We need to promote more responsible corporate and consumer behaviour through greater transparency and by harnessing the insights from behavioural economics and social psychology. (HM Government, 2010)

Introduction

There is increasing doubt that the traditional levers of public policy, regulation and economic instruments (taxes and expenditure) are sufficient tools of government to achieve a desired set of outcomes. This has prompted a growing interest in the role of individual behaviour and how altering that behaviour might help improve policy outcomes. Such arguments were reflected in the policies of the New Labour Government that emphasised greater personal responsibility in healthcare (e.g., tackling obesity) and welfare (e.g., conditionality on unemployment benefit claimants).

They also underpinned much of the policy formulation of the Conservative Party when in opposition, notably in forming proposals around energy efficiency, household recycling and consumer credit. Of particular note here has been the influence of Thaler and Sunstein’s (2008) bookNudge. Since the formation of the Coalition Government the ideas of Nudge appear to have gained a subtle hold on policy makers, as the above quote from the Coalition Agreement reveals.

Nudge has gained considerable political and cross-disciplinary academic interest in the United States and United Kingdom, as it appears to offer politicians the opportunity to recast the relationship between state and society. The influence of these authors’ ideas should not be underestimated: in the United States, Sunstein has been appointed as head of the White House Office of Information and Regulatory Affairs whilst Thaler has been advising the Conservative Party on regulation, and, since the change in government, has been appointed adviser to a revamped Behavioural Insights Team in the Cabinet Office.

This paper provides an extended review of Thaler and Sunstein’s work and its relevance for politicians and policy makers. It starts with an exploration of the main dimensions of their argument (section 2), and then continues with a consideration of its critical reception (section 3), a review of its influence of United Kingdom government policy since the formation of the Conservative Party-Liberal Democratic Party Coalition Government in May 2010 (section 4) and by way of conclusion a summary of its apparent strengths and weaknesses for policy makers (section 5).

A Nudge One Way: about Nudge

Nudge is emblematic of a series of high profile texts from mainly American authors (see also Ariely, 2008; Cialdini, 1998; Shiller, 2000) who explore the importance of behavioural and psychological factors in shaping social, environmental and economic decisions and outcomes. At the heart of these texts is a critique of economic theory which is based on rational assumptions; these authors argue that social and behavioural factors matter, and in particular expound arguments around the significance of a human being’s bounded rationality. In making choices humans do not have ‘full attention, perfect information, unimpaired cognitive ability and complete self control’ (Sugden, 2009). Thaler and Sunstein argue that a third party, whether the state or an independent private agent, can shape, process and filter signals so that humans make ‘better’ choices – either in terms of their own welfare or a collective good (for example, environmental protection).

The central argument of Nudge is around what Thaler and Sunstein call libertarian paternalism. As the authors point out in an earlier article, this is an apparent oxymoron, for libertarians despise paternalism, and vice versa (Sunstein and Thaler, 2003). This position is a powerful critique of libertarian arguments and their central tenets which favour a minimal role for government (Brennan and Buchanan, 1977), arguing instead that it can have more active roles in a range of areas, from commercial transactions to environmental protection to pension provision. As Thaler and Sunstein argue in their earlier work:
Our emphasis is on the fact that in many domains, people lack clear, stable, or well-ordered preferences. What they choose is strongly influenced by details of context in which they make their choice … These contextual influences render the very meaning of the term “preferences” unclear. (Thaler and Sunstein, 2003: 1161)
However, the authors are first and foremost libertarian because they seek to protect individual choice.

Nudging may be through changing the decision structures faced by individuals so that either individual or collective welfare is increased. The authors give the example of differences between national positions on organ donation: countries (e.g., France) which require individuals to opt outof organ donation have considerably higher levels of organ donation than countries where individuals are required to opt in to an organ donation scheme (e.g., the United States and United Kingdom). Both positions, they argue, are consistent with libertarian thinking, but the former is a type of libertarian paternalism. This is also an example of what the authors term ‘status quo bias’ – other things being equal, the force of inertia will mean that humans are likely not to make choices.

The authors encourage governments to nudge, but not to ban or mandate, that is to enforce compulsion. So, under libertarian paternalism there would be no smoking ban in public places, but there would be information on cigarette packets alerting the smokers of the damage they do to others; and speculatively food manufacturers would not be banned, or taxed, for producing unhealthy foods, but retailers encouraged to stock fruit and vegetables in convenient locations and chocolates and crisps in difficult to reach parts of the shop. For libertarians individual choice in these scenarios can be maintained, despite only relatively small improvements to welfare benefits from the changes in behaviour.

Throughout Nudge, Thaler and Sunstein discuss an array of examples which demonstrate the failings of rational economics and the significance of behavioural factors. For instance, drawing on literature on cognitive behaviour they contrast automatic and reflective thinking: both are required for our everyday functioning and existence, but they operate in different ways and over different time periods. Thus, automatic thinking allows humans to make quick decisions in response to immediate stimuli or signals, and base decisions on often imprecise or inappropriate rules of thumb. In contrast, reflective thinking is slower, more deductive and follows established rules and conventions. Here links can be drawn to the Coalition Government’s emphasis in proposals for climate change and environmental sustainability which seek to incentivise or reward certain behaviours, rather than mandate against, for example, carbon emissions.

The authors also highlight the influence of contextual factors on our decision making processes, and in particular the influence of other humans.  Here, Thaler and Sunstein discuss the influence of groups and in particular the tendency to conform.  Therefore, proposals have been made for information to be published on neighbourhood recycling which also shows the contribution of individual households; the rationale being that individuals are by nature competitive and will want to improve their ‘league table position’.

The lessons Thaler and Sunstein draw for policy makers are around what they term choice architecture. The following are the main elements in a positive, welfare enhancing, choice architecture:

setting defaults and the path of least resistance: the core principle here is to overcome status quo bias. Examples here include organ donation (outlined above) and the setting of the default position for pension payments to ‘opt in’ where payments are deducted automatically unless individuals choose to opt out
expect error: human fallibility suggeststhat choices should be structured to minimise the risks of adverse outcomes. Examples given here include ATMs where cards are returned before money is dispensed or self administered medical treatments where a preference is given to taking drugs once a day, the argument being that simple routines minimise error
give (immediate) feedback: central to making choices, the authors believe, is appropriate and timely feedback. The oft cited example here is the use of household smart energy meters where information is provided on energy usage (and expressed in monetary terms), including warnings if usage is above a pre-specified level
mind maps and structuring information:although feedback is important it needs to be structured to allow action. Where pricing signals are complex (e.g., utility or telecommunications tariff structures) the authors suggest using ‘RECAP’ (Record, Evaluate and Communicate Alternative Prices) whereby instead of government regulating prices (a paternalist position), they argue for information to be better structured, timely and comparable
structuring complex choices: the authors highlight that individual consumers need help to make better choices. They point to companies such as Amazon which uses filters to recommend books customers may be interested in, based on their previous purchases and the purchases other customers made who have bought similar books. Again, this reflects the central tenet of bounded rationality
incentives: at this point the authors restate their commitment to freely functioning markets, although they also highlight that there can exist incentive conflicts. The public policy example given here is the use of household thermostats which show the cost of energy being used (an incentive to curtail energy use) as well as room temperature

Thaler and Sunstein’s work has a strong intuitive appeal to policy makers. The support of libertarian paternalism and the notion of the state as a choice architect allow policy makers to make the case for government action which is more responsive to citizens: moreover, throughout the authors’ work there is a strong case made that a nudging state is likely to be a much smaller one. Towards the end of Nudge, Thaler and Sunstein also claim that Nudge provides the real third way, for libertarian paternalism provides the basis for the state’s involvement in limiting the worst excesses of markets, but at the same time calls for a smaller and more specified role for the state in other areas.

Reception and Review

Nudge has been the subject of extensive review by academia and in the media. It has stimulated debate within particular policy domains (for instance, health and environmental protection), academic disciplines (economics, politics, sociology and psychology) and political parties (notably the Conservative Party in the United Kingdom). It is an idea which has garnered considerable interest and is seen as something which should be at the very least part of the policy maker’s toolkit, but it is also contested both on theoretical grounds (notably by Sugden, 2009) and on its empirical basis. In terms of the latter, it is argued that Thaler and Sunstein’s work rests on the application of Nudge principles to very specific policy problems, notably on pension provision and the shift to opt out clauses.

Of the critiques of Nudge, the work by John et al. (2009), Sugden (2009), Amir and Lobel (2009) and Prabhakar (2010) stand out, and this review draws on them to examine three key concerns around: individualism, learning and deliberative democracy; problems for choice architects; andthe absence of inequality and welfare considerations. Each is taken in turn.

Individualism, Learning and Deliberative Democracy

Amongst the instruments Thaler and Sunstein advance, the role of default and opt out come to the fore. They draw particularly here on their work on pensions and savings. However, what is striking is that rational and irrational decision making is essentially an individual act, often in response to limited information and sometimes poor price signals. John, Smith and Stoker (2009) compare a normative political science approach of ‘think think’ based deliberative democracy to the economic libertarian paternalism of ‘nudge nudge’. In contrast to Nudge principles based on individuals, they emphasise that decision making which takes place in groups, with discussion and deliberative action by citizens, potentially leads to better collective decisions. They emphasise that individuals do not have fixed preferences, but rather through expressing different points of view consensus may be reached. The authors point to United Kingdom and international examples such as participatory budgeting and citizens’ juries and panels. Another example would be the New Deal for Communities programme whereby neighbourhood elections were used to select board members of local partnership bodies (see Duffy, Platts-Fowler and Wells, 2008).

In a similar vein, Prabhakar (2010) highlights the role of education and learning in making informed choices, rather than the prescriptive filtering of information by choice architects suggested by Thaler and Sunstein. Again, this contrasts the conservatism and libertarian position of a narrow state-individual relationship with a more social democrat and progressive position of a broader more complex relationship between the state and citizen. What both Prabhakar and John et al. highlight is that whilst Nudge appears to go with the grain of (low cost) decision making, the more costly deliberative approach at least has the potential to challenge deep seated issues and provide the basis for innovation.

Problems for Choice Architects

Amir and Lobel (2009) draw attention to the assertion in Nudge that, ‘in every possible situation a choice architecture already exists [and] therefore, one must begin by asking whether such architecture is appropriate, beneficial or even optimal’ (p. 2114). Such a process would involve asking what mechanism underlies potential biases and what mechanism the potential solution might require. The archetypal case here is where individual decision makers are swamped with too much information: the response may therefore be around defaults (the opt out clause discussed above) or through improving information (the examples of RECAP and structuring complex choices above). As Amir and Lobel highlight (p. 2116), restructuring choices may lead to unintended consequences – for instance, because individuals react adversely to having their decisions manipulated.

More critically, Amir and Lobel argue that the use of defaults and opt outs, the first recourse position and least interventionist actions for libertarian paternalists, is limited and forecloses the use of other instruments such as incentives, deliberation (as discussed), enforcement, education and learning, and agreement of common standards. Through preserving the libertarian position of not infringing on individual choice, Thaler and Sunstein are thus overly concerned with ensuring that there is a near zero-cost to opt out.

Sugden and Amir and Lobel take issue with the notion of the libertarian paternalist choice architecture. A central concern for Thaler and Sunstein is that there exists sufficient information and evidence from which choices can be structured.  Unfortunately, choice is a contingent process from which individuals will not know the full outcomes at the time of decision making. For Sugden this is where the libertarian and paternalist aspects of Nudge pull in competing directions. From a libertarian perspective, even where an individual’s choices are incoherent, they remain the property of the individual and therefore leave unclear scope for the ‘choice architect’s’ paternalist intentions.

Inequality and Welfare

Nudge gives no attention to issues of inequality and welfare. The authors’ concern instead is with a libertarian critique of the state and suggestions for the design of certain policies, but not ones which address inequality. The policy areas where Nudge as a set of principles has been discussed, albeit theoretically, are in areas such as the reduction of obesity, combating knife crime or recycling (Osborne, 2008) and more generally around pensions reform. As Prabhakar (2010) implies, Nudge can be seen to recast the relationship between state and society, and brings to mind the influence of libertarian thinking on the Conservative Party in the 1980s (Gamble, 1994). For Prabhakar libertarian paternalism implies that the state has sufficient information by which to act as a ‘choice architect’ through the range of mechanisms outlined above; but that what the state is nudging is human behaviour, implying that the responsibility for decisions and the consequences of those decisions, however bad, lie with individuals. Under libertarian paternalism the state therefore withdraws from addressing any adverse consequences. The further concern is that scant attention is given to the variation in the capability of individuals to make decisions, let alone the material resources individuals may have to exercise agency or a behavioural change when confronted with a nudge.

Early Signs of Nudging in UK Government Policy

Individual behaviour and the balance between the rights and responsibilities of the state and the individual has been a long standing interest of all main political parties in the UK. For instance, a Cabinet Office paper in 2004 discusses these issues directly and leads to conclusions around a joint responsibility for service delivery and to the personalisation agenda of ‘service co-production’. Whilst the Cabinet Office paper acknowledges risks around widening social inequalities, it does not suggest that these act as a brake on a personalisation agenda.

However, in opposition and in Government, the Conservative Party has embraced the insights of Nudge more wholeheartedly, and allied it to their policy axis of deficit reduction and the Big Society. In September 2010 new life was breathed into the Cabinet Office’s Behavioural Insights Team, which had been established by Tony Blair but had focused on law and regulation under Gordon Brown. Richard Thaler is to act as an adviser to this team. As the quotation from the Coalition Agreement at the start of this paper suggests, behavioural change and behavioural economics forms an area of common interest within the coalition. With a government committed to a withdrawal of the central state from certain areas and a commitment to untying centre-local relations (the localism agenda), the advancement of the Nudge agenda is likely to be driven through welfare reform, pensions policies and the appetite for nudges in some local authorities.

Conclusion

Recent popular texts on behavioural economics have garnered the interest of policy makers from the centre left and centre right.  Nudge comes from the libertarian critique of the state, and propounds the advancement of a smaller state. It therefore provides a key reference point for the Conservative-Liberal Democrat Coalition Government’s agenda of drastic deficit reduction through unprecedented cuts to the public sector. In doing so it also chimes well with the Coalition’s emphasis on individual freedom and civil responsibility. Indeed, the financial crisis, recession and subsequent growth in the deficit may provide a veil to implement a strategy rooted in libertarianism.  Such discourse is redolent in policy pronouncements on welfare reform, pensions and the ‘Big Society’.

This review has also sought to explore the theoretical and empirical basis for Nudge. In both respects it finds a relatively narrow set of theoretical principles and a very limited evidence base to support its ideas. Of course, a deeply theoretical or empirically rich text may not have helped Thaler and Sunstein’s progress towards making the book something of an international best seller or garnering the interest of policy makers. To be fair to the authors, however, policy making, rightly, is not a purely technocratic exercise.

Indeed, the earlier work by Thaler and Sunstein raises important issues across a spectrum of policies, notably on pension reform. The popularity of the text has also led to interest in applying libertarian paternalism to different fields, such as environmental protection. Its ideas should not be written off lightly. Given its libertarian roots it is therefore not surprising that the text gives scant regard to issues of welfare, redistribution and inequality. Issues which highlight structural and geographical factors in explaining inequality are entirely absent.

Libertarian paternalism as outlined by Sunstein and Thaler should not be dismissed out of hand, and indeed the broader ideas of behavioural economics should be of interest to policy makers across an array of policy fields, both in policy design and implementation. However, allied to a libertarian agenda which seeks to curtail the present and future role of the state, it should be assessed more cautiously. It can be seen to ‘lock in’ a state position which blocks off its role in addressing fundamental social and economic inequalities, in policy innovation and in advancing a progressive agenda.

The starting position for policy makers varies. For libertarian paternalists it is essentially one of reducing the state. For centre and centre left, state action is more active and concerned less with mere wealth maximising but also with issues of redistribution – whether purely through fiscal measures (the traditional state role), or more recent variants such as equality of opportunity and levelling up factor endowments and individual capabilities (Sen, 2002), or agendas around asset based welfare and the equality around capital endowments (Prabhakar, 2008). Libertarian paternalism has little to say on these issues.