several European countries, to a prohibition of indirect advertising.

Moreover, as soon as policy-makers across industrialised countries realized that tobacco companies had

begun using cigarette packs and specialized retail outlets as advertising vehicles, they adopted space

appropriations rules. Thus, for instance, laws were adopted mandating companies to print large messages

on cigarette boxes, such as “Smoking causes peripheral vascular disease”, “Don't let children breathe your

smoke” or “Smoking causes lung cancer”. This kind of message had been written on boxes since the late

Sixties, but the legislation begun in the 1990s was more ambitious. In line with the FCTC, for instance,

warnings must cover a given percentage of the packet front and back182. In most countries, messages are

written in large, standardized black letters on a white background, surrounded by a black frame. Within

the EU, some countries decided to go further by introducing mandatory graphic warnings and combining

these warnings with images of tobacco-linked diseases. It seems undisputed that the combined adoption

of this first set of tobacco control measures has proven effective, since there is evidence suggesting that

they have helped reduce tobacco consumption183. Yet, as previously illustrated, to ascribe to each measure

a specific reduction in tobacco prevalence is a difficult task. It is particularly difficult to distinguish the

causes and effects of the ‘cultural’ shift against smoking – first among the middle class – from regulatory,

pricing and other factors184.

By the beginning of the new century, European and other developed countries had implemented most of

these first generation measures, and governments were ready to take a new, historical step. By far the

largest achievement in the fight against tobacco was the conclusion of the Framework Convention on

Tobacco Control as it paved the way to a second generation of tobacco control measures. Addressing

181 Kip Viscusi controversially argued that – as smokers tend to die a few years before non-smokers – states save costs in health coverage, pensions and social security and, as a result, tobacco could be argued to be a net benefit to society. See Kip Viscusi, Regulation through Litigation (Washington, DC: AEI Brookings 2002), pp. 24–25. 182 Thus, for instance, in the EU, the general warnings (‘Smoking kills’ or ‘Smoking seriously harms you and others around you’) should cover no less than 30% of the most visible surface of the unit pack; whereas the additional warnings (to be chosen from an agreed list and to be rotated on a regular basis) should cover no less than 40% of the other most visible part of the packet. See Article 5 of the Tobacco Products Directive.

183 David Hammond, Geoffrey T. Fong, Ron Borland et al., “Text and Graphic Warnings on Cigarette Packages: Findings from the International Tobacco Control Four Country Study”, 32 American Journal of Preventive Medicine (2007), pp. 202–209; Marc C. Willemsen, “The New EU Cigarette Health Warnings Benefit Smoker Whoe Want to Quit : Results from the Dutch Continuous Survey of Smoking Habits”, 15 European Journal of Public Health (2005), pp. 389–392. 184 I would like to thank Adam Burgess for drawing my attention to the class aspect of smoking. Smoking being now a working class habit in the West, this class distinction suggests that conscious efforts at curbing smoking tend to be severely constrained by socio-economic reality. For an excellent illustration, see Eric A. Feldman and Ronald Bayer, “The Triumph and Tragedy of Tobacco Control: A Tale of Nine Nations”, 7 Annual Review of Law and Social Science (2011), pp. 79–100.


complex issues such as smuggling and illegal traffics, besides the more conventional marketing and

packaging aspects of tobacco, this convention identifies a comprehensive series of actions whose

implementations have been further detailed by a set of Guidelines.

These new, radical measures include inter alia standardized plain packaging as well as visual display bans.

While the former implies the removal of trademarks, logos, pictures graphics and other promotional

elements from the pack, the latter consists of prohibiting the display of tobacco products at points of

sale185. Through the adoption of these measures, policy-makers aim not only at diminishing the

attractiveness of the brand logo and graphical world but also at encouraging smokers to pay more

attention to health warning messages. This seems increasingly crucial today because there is evidence

suggesting that tobacco companies’ efforts to associate themselves with positive images are effective in

mitigating the impact of health warnings186. An important focus has developed around challenging the

industry’s ability to advertise its products so as to reduce their appeal and attractiveness among

consumers. Both plain packaging and visual display bans – as well as graphic warnings and smoking bans

– seem potentially capable of nudging citizens away from consumption of tobacco products.

An interesting example of further anti-smoking initiatives is the removal of information regarding tar,

nicotine and carbon monoxide yields (commonly referred to as TNCO)187. By prohibiting the indication

of the TNCO information, the idea is to nudge smokers towards believing that all cigarettes are equally

dangerous. Because smokers (be they current or future) would no longer find any information about

TNCO yields on their packs, they might consider all tobacco products similarly threatening. Yet,

interestingly enough, the factor that triggered the idea that ‘safer’ cigarettes may exist was a reaction to

regulatory action aimed at disciplining tobacco ingredients by, for instance, establishing maximum levels

of TNCO. In particular, the development of low-tar cigarettes has been the tobacco industry’s response to

increased health concerns. It is no surprise that low-tar cigarettes have immediately sparked enduring

scientific debates over their supposedly safer use. In particular, by the 1980s the methods commonly used

to measure tar and nicotine were found to be limited in accurately reproducing smoking behaviour. This

has been ascribed to the fact that smokers are able to take larger, more frequent and higher velocity puffs

than machines do, giving rise to the so called ‘compensation effect’, i.e. compensatory adjustments capable

185 For a detailed legal and policy analysis of plain packaging and visual display ban under EU Law, see Alberto Alemanno, “Out of Sight Out of Mind – Towards a New Tobacco Products Directive”, 18(2) Columbia Journal of European Law (2012), forthcoming. 186 See, e.g., George Lowenstein, “Out of Control: Visceral Influences on Behavior”, 65 Journal Org. Behav. And human. Dec. Proc. (1996), pp. 272–292; Jon D. Hanson and Douglas A Kysar, “Taking Behavioralism Seriously: Some Evidence of Market Manipulation”, 112 Harvard Law Review (1999), pp. 1420–1569. 187 TNCO labelling should be printed on one side of the cigarette packet in the official language(s) of the Member State where the product is placed on the market so that at least 10% of the corresponding surface is covered.


of turning nominally lower ‘tar’ and nicotine cigarettes into ‘higher’ tar and nicotine cigarettes188. Thus,

TNCO information – currently mandated in Europe and most industrialised countries – is not only

incomplete but also positively misleading, as it suggests greater safety of low-tar cigarettes. Policy makers

around the world have responded with bans on the use of descriptors such as ‘light’ and ‘mild’, and the

EU Commission is set to eliminate the obligation of tobacco manufacturers to display the results of the

commonly measured yields from tobacco smoke on products189. It is believed that eliminating the existing

TNCO quantitative labelling might prevent them from being misread by consumers who might “think

that lower levels indicate that a product is less risky to their health” 190. However, the ensuing no-

information policy sits uneasily with another well-established tobacco control policy aimed at establishing

maximum content levels of tar, nicotine and carbon monoxide. Either tar, nicotine and carbon monoxide

yields have a role in tobacco safety or they do not. Tertium non datur.

IV. Tobacco control measures as Nudges

As recently stated, “There is no single science of behaviour change. A number of scientific disciplines,

including neuroscience, psychology, sociology and behavioural economics, contribute to what is known

about human behaviour and we refer to these sciences collectively as the ‘sciences of human behaviour’.

Behaviour change interventions apply findings, drawn from these various sciences, in order to influence

human behaviour”191. But what is a ‘Nudge’?

A nudge, according to its theorisers, “…is any aspect of the choice architecture that alters people’s

behaviour in a predictable way without forbidding any options or significantly changing their economic

incentives”192. However, the exact contours of this innovative approach to policymaking are not clear and

often differ from the standard definition provided for by both academics and public authorities193.

Typically, ‘nudges’ prompt choices without motivating people to consider their options consciously, and

therefore do not include openly persuasive interventions such as media campaigns and the provision of

188 US Surgeon-General Report, The Health Consequences of Smoking (Bethesda, MD: US Department of Health and Human Services 1981). 189 European Commission, Public Consultation Document on the Possible revision of the Tobacco Products Directive (Health and Consumer Protection Directorate-General, 2010).

190 Ibid., p. 7.

191 UK Parliament, “Science and Technology Committee – Second Report Behaviour Change”, 11 July 2011, available on the Internet at <http://www.publications.parliament.uk/pa/ld201012/ldselect/ldsctech/179/17902.htm> (last accessed 24 January 2012).

192 Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth and Happiness (London: Yale University Press, 2008), at p. 6. 193 Ibid., at para 2.5.


information. Furthermore, ‘nudges’ themselves may be provided through regulatory means and thus are

not necessarily an alternative to regulatory tools. In particular, by building upon Luc Bovens’s proposed

taxonomy, a policy instrument qualifies as a nudge when it satisfies the following features:

- its intervention must not restrict choice;

- it must be in the interest of the person being nudged;

- it should involve a change in the architecture or environment of choice;

- it implies the strategic use of some pattern of human irrationality (e.g. cognitive biases);

- the action it targets does not stem from a fully autonomous choice (e.g. lack of full

knowledge about the context in which the choice is made)194.

If measured against this definition, several tobacco control policies seem to qualify as ‘nudges’:

- graphic warnings (typically portraying dramatic images suggesting a direct link between

consumption and death or morbidity);

- plain packaging (implying the removal of brand, colours and other design features of the

product to decrease attractiveness and increase visibility of graphic/textual warnings);

- visual display ban (changing the context of access to the products by hiding them from

public view)195.

Given the increasing salience of the danger of tobacco smoking, none of these policies, not even the

health warnings, is set to correct ignorance. Instead they aim at fighting inertia – by discouraging

automatic behaviour – and akrasia, i.e. the weakness of the will, by breaking social norms. Moreover,

unlike social advertising, these policies are all designed to affect the present choice situation and not

future, indeterminate situations196.

V. A paradigm shift: the behavioural turn of tobacco regulation

194 Luc Bovens, “The Ethics of Nudge”, in Till Grüne-Yanoff and Sven Ove Hansson (eds.), Preference Change: Approaches from Philosophy, Economics and Psychology (Berlin and New York: Springer, 2008). 195 For a detailed legal and policy analysis of these regulatory measures, see Alberto Alemanno, “Out of Sight Out of Mind – Towards a New Tobacco Products Directive”, 18(2) Columbia Journal of European Law (2012), forthcoming. 196 See Luc Bovens, “The Ethics of Nudge”, in Till Grüne-Yanoff and Sven Ove Hansson (eds.), Preference Change: Approaches from Philosophy, Economics and Psychology (Berlin and New York: Springer, 2008).


So what emerges nowadays is a new paradigm in tobacco control policies. The first generation measures

were based on incentives (such as excise duties) as well as on the provision of detailed information (about

product contents and their adverse effects), which assumed that consumers were rational decision makers.

The individual targeted by those policies was a free, responsible, autonomous subject, capable of making

the right choice insofar as she received the correct information. With the emergence of the new

generation of measures, this neoclassical figure of a reasonable decision-maker becomes irrelevant. The

idea is to keep consumers away from temptation, by preventing them from having to make a choice. The

choice itself is in no way presented in a balanced way: the act of buying occurs in an entirely new context.

Those who want to buy a cigarette box face a series of physical and moral obstacles and, as exemplified by

the elimination of TNCO labelling, they often receive less information than in the past about the product

itself. Indeed, in line with behavioural economics findings, policy makers seem to increasingly question the

assumption that consumers will always react rationally to warnings197. There is indeed evidence suggesting

that consumers tend to overvalue immediate benefits and downplay delayed costs.

VI. Testing the validity of Burgess’ concerns about Nudging in tobacco control

This is where some of the criticisms elaborated by Adam Burgess on ‘social architecture’ kick in: To what

extent might ‘nudging’ allow policy-makers to hide the facts in the name of a higher interest? When may

induced behavioural changes be ethically and morally justified? To what extent may the public health

objective compromise the right to be informed? Is the government justified in nudging citizens not to

smoke without providing them facts? How can governments ‘know’ it better? Does ‘nudging’ induce

genuine preference change?

1. Why Nudging smokers?

The success of nudge policies within tobacco control neither seems driven by the economic climate nor

by the attempt to embrace an alternative, less heavy-handed regulatory approach to risk issues, as

illustrated by Burgess. Rather its rapid acceptance within the tobacco control movement seems to have

more to do with the process aimed at counter industry efforts to influence and manipulate individual

choice. This idea of opposing corporate influence can also be traced as one of the main drivers behind the

conclusion of the FCTC198. However, it is true that, as suggested by Burgess’ analysis, these nudge-

inspired tools are designed to depart from the ‘process of risk dramatization’. In line with the objective of

197 See, e.g., Kris N. Kirby and Richard Herrnstein, “Preference Reversals Due to Myopic Discounting of Delayed Reward”, 6 Psyc. Sci. (1995), pp. 83–89. 198 Stephen D. Sugarman, International Aspects of Tobacco Control and the Proposed WHO Treaty, in Robert L. Rabin and Stephen D. Sugarman (eds.), Regulating Tobacco (Oxford: Oxford University Press 2001), at p. 245.


the ‘de-normalisation’ of tobacco, these policies aim at making the individual believe that the social norm

is different, typically lower, than they may have assumed. Thus, plain packaging (whereby a cigarette no

longer looks like a consumer product), visual display bans (whereby a cigarette is no longer assimilated to

other products presented on the shelves) as well as smoking bans (nobody smokes in public places in its

most radical form) portray the smoking reality in a way that differs from what it actually is: tobacco

products are still legally manufactured products that are largely consumed. As sharply suggested, the

objective is to encourage individuals “to adjust their behaviour and eat, drink and smoke less, or consume

less energy”. Therefore, the contribution that nudging might offer to tobacco control is significant

because it favours the de-normalisation of smoking without dramatizing its effects on society. It has

indeed been argued that the social norms approach is gradually emerging as an alternative to the

unsuccessful approach of “health terrorism”199. Instead of scaring people into abstaining from tobacco,

this approach seeks to change the context within which they are exposed to it. This in turn may help

young people to break free from an “imagined false norm of behaviour”200.

2. The libertarian argument

The most immediate argument raised against nudging is that ‘libertarian paternalism’, despite the original

claim of its theorisers201, is an oxymoron because its policies are too paternalistic. However, as noted by

Burgess, the overall resistance to public policies on grounds of individual liberty and autonomy seems to

have weakened in recent times. If attachment to the concepts of liberty and autonomy were dismantled in

the 1990s, the libertarian argument against tobacco control was undermined much earlier and, as a result,

plays only a residual role in the actual debate about tobacco regulation202. Historically, the industry has

systematically opposed tobacco control initiatives as paternalistic and patronizing203. Being that smoking is

deeply rooted in individualism, it has been constantly held that smoking is a voluntary behaviour engaged

in by consenting adults who are well-aware of the risks inherent in their choice204. In Western societies it is

indeed generally believed that individuals should be free to choose their own lifestyles, including the risks

that a lifestyle may entail205. This seems especially true for a product like tobacco (or alcohol) that is

associated with enjoyment, relaxation, celebration and entertainment. Although this ‘libertarian’ argument

199 David J. Hanson, “Social Norms of Marketing is Highly Effective”, available on the Internet at: <http://www2.potsdam.edu/hansondj/YouthIssues/1093546144.html> (last accessed 24 January 2012). 200 Ibid. 201 Richard Thaler and Cass Sunstein claim that, “the libertarian aspect…lies in the straightforward insistence that, in general, people should be free to do what they like…The paternalistic aspect lies in the claim that it is legitimate for choice architects to try to influence people’s behaviour in order to make their lives longer, healthier and better”. See, Nudge: Improving Decisions About Health, Wealth and Happiness (London: Yale University Press, 2008), at p. 232. 202 For a recent attempt, see Scott Crosby, “The New Tobacco Control Directive: An Illiberal and Illegal Disdain for the Law”, 27 European Law Review (2002), pp. 177–193. 203 Interestingly enough, sociological research suggests that the modern anti-smoking movement began, similarly to Prohibitionism of the mid-30s, as a typically upper-middle-class phenomenon intended to impose its values on the lower orders of society by “education”, if possible, or by government intervention, if necessary. 204 Robert L. Rabin and Stephen D. Sugarman (eds.), Regulating Tobacco (Oxford: Oxford University Press 2001), p. 3. 205 Ibid.


has been widely and successfully used by the tobacco industry for some time, it came under attack several

decades ago, thus loosing some of its normative power206. The emergence of evidence showing the health

risks of second-hand smoking shifted the debate from the rights of smokers to the rights of non-

smokers207. The ensuing paradigm shift in developing demand reduction strategies, by freeing the tobacco

movement from this remaining ideological obstacle, served as a catalyst for embracing the more radical

second-generation measures. As a result, despite the existence of some scepticism about the hazards of

second-hand smoke208, the libertarian argument does not seem to weaken as such the normative power of

nudges in tobacco control. By now a consensus has arisen that governmental intervention to reduce

tobacco consumption is “not only legitimate but essential”209. This seems confirmed by the fact that the

tobacco industry itself, facing uncontroversial evidence and an increasingly critical public, no longer

contests the need to curtail youth access and to inform adults about the dangers of smoking.

3. The legitimacy argument

The other criticism generally levelled against nudging policies, and here further elaborated by Burgess,

pertains to legitimacy: who makes the choices about people’s choices? While ‘libertarian paternalism’ is not

an oxymoron, it is inherently technocratic, and it cannot be otherwise since nudging tends to work best

when users are unaware that their behaviour is shaped by choice architecture. Therefore, like any form of

technocratic intervention, the legitimacy of nudges depends on citizens’ acceptance of the manipulative

character of the policy as well as on citizens’ trust in the ‘choice architects’, also called ‘nudgers’. In other

words, how comfortable are citizens with having experts and bureaucrats designing policy to prevent them

from smoking?

Whilst an important question, this argument does seem to loose some normative sharpness when invoked

in the framework of tobacco regulation. Given the general knowledge of the adverse effects stemming

from tobacco consumption, nobody – not even the tobacco industry – seems ready to challenge

governmental action aimed at reducing tobacco consumption on legitimacy grounds210. This is not the

same in other areas, such as obesity prevention, in which due to the multifactorial character of the

problem at stake (e.g. genetic, lifestyle, socio-economic factors) the scientific evidence proving the

206 The evidence about the hazards of second-hand smoke began to emerge in the 1980s. See, Howells, The Tobacco Challenge, supra note 16, at pp. 72–73. 207 The US Surgeon General’s 1972 Report identified the hazards of second-hand smoke before its 1986 Report decisively concluded that, “Involuntary smoking is a cause of disease, including lung cancer, in healthy non-smokers”. In 1998, the UK Scientific Committee on Tobacco found that exposure to second-hand smoke led to a 24 per cent increased risk of lung cancer. See UK Report of the Scientific Committee on Tobacco and Health, 1998, available on the Internet at <http://www.who.int/tobacco/mpower/en_tfi_scientiifc_committee_dh.pdf> (last accessed 24 January 2012). 208 See, e.g., Virginia Berridge, “Passive smoking and its pre-history in Britain: policy speaks to science?”49 Social science and medicine (1999), pp. 1183–1195.

209 Robert L. Rabin and Stephen D. Sugarman (eds.), Regulating Tobacco (Oxford: Oxford University Press 2001), p. 3. 210 On the corporate approach of tobacco companies towards regulation, see, e.g., Patricia McDaniel and Ruth Malone, “Understanding Philip Morris’s Pursuit of US Government Regulation of Tobacco”, 14 Tobacco Control (2005), pp. 193–200. See also, Robert L. Rabin and Stephen D. Sugarman (eds.), Regulating Tobacco (Oxford: Oxford University Press 2001).


existence of a causal relationship between overconsumption and adverse effects is still lacking. Although

our understanding of the causes of both overweight and obesity has increased, there remain scientific

uncertainties: the role played by genetic and ethnicity; the role of physical activities in preventing weight

gain; the changing nature of food; the relative importance of each cause and their interactions211. In these

circumstances, the legitimacy of a nudge aimed at discouraging food consumption seems more limited

than with tobacco control.

4. The short-term argument and the infantilisation effect

Nudges, by aiming at behavioural change through strategic manipulation of cognitive patterns, achieve

their objective neither through education nor personal will. The ensuing risk is that nudges do not build

moral character, but rather weaken our capacity for self-control212. As elaborated by Burgess, this

argument might imply that while nudging may produce benefits in the short term, its long-term effect may

be ‘infantilisation’, i.e. decreasing responsibility in matters regarding one’s own welfare.

However, the history of tobacco control seems to contradict such a conclusion insofar as it suggests that

induced behavioural changes and de-normalisation may also bring about long-term effects. In other

words, nudge-inspired policies such as visual display bans, plain packaging of cigarettes as well as smoking

bans may induce genuine preference changes through a permanent alteration of the context in which

decisions are taken. Given the possibility of individualized reactions to these policies, only an empirical

investigation will be able to show whether the long-term effects will systematically be matched by their

short-term successes.

As for the contended infantilisation process, this seems prima facie of no concern to an anti-smoking

lobby exclusively driven – in line with the public health approach – by the imperative of reducing the

number of smokers. Its policies do not pursue the goal of obtaining informed choices and informed

behavioural changes. Rather they are only driven by the need to reduce the number of smokers. As a

result, their posture vis-à-vis the tobacco issue has no other justification than an instrumental one. This is

exemplified by the actual inconsistent policy pursued and promoted regarding low-tar products. On the

one hand, the legislation establishes maximum levels of tar, nicotine and carbon dioxide contents – thus

suggesting that lower contents might be safer products – and, on the other hand, it prohibits tobacco

manufacturers from listing the amount of tar contents and other ingredients to consumers.

211 Amandine Garde, EU Law and Obesity Prevention (Alphen aan de Rijn: Wolters Kluwer 2011), pp. 10–16. 212 Luc Bovens, “The Ethics of Nudge”, in Till Grüne-Yanoff and Sven Ove Hansson (eds.), Preference Change: Approaches from Philosophy, Economics and Psychology (Berlin and New York: Springer 2008).


In the light of the above, neither the short-term argument nor the infantilisation effect seem relevant to

the extension of nudging policies to tobacco control.

5. The evidence argument

Although Adam Burgess does not seem to share optimism about the ability of ‘choice architecture’ to

increase rational living, he concedes that nudge is presented as a potentially cheaper means of dealing with

social problems213. Yet, as highlighted in his opening essay, no compelling evidence supports this

alternative regulatory approach. As it was recently concluded by the UK Science and Technology Select

Committee, “There is a marked lack of information about what works to change behaviour at policy level”

214. In particular, although most of the proposed schemes are designed to address human misperceptions,

their elaboration is based on experimental settings215. This is certainly the case for plain packaging of

tobacco products, a policy that has never been implemented, but only partly for smoking bans and visual

display bans, which have been in force in several jurisdictions for some time already. Hence, the question

arises whether nudge-inspired measures as applied to tobacco control are capable of changing behaviour?

While it is undisputed that following the entry into force of an overall set of tobacco control measures,

smoking prevalence in the West began to decline after years when it had flatlined216, there has been

insufficient evaluation of the overall strategy to be able to determine the relative impact of each individual

element. This seems particularly sensitive in the area of tobacco control since governments have reporting

obligations as a party to the WHO Framework Convention on Tobacco Control. The protocols and

guidelines to the FCTC provide guidance on the implementation of evidence-based behaviour change

interventions to reduce smoking prevalence and uptake. Although no specific targets are listed, there is a

clear commitment to evidence-based policies. However, the general impression is that, given the highly

instrumental character of the anti-smoking efforts, this commitment is not always taken as seriously as it

should.

VII. Conclusions

213 According to Thaler and Sunstein, “Many of these policies cost little or nothing ; they impose no burden on the taxpayers at all”. See Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth and Happiness (London: Yale University Press, 2008), at p. 14. 214 Science and Technology Select Committee at the House of Lords, Behaviour Change, 2nd Report of Session 2010-12, pp. 18–19. 215 For an early recognition of the lack of both a theoretical and an applied experimental basis for integrating cognitive biases into policy making, see William Eskridge and John Ferejohn, “Structuring Lawmaking to Reduce Cognitive Bias: A Critical View”, 87 Cornell Law Review (2001), at p. 647. 216 Prevalence of current tobacco use among adults aged ≥ 15 years (percentage), World Health Organization. Retrieved 26 January 2012.


The most recent tobacco control policies being inspired by the ‘permit but discourage’ approach all seem

to be supported by ‘libertarian paternalism’: they let individuals smoke but they do not miss a chance to

nudge them towards less consumption in order to reduce morbidity and mortality generally associated

with tobacco smoking. They aim to achieve this not by informing individuals about how harmful smoking

is but by changing the context within which all smoking choices are ‘made’. As such, ‘libertarian

paternalism’ seems capable of accommodating (and justifying) the rather contradictory governmental

stance towards tobacco products, which allows on the one side the legal marketing of a deadly product

capable of generating significant income for both the industry and the government through taxation, while

on the other it heavily regulates and disincentivises its consumption.

While the libertarian argument has lost ground in opposing tobacco control since the emergence of

studies proving the adverse effects of second-hand smoke, the paternalistic argument is not yet fully

established. Although the number of tobacco control policies has progressively increased in number and

significance, there is still resistance to the introduction of tobacco control measures that remain perceived

as contradictory. Libertarian paternalistic-inspired policies may be able to overcome such a last frontier in

the social and intellectual acceptance of tobacco control policies. As demonstrated above, most of the

arguments levelled against nudging do not necessarily weaken the use of this regulatory approach to

tobacco control: manipulation of individual choice and public intervention seem both increasingly

accepted in the area of tobacco control. However, whilst nudging offers a badly-needed intellectual

underpinning to the current regulatory stance against tobacco, it might encounter some of the same

obstacles that it faces in other less contentious areas of policy-making: the effectiveness of its policies and

the risk that, in the long term, they might backfire. Until society decides to ban smoking completely, adults

will continue to make their individual choices. Should these choices not occur in an informed way, there is

a risk that governmental efforts will loose credibility vis-à-vis both current and new, potential smokers. In

sum, despite the specificity of the tobacco industry, this account of the latest nudge-inspired control

measures confirms Burgess’ conclusion: “nudge is only the latest addition to the portfolio of

interventionist approaches, rather than an alternative to it”217. Indeed, by calling for an original regulatory

‘mix’ of adequate intervention, regulating lifestyle choices clearly defies “the power of law to correct all of

the evils”218. There is an emerging consensus that although nudging people along cannot be a self-

sufficient strategy, it should necessarily become part of a broader and multifaceted strategy. Therefore, the

remaining question is how to select the right components in order to define the regulatory ‘mix’ capable of

achieving the intended goals. Unfortunately, neither Thaler/Sunstein nor Burgess provide the recipe for

the optimal ‘regulatory mix’ that would both shift norms away from consumption and use those changing

norms to support more extensive legal intervention which also contributes to reduced consumption.

217 This is also the conclusion reached by the Annual Update 2010-11 of the Behavioural Insights Team (BIT) established at the UK Cabinet Office. See p. 29. 218 Joseph R. Gusfield, Symbolic Crusade (Champaign, IL: University of Illinois Press 1963), at p. 1.


As a result, policy-makers dealing with regulatory efforts to curb tobacco consumption are likely to

continue their desperate search to identify the right regulatory ‘mix’ of policy tools holding the key to

further curtailment. This is likely to remain the case until society determines that the advantages derived

from a ban on tobacco outweigh the costs that it would impose.


Chapter V – Real Nudge

Luc Bovens*

The novelty in Adam Burgess’ paper is that he assesses nudge policies in the context of the shift in the UK government’s approach to risk from the nannying policies of Labour to the nudge policies of the Conservatives. There is a wealth of ideas in this paper. I find it useful to disentangle some of these ideas focusing on the following two questions:

1. In what respects do Labour’s nannying policies and the Conservatives’ nudge policies differ?

2. What is problematic about Labour’s nannying and the Conservatives’ nudge policies?

Subsequently I will reflect on how a particular strand of research in the social sciences can be made relevant to designing a more responsible way of dealing with societal risk and show how this approach can evade some of Burgess’ concerns.

I. Differences between nannying and nudge

1. Regulation and economic incentives versus behavioural modification

Nannying involves regulation to reduce the option set by legislating against risky behaviour or it provides economic incentives and disincentives to make risky behaviour more costly (be it through accounting costs or through opportunity costs). Nudge aims to change the choice architecture, i.e. the environment in which the choice is made, so that people who are placed in this environment would be less likely to display risky behaviour.

2. Rational versus a-rational agents

Nannying rests on the assumption that people behave rationally: a rational agent will avoid risky behaviour when it is met with punitive measures or when it becomes more costly. Nudge rests on the assumption that people behave a-rationally: an a-rational agent changes her behaviour conditioned by environmental cues that she would not be affected by if she were a fully rational agent.

3. Centralised, bureaucratic and costly interventions versus small scale, osmotic and cheap interventions

Nannying is top-down, requires bureaucratic oversight and is costly in terms of public funding. Nudge is bottom-up – it aims to inspire local communities to institute changes in choice environments and to emulate successful initiatives. This approach requires less public funding.

* Professor and Head of Department in the Department of Philosophy, Logic and Scientific Method at the London School of Economics and Political Science.


4. Scare tactics versus social norm enforcement through conformity Nannying intervenes by exaggerating the prevalence and the negative consequences of the risky behaviour. This is in contrast to one particular nudge strategy which intervenes by drawing attention to the fact that the majority of the population does not engage in risky behaviour implying that this is the social norm. This nudge strategy aims to bring about behavioural change by tapping into the common psychological disposition to conform to social norms.

5. Non-laboratory science as evidence base versus laboratory science as evidence base

Nannying is based on evidence that is gathered in the field. This may include evidence about the consequences of risky behaviour (as studied, say, in epidemiology) or evidence about the success of certain interventions to restrict risky behaviour by changing rational expectations (as studied in neoclassical economics). Nudge relies on evidence about how changes in the environment can bring about behavioural modification through a-rational processes. This kind of evidence is typically obtained through laboratory experiments conducted by psychologists and behavioural economists.

6. Adversarial versus cooperative relationship with industry

Nannying’s relationship with industry is adversarial in that it imposes restrictions on production and sales through regulation. Nudge engages industry by coaxing them to institute changes in the choice environment that bring about desirable behavioural modifications.

II. Critiques of Labour’s and the Conservatives’ policies

Burgess raises a number of critiques. Some of them address risk policies in general and apply equally to nannying and nudge whereas others are specific to nannying or to nudge.

1. Politicising risk

The government is overly eager to identify risk in society and may even invent risk factors in order to create the illusion of being proactive in dealing with societal problems. This critique applies both to nannying and nudge.

2. Invasion of privacy


The government transgresses its mandate by intervening in matters that belong to a person’s private sphere. This critique applies both to nannying and nudge.

3. Threat to liberty

Nannying and nudge both pose a threat to liberty. But they do so in different ways. Nannying reduces liberty by reducing the option set or by rearranging our preferences over the option set through adjusting costs and benefits. Nudge poses a threat to liberty as autonomy. It aims to modify behaviour by appealing to a-rational processes. A person who is subject to such a-rational processes is not fully in control of her agency and hence does not act in a fully autonomous manner.

4. Unintended side effects

We do not fully understand how interventions will affect behaviour and how behaviour will affect outcomes. This kind of critique applies both to nannying and nudge. In the case of nannying, interventions may not work because agents may not respond rationally or because we do not fully understand the agents’ belief and desire structures. In the case of nudge, certain behavioural mechanisms identified in the laboratory may not transpose to real-world decision-making. Furthermore, even if we are able to bring about the targeted behaviour, either through nannying or nudge, the behavioural change may fail to bring about the expected outcomes.

5. Infantilisation

An unintended side effect that deserves special attention is the issue of infantilisation and the hindrance of moral development. Both nannying and nudge may have this effect. Regulation as well as environmental cues to discourage or encourage certain behaviour may leave the agent with a lack of moral strength to implement the target behaviour once the regulation or the environmental cues are no longer present.

6. Corruptibility

Nannying is corruptible in that it may lead to government sanctioned bureaucratic sprawl. Nudge may be exploited by self-interested commercial actors or may become mere window-dressing for the government to do nothing about issues of risk in times of austerity.

III. Ways forward: Real nudge

I wish to make a plea for more social science research in which outcomes of risk behaviours are compared across relevantly similar societies. For example, rates of biking accidents, liver disease, diabetes, suicide…could be compared between EU member states. Mechanisms that may be causally responsible


for these differences could then be identified through causal search techniques. Such mechanisms may be cultural features, legislation, economic incentives, features in the choice architecture etc. Some of these mechanisms may be transposable, some may not be transposable. Insight in the causal mechanisms that bring about these differences in outcomes permits us to make informed choices about whether intervention from the government is advisable and what kind of intervention would be most effective.

Comparative causal efficacy studies of regulation and economic incentives are of course commonplace in social policy studies. What is novel here is that I propose to search for features of the choice architecture that are causally responsible for differences in risk behaviours across comparable societies. I propose to name this approach ‘Real Nudge’, since we focus on choice architecture differences that are operative in real-world contexts (and not on policy initiatives that find their inspiration in laboratory experiments).

What are the advantages of such an approach?

1. Real risks

Risk management would become less politicised. It is no longer possible for the government to concoct risks out of thin air and to implement needless policies to give the impression of being proactive. What qualifies as risk-to-be-managed must be underpinned by cross-cultural differences. There is a prima facie cause for concern for certain problem behaviours in our own society only if we can show that we have alarmingly high rates of undesirable outcomes caused by these problem behaviours relative to societies that are similar to ours in many other respects.

2. Non-ideological solutions

The tension between nannying and nudge is entangled in an unproductive ideological debate about the role and scope of government. An appeal to this strand of research in the social sciences to support policy making is non-ideological and does not favour nannying or nudge. Whatever works – be it a mechanism based on regulation or tweaking choice architectures – is open for consideration as a policy measure to be implemented.

3. Solutions that work

Burgess objects that many of Labour’s laws and regulations to manage risk are ineffective. If we restrict the examples in Thaler and Sunstein219 to examples that genuinely qualify as real-life successful nudges then we are left with a rather small and anaemic set. Furthermore, there is little reason to believe that nudge policies that are inspired by research on behavioural patterns observed in the laboratory would be successful in the real world. The advantage of starting from actual outcome differences and uncovering

219 Richard Thaler and Cass Sunstein, Nudge – Improving Decisions about Health, Wealth and Happiness (New Haven, CT: Yale University Press 2008).


causal mechanisms that produce these outcome differences is that these mechanisms have at least some proven track record.

4. Managing transposability

Granted, these causal mechanisms may not be transposable. First, they may include deeply-seeded cultural differences that are simply not transposable. Second, they may be effective in one context and ineffective in other contexts. And third, they may have unintended side effects when transposed to other contexts. But having seen them in operation in some societal contexts should be at least somewhat informative and encouraging. The bridge between relatively similar societies is expected to be less wide than the bridge between rational choice models (for nannying) and the laboratory (for nudge) on the one hand and the real world on the other hand. But clearly, the issue of transposability will need to be decided on a case by case basis with a sensitivity for cultural singularities and through cautious and small scale experimentation.

5. Privacy and Liberty

As to privacy and liberty issues, comparative studies between countries that are roughly similar in their respect for civil rights have a creative potential of revealing mechanisms that are minimally invasive. Certainly, what may pass the bar of respecting privacy and liberty in one liberal democracy may not in another due to cultural differences. Again, this is an issue to be decided on a case-by-case basis.

How would this work in practice? Think of alcohol policy. There is a range of studies that try to link government policies about alcohol pricing, advertising, availability… to alcohol consumption rates and patterns and to alcohol-related morbidity and mortality rates across OECD countries.220 Policy changes should be maximally informed by such cross-cultural studies.

Furthermore, it would be interesting to include features of the choice architecture of alcohol consumption among the potential causes in such cross-cultural studies. Are non-alcohol and low-alcohol beers made available in pubs and off-licences? Are they made available on tap (so that they can be consumed non-conspicuously) in pubs? What food choices are made available along with the beer? What are the serving sizes of drinks? Do customers pay at the time of the order or can they run up tabs? Do customers pay for themselves or is there a culture of rounds? What feedback is made available e.g. through alcohol content labelling and through the sale of breathalyzers? Do any of these choice architecture features causally affect alcohol consumption rates and patterns and alcohol-related morbidity and mortality rates? Nudge initiatives should be informed and backed up by such studies.

Similar cross-cultural effectiveness studies of nannying and nudge policies could be carried out in areas of energy consumption, littering, obesity, education, integration of ethnic minorities … Such studies (and not

220 See, e.g., Donald A. Brand, Michaela Saisana, Lisa A. Rynn et al., “Comparative Analysis of Alcohol Control Policies in 30 Countries”, 4(4) PLoS (2007), pp. 752–759; Alison Ritter “Comparing Alcohol Policies between Countries – Science or Silliness?”, 4(4) PLoS (2007), pp. 616–619; Nick Sheron, Chris Hawkey and Ian Gilmore, “Projections of Alcohol Deaths: A Wake-Up Call”, 377 The Lancet (2011), pp. 1297–1299.


ideological commitments to the proper role and scope of government or philosophical views about the nature of human agency) should provide the backbone for policy initiatives.

alberto alemanno

Documents