This is a pre-refereed version of the article that appears in (2012) 75(1) Modern Law Review 122148 that can be found via the publisher’s website at http://onlinelibrary.wiley.com/]
Review Article
Nudge as Fudge
Richard Thaler and Cass Sunstein
Nudge – Improving Decisions About Health, Wealth and Happiness
(Penguin Books, London 2008)
Karen Yeung∗
1.
Introduction
Nudges are often used in my household. Directed primarily at my two year old daughter, these
practices range from pointing to other children riding happily in their pushchairs in order to prompt her
to climb aboard her own, reminding her of the fun she had the previous day playing with a particular toy
to enhance her interest in it, to fashioning vegetables into animal shapes to encourage her to eat them.
What unites these diverse practices is their common aim: to elicit my daughter’s co-operation in
complying with my (typically unarticulated) wishes without resort to more punitive, coercive
techniques which generally end in tears and tantrums. My techniques are, I suspect, as familiar to
parents everywhere as they are to policy-makers. The carrot shaped in the form of a duck and the speed
hump installed in a residential street can both be understood as techniques that deliberately seek to elicit
a particular behavioural response from another, whilst formally preserving the latter’s freedom of
choice. While there is arguably nothing particularly new or unusual about these techniques, what is
novel is the claim by various scholars that their effectiveness (at least when directed at adult decisionmakers of ordinary mental competence) rests on scientific foundations. More specifically, recent
research into the ‘science of decision-making’ has spawned a new academic sub-discipline often referred
to as ‘behavioural economics’ and its off-spring, ‘behavioural law and economics’.
Although Richard Thaler is merely one among many economists working from within this burgeoning
field, it is his partnering with lawyer and legal academic, Cass Sunstein, to produce the best-selling and
highly readable paperback, Nudge – Improving Decisions About Health, Wealth and Happiness1 that has the
∗
Professor of Law, Director of the Centre for Law, Ethics and Technology in Society (‘TELOS’), King’s
College London. I am indebted to Roger Brownsword, John Coggon and an anonymous reviewer for comments
on earlier drafts. All errors remain my own.
1
R. Thaler and C. Sunstein. Nudge (Penguin Books, London 2008) (hereafter ‘Nudge’).
1
catapulted the ideas underpinning so-called ‘nudge’ techniques into political prominence. Although the
book first appeared in 2008, this review timely in light of its extraordinary success in recent years,
capturing the imagination of politicians from both ends of the political middle-ground and spawning a
plethora of nudge-based policy proposals across a varied range of social sectors. In the UK, David
Cameron has energetically promoted the use of nudges as a technique for solving social ills, claiming that
Nudge ideas can be adopted to implement the policies of his “Big Society”2, marked by the establishment
of the ‘Nudge Unit’ (formally known as the Behavioural Insight Team) within the Cabinet Office.3 On
the other side of the Atlantic, the Democratic Obama administration’s active endorsement of Nudge
strategies are evident in his appointment of Cass Sunstein as his so-called ‘Regulation Czar’, head of the
White House Office of Information and Regulatory Affairs4, and in Obama’s recent promulgation of
Executive Order EO 12866.5
According to Thaler and Sunstein, a nudge is
‘an aspect of choice architecture that alters people’s behaviour in a predictable way without forbidding any options
or significantly changing their economic incentives’.6
The idea of nudge is best grasped by reference to specific examples, rather than by formal definition.
One of Nudges’ most frequently cited examples is the etching of the image of a housefly into the men’s
room urinals at Amsterdam’s Schipol Airport, which is intended to ‘improve the aim’.7 Apparently, the
fly etchings have reduced spillage by 80%.8 Other examples include arranging the food in a cafeteria so
that the healthy items are displayed prominently at eye level, with the fattier, sugar-laden options
2
P. Omerod, 'A Network Is As Good as A Nudge for a Big Society' Financial Times (London, 15
September 2010).
3
Headed by behavioural economist David Halpern, it has been directed to focus on problems of obesity,
diet and alcohol, to work alongside the Health Secretary's Responsibility Deal Behaviour Change group: F.
Lawrence, 'First Goal of David Cameron's Nudge Unit is to Encourage Healthy Living' The Guardian (London, 12
November 2010).
4
J. Weisman and J. Bravin, 'Obama's Regulatory Czar Likely to Set a New Tone' Wall Street Journal (New
York, 8 January 2009).
5
Section 4, titled “Flexible Approaches” provides that, “Where relevant, feasible, and consistent with
regulatory objectives, and to the extent permitted by law, each agency shall identify and consider regulatory
approaches that reduce burdens and maintain flexibility and freedom of choice for the public. These approaches
include warnings, appropriate default rules, and disclosure requirements as well as provision of information to the
public in a form that is clear and intelligible.” See B. Obama, 'Towards a 21st Century Regulatory System' Wall
Street Journal (New York, 18 Jan 2011).
6
Nudge, 6.
7
Nudge, 4 and 91.
8
Nudge, at 261 citing http://www.coathanger.com.au/archive/dibblys/loo.htm.
2
displayed further back in order to encourage customers to choose the healthy options (‘Cafeteria’)9,
automatic enrolment of employees into a scheme which commits them to allocating future salary
increases into a retirement savings plan (‘Save More Tomorrow’)10, and the painting of white stripes on
road bends spaced more closely together at the most dangerous points to create the illusion that the
vehicle’s speed is increasing thereby prompting the driver to brake before the apex of the curve
(‘Painted Road Stripes’).11 According to Thaler and Sunstein, the core feature of these (and a host of
other policy interventions which they advocate) is their reliance upon ‘choice architecture’, referring to
the conscious and deliberate attempt to shape the context in which people make decisions.12 In
advocating the use of choice architecture to ‘improve’ individual decision-making, Thaler and Sunstein
seek to establish ‘our new movement’ which they dub ‘libertarian paternalism’.13
While the ‘paternalistic’ dimensions of Thaler and Sunstein’s policy proposals have sparked lively debate
about the virtues and vices of paternalism, which is currently enjoying something of a renaissance in
legal scholarship, fuelled by the growing popularity of behavioural law and economics14, it is clear that
many of the proposals advocated in Nudge are concerned with shaping other-regarding decisions in order
to promote collective welfare rather than with ‘improving’ an individual’s self-regarding actions.15 So,
for example, the benefits arising from spillage reduction in Schipol airport urinals accrue not to those
whose aim is thereby improved, but to subsequent travellers who encounter clean facilities. Likewise,
although the customer who chooses the fruit salad rather than the chocolate cake enjoys direct health
benefits, the broader community also benefits indirectly through savings in the healthcare and other
costs associated with obesity-related health conditions. In other words, nudge strategies can be directed
at individuals’ other-regarding actions and decisions, to which objections to paternalism do not apply,
and need not be (or not merely) directed towards the self-regarding decisions of individuals.
Accordingly, my primary concern in this review essay is to interrogate the legitimacy of ‘choice
architecture’ as a means for implementing public policy, although in so doing, some of the uncertainties
9
Nudge, 1-4.
10
Nudge, 118-9.
11
Nudge, 41.
12
Nudge, 3.
13
Nudge, 6.
14
M. A. Edwards, 'The FTC and the New Paternalism' (2008) 60 Administrative Law Review 323, at 323.
15
O. Amir and O. Lobel, 'Stumble, Predict, Nudge: How Behavioural Economics Informs Law and Policy'
(2009) 108 Columbia Law Review 2098; R. Korobkin, 'Libertarian Welfarism' (2009) 97 California Law Review 1651;
M. A. Smith and M. S. McPherson, 'Nudging for Equality: Values in Libertarian Paternalism' (2009) 61
Administrative Law Review 323.
3
inherent in the concept of paternalism will be brought to light. I will argue that although Nudge contains
some quite sensible, valuable ideas, Thaler and Sunstein make a number of ambitious and
unsubstantiated claims that do not withstand critical scrutiny. Foremost among them is the claim that
nudges are liberty-respecting. Yet despite the serious shortcomings of Nudge as a set of empirically
grounded, theoretically rigorous policy prescriptions, it is unlikely to dampen its popularity with
politicians or public administrators given the allure of simple, cheap and effective solutions to policy
problems.
2.
Nudge’s Intellectual Heritage: Experiments in Cognitive Psychology
The theoretical foundations of Nudge rest primarily on findings from laboratory experiments conducted
by cognitive psychologists concerned with understanding human decision-making. These studies, which
Thaler and Sunstein describe as the “emerging science of choice”16, demonstrate considerable divergence
between the rational actor model of decision-making that is assumed in microeconomic analysis, and
how individuals actually make decisions. The resulting discrepancies are often somewhat pejoratively
labelled as cognitive “defects”. A basic understanding of these defects is required in order to undertake a
critical assessment of the nudge proposals which are claimed to spring from them. For the purposes of
this review essay, these defects are classed into four groups, (a) decision-making heuristics and judgment
biases; (b) akrasia; (c) endogenous preferences and framing effects and (d) information problems.
2.1
Decision Making Heuristics that Generate Judgement Bias
Heuristics are tools for learning. Decision-making heuristics are mental devices or shortcuts that are
used by individuals in making decisions. They are often unreflective, some even unconscious, yet they
are pervasive in human decision-making. To economise on the mental costs of concentration,
information acquisition, and evaluation, we employ judgment heuristics that enable us to make decisions
efficiently. One only need compare the ease and skill with which an experienced driver negotiates
traffic in all but the most exceptional of driving conditions with the level of mental and physical effort
required of a novice driver to display quite basic driving competence, to understand the importance of
heuristics in managing the complexities of daily decision-making. But these instinctive judgments are
not always accurate. The seminal work is by Kahneman and Tversky, who devised a series of
experiments which identified situations in which people’s heuristics violated the rules of probability
theory and the axioms of rational choice 17. The first three chapters of Nudge provide a highly readable
16
Nudge 8.
17
A. Tversky and D. Kahneman, 'Judgment Under Uncertainty: Heuristics and Biases' (1974) 185 Science
1124-1130.
4
account of many of these decision-making flaws, employing an engaging blend of familiar examples and
personal anecdote to explain some of the findings emerging over the last four decades.18 Some of the
more well-known heuristics relied upon by Thaler and Sunstein are summarised below.
According to the availability heuristic, we tend predict the frequency of an event, or a proportion within
a population, based on how quickly past instances can be brought to mind. For example, in the
aftermath of a natural disaster, purchases of insurance covering that kind of disaster rise sharply, but
then rapidly decline as vivid memories recede.19
Similarly, the anchoring heuristic refers to the common human tendency to rely too heavily, or "anchor"
on one trait or piece of information when making decisions, and this influences the way people
intuitively assess probabilities. According to this heuristic, people start with an implicitly suggested
reference point (the "anchor") and make adjustments based on additional information to it to reach their
estimate. In one of Tversky and Kahneman’s early studies, people were asked to guess the percentage of
African nations that were members of the United Nations. Those who were first asked "Was it more or
less than 10%?" guessed lower values (25% on average) than those who had been asked if it was more or
less than 65% (45% on average).20
The representative heuristic refers to the tendency of individuals to judge the probability or frequency of a
hypothesis by considering how much the hypothesis resembles available data, as opposed to using
rational probability-based calculations. This heuristic explains the “hot hands fallacy”, by which many
basketball fans believe that players go on winning streaks, such that a player is more likely to shoot a goal
if that player’s previous shot (or, even better, last few shots) were successful: yet there is no evidence of
this, each shot is independent of the previous shot.21
Many lawyers and academics will, I suspect, be personally familiar with optimism bias. This refers to the
systematic tendency for individuals to be unduly optimistic about the outcome of planned actions,
thereby over-estimating the likelihood of positive events and under-estimating the likelihood of negative
event. Hence we should not be surprised by our systematic tendency to under-estimate the length of
time it will take to complete a project, whether it be writing a review article or packing for a holiday.
18
Thaler and Sunstein discuss only a small subset of these. For a more extended discussion, see F. H.
Buckley, Fair Governance - Paternalism and Perfectionism (Oxford University Press, Oxford 2009) at Chapter 3.
19
Nudge 28.
20
Ibid.
21
Nudge 31-32.
5
Equally familiar is status quo bias, referring to our tendency not to change an established behaviour unless
the incentive to change is compelling. This phenomenon can be readily exploited by commercial service
providers, particularly providers of energy, telecommunications or insurance services, who exploit the
tendency of subscribers to remain with their existing provider after the conclusion of the service
contract, even though considerably cheaper substitutes are available. According to some behavioural
economists, status quo bias may create an endowment effect, which suggests that people value a good or
service more once their property right to it has been established, placing a higher value on objects they
own than objects that they do not. For example, in one experiment, people demanded a higher price for
a coffee mug that had been given to them but put a lower price on one which they did not yet own.
The decision-making ‘flaw’ that is relied upon most frequently by Thaler and Sunstein is the so-called
“framing effect”, which describes the way in which presenting the same option in different formats can
alter people's decisions. Hence, if asked to choose between treatment options in which is Option A is
presented as having a 70% chance of success, and Option B which is presented as having a 25% failure
rate, a significant number of individual will select Option A, even though Option B has a higher
probability of success (75%).22
2.2
Endogenous Preferences
Laboratory experiments clearly demonstrate that in many situations, individuals’ preferences are
endogenous, in that they are influenced by the choices of others.23 Unlike decision-making heuristics,
however, the resulting decisions may not necessarily result in choices that would depart from the
rational actor model of decision-making. There are many sound reasons why it may be rational to do
what others are doing: we may rationally believe that the behaviour of others signals valuable
information. A rather commonplace example can be found in the phenomenon by which, in deciding
which of two equally appealing adjacent restaurants to dine in, one of which is empty, and the other
with a handful of customers. Other passers-by then choose the restaurant which is currently patronised,
reasoning that other people consider it the better choice. Yet it may be the case that none of the diners
have any additional information available to them in making their decision, in which case the popularity
of one restaurant over another is explained wholly by reference to the social influence of other people’s
choices (known as ‘information cascades’).
2.3
Akrasia or Weakness of the Will
22
A. Tversky and D. Kahneman, 'The Framing of Decisions and the Psychology of Choice' (1981) 211
Science 453-458.
23
See generally Buckley, supra n. 18, chapter 7.
6
Unlike decision-making heuristics, which involve some kind of mental or computational error, akrasia
refers to the state of acting against one’s better judgment. For example, I may clearly prefer to eat a
healthy meal, yet when lunchtime arrives I succumb to temptation and choose the lasagne rather than
the salad. Whether akrasia, or weakness of the will, should properly be regarded as ‘irrational’ is highly
debatable. There may be some decisions that can be partly explained by decision-making biases, such as
the hot-cold empathy gap.24 This refers to an individual’s tendency to underestimate the influence of
visceral drives rather than to other non-visceral factors, so that an individual may resolve not to drink
too heavily at a social engagement, but is unable to refuse when the host offers a top up of wine. On the
other hand, these decisions may well reflect an individual’s rational assessment of the value of present
and future costs and benefits. Hence I may rationally decide that the enjoyment of consuming the
lasagne today outweighs the benefits of being marginally slimmer and healthier tomorrow. In other
words, pleasure has a temporal dimension.25 Yet, in contrast to their discussion of decision-making
heuristics, Thaler and Sunstein’s discussion of “temptation” is remarkably thin, particularly given the
instability and heterogeneity of human preferences.26 Their brief discussion simply refers to the
tendency of individuals to choose “mindlessly”27, and is primarily concerned with outlining a series of
practical “self-control strategies” that can help an individual resist temptation, such as an alarm-clock
that runs around the room if the snooze button is pressed after the alarm sounds to prompt the snoozer
to get up and find the alarm clock, and making reciprocal commitments to transfer something of value
to another if she fails to attain her specified goal.28
24
Nudge, citing G. Loewenstein, 'Out of Control: Visceral Influences on Behaviour' (1996) 65
Organizational Behavior and Human Decision Processes 272-296.
25
Buckley, supra n 18, 38.
26
R. Hollander-Blumoff, 'Law and the Stable Self' (2010) Saint Louis University Law Journal 1173; C. A. Hill,
'The Rationality of Preference Construction (and the Irrationality of Rational Choice)' (2008) 9 Minnesota Journal of
Law, Science & Technology 689.
27
This claim is backed up by a reference to B. Wansink. Mindless Eating: Why We Eat More Than We Think
(Bantam, New York 2006).
28
Nudge, 47-50. Thaler and Sunstein’s discussion of self-control strategies has its intellectual roots in the
work of an earlier generation of economists working within the rational choice tradition who were concerned with
understanding the phenomenon whereby an individual may employ various ‘precommitment’ (or ‘self-binding’)
strategies, restricting their own future freedom in order to protect themselves from succumbing to the temptation
to engage in actions that they would prefer to avoid. So, for example, an individual committed to giving up
smoking may resolve to avoid keeping cigarettes in the house, knowing that otherwise the urge to give in to his or
her desire to smoke will be too difficult to resist. See T.C. Schelling, ‘Self- Command in Practice, in Policy, and
in a Theory of Rational Choice’ in T.C. Schelling, Strategies of Commitment and Other Essays, 2006 (Harvard
University Press, Cambridge MA) 63-81; J Elster, Ulysses Unbound: Studies in Rationality, Precommitment and
Constraints, 2000 (Cambridge University Press, Cambridge) and T Kuran, Private Truths, Public Lies: The Social
Consequences of Preference Falsification, 1995 (Harvard University Press, Cambridge MA).
7
2.4
Information Problems and Bounded Rationality
Somewhat surprisingly, Thaler and Sunstein’s survey of decision-making flaws does not include an
explicit discussion of the problem of information deficits and bounded rationality, even though some of
the proposals they advocate can be readily understood as attempts to overcome these problems.29
Individuals sometimes agree to unduly burdensome obligations that do not accurately reflect their
preferences towards those obligations, due to the excessive search costs involved in finding the
information necessary to make a properly informed judgment. This is often the case in relation to
financial decision-making, where an enormous volume of information may need to be acquired and
processed in order to make a fully informed decision. Hence an individual might rationally decide that
the benefits of undertaking a comprehensive search are not justified by the likely benefits that would
accrue relative to an attenuated evaluation based on more limited information.
2.5
Behavioural Law and Economics
The findings of experimental cognitive psychologists identifying these and other systematic decisionmaking flaws have been seized upon by economists, generating a body of work which has become known
as ‘behavioural economics’ and its off-spring ‘behavioural law and economics’ (or the ‘new law and
economics)’. Unlike orthodox law and economics methodology, new law and economics seeks to
challenge the standard economic model by pointing to systematic divergences from the premise of the
rational self-interested decision-maker that orthodox economic modelling takes as its starting point.
Thaler and Sunstein represent these divergences by contrasting two kinds of actor: real people or
‘humans’ and ‘econs’ (a short form for ‘economic man’ or homo economicus’). While the latter are actors
who think and choose rationally and thus fit the picture of the decision-maker assumed by traditional
economics, the former are fallible individuals who have inescapable difficulties in making decisions that
conform to the rational actor model.30 By taking account of the systematic flaws in human decisionmaking identified in laboratory experiments conducted by cognitive psychologists, proponents of
behavioural law and economics hold out the prospect of a new and improved approach to legal and
social policy. Their ambition is to develop an approach to legal and social policy questions that
integrates the findings from cognitive psychology into an economic framework in order to yield an
approach with greater descriptive accuracy and predictive reliability.31 It offers the prospect of
29
See Buckley,supra n 18, chapter 5.
30
Nudge 7.
31
T. Rostain, 'Educating Homo Economicus: Cautionary Notes to the New Behavioral Law and Economics
Movement' (1999) 34 Law & Society Review 973-1006, 979.
8
accurately modelling how human beings will react to varying rules and policy interventions. Small
wonder that it has proved so attractive to politicians of various stripes as a set of policy prescriptions for
curing social ills.
3.
Choice Architecture as a Tool of Government
Nudge is an exemplar of the ‘new’ law and economics, albeit presented in a popularly accessible form
rather than written primarily for an academic audience. It draws on the findings of experimental
cognitive psychology as the basis for policy prescriptions that are claimed to lead to predictable and
replicable alterations in human behaviour and, hence, social outcomes. The core idea underpinning
nudge policies are grounded in the notion of ‘choice architecture’, referring to the conscious and
deliberate attempt to shape the context in which people make decisions, rather than altering or extending
the available range of choices.32 According to Thaler and Sunstein, choice architects are those
responsible for organising the contexts in which decisions are made, including doctors who describe
alternative treatments available to patients, the food services director of a cafeteria who is responsible
for the layout of food items, and a parent describing possible educational options to his or her child.33
Direct parallels are drawn between the way in which building architects make design choices they
expect will yield beneficial effects: just as open stairwells are likely to produce more workable
interactions and more walking, the food services director of a cafeteria can choose a particular food
arrangement that is likely to influence what people will eat.34 By drawing on the cognitive
psychologists’ findings concerning systematic human decision-making behaviours, choice architecture
appears to offer a potentially powerful instrument to influence social behaviour by seeking to prompt or
‘nudge’ individuals to make decisions that the choice architect deems desirable.
But not all forms of choice architecture are nudges.
‘To count as a mere nudge, the intervention must be easy and cheap to avoid. Nudges are not mandates. Putting
the fruit at eye level counts as a nudge. Banning junk food does not.’35
Hence the installation of a six foot perimeter wall around a new housing development with only one
open-access entry point might constitute a form of choice architecture, but because alternative access
32
Nudge 3.
33
Ibid. Although Thaler and Sunstein’s policy proposals are directed at the use of choice architecture by
state actors as an instrument for implementing public policy, it is important to note that choice architects are not
limited to state actors.
34
Nudge 4.
35
Nudge 6.
9
strategies are not cheaply and easily available, it would not count as a nudge. In contrast, filling in and
returning a form to opt out of a default savings scheme, or reaching over the fruit to obtain the
chocolate cake, impose relatively trivial costs on those who wish to avoid the default options, and thus
constitute nudges. Thaler and Sunstein provide their view of the policy implications of systematic
cognitive defects for choice architecture in Chapters 4 and 5, culminating in the following set of ‘basic
principles for choice architecture’ which choices architects are enjoined to follow:
ƒ
Incentive effects of particular choice architecture should be considered (particularly who uses,
who chooses, who pays and who profits);
ƒ
Understand that good choice architecture helps people to improve their ability to map
alternative options;
ƒ
Default options are ubiquitous and powerful. Hence, careful attention to default options is
important because a large number of people can be expected to end up with the default option;
ƒ
Give feedback to help improve individual decision-making; and
ƒ
Error is ubiquitous and expected, and thus should be taken account in designing the decisionmaking context.
These policy prescriptions are then summed up in a simple mnemonic– ‘nUDGE’.36
Part II of the book proceeds to elaborates on these core ideas, comprised of a set of more detailed nudge
proposals focused primarily around measures concerned with financial decision making (Chapters 6 to 9)
and a range of disparate proposals which are grouped together under the broad heading ‘society’ in
Chapters 10 to 13, concerned with prescription drugs plans (Chapter 10), increasing organ donations
(Chapter 11), reducing environmental degradation (Chapter 12) and default rules to apply in the event
of marital breakdown (Chapter 13). A further series of small nudge suggests are strung together and
presented in Chapter 14 under the heading ‘A Dozen Nudges’.
3.1
A Typology of Nudges
In order to interrogate the legitimacy of ‘choice architecture’ as an instrument of government, the
nudges which Thaler and Sunstein describe can be roughly classified into three groups, according to the
36
Nudge 109.
10
underlying architectural mechanism through which they are intended to work: setting defaults and
anchors, physical architecture, and deliberation tools. Figure 1 summarises the nudges described in the
book according to class while identifying the policy goal which each nudge is designed to promote.
(a) Defaults and Anchors
Defaults and anchors seek to exploit several decision-making heuristics, particularly status-quo bias (the
tendency not to change an established behaviour unless the incentive to change is compelling), the
human tendency to “anchor” on one trait or piece of information when making decisions and the
endogeneity of preferences (the tendency of individual decision-making to be influenced by the decisions
of others). For example, both Save More Tomorrow and Cafeteria are intended to encourage more
conscientious saving and healthier diet choices respectively, by setting baseline defaults that reflect these
preferences.37 Similarly, tax-payers can be encouraged to file their tax returns by advising them of the
fact that 90% of the public comply with tax laws, in the same way students may be discouraged from
smoking if they are advised of the fact that most college students do not smoke.38 Defaults and anchors
are premised on the basis, confirmed by studies in experimental psychology, that many decisions which
individuals confront are taken in a passive and unreflective manner, rather than making an active,
conscious decision following a process of careful reflection and evaluation. It is this behavioural
tendency to ‘do nothing’ that makes the default option ‘ubiquitous and powerful’,39 and effect which can
not only be harnessed by policy-makers, but also magnified if combined with some implicit or explicit
suggestion that it represents the normal or even the recommended course of action.40
(b) Physical Design
The image of the fly etched into the Schipol airport urinals clearly demonstrates how the design of the
physical environment, and the household products and artefacts that are employed by humans in their
regular activities, can be designed in ways that will promote particular kinds of social outcomes. Thus,
stripes can painted on roads to create the illusion that the vehicle is speeding up when approaching
dangerous bends,41 petrol caps can be attached to motor vehicles to avoid drivers leaving them at petrol
37
Nudge 1, 117-119.
38
Nudge 74.
39
Nudge 93.
40
Nudge 93.
41
Nudge 41.
11
stations when refuelling,42 and anaesthetic connectors can be designed to preclude drugs being
inadvertently delivered via the wrong route.43
(c) Deliberation Tools
In a rather different vein, choice architecture can facilitate more informed, thoughtful decision-making
aimed at helping individuals to comprehend the range of options available. These include government
information campaigns, intended to provide individuals with information that will help inform their
decisions, as well as mandatory disclosure laws, such as truth in lending laws requiring the uniform
publication of annual percentage return (APR) borrowing rates to make the cost of borrowing more
comprehensible to individuals. The law can also mandate that a specific ‘cooling off period’ should be
available following certain kinds of transactions that are deemed by policy makers to be associated with a
significant risk that the transaction will be unduly burdensome, or fail to reflect individual’s more
considered choices.44
Figure 1
Nudges and Their Policy Goals
Nudge
A. Defaults and Anchors
Automatic enrolment of new employees into
employer pension savings plan (Ch 7, Ch 9 )
Intelligent assignment of prescription drug plans
(p 180)
Default healthcare plans that reflect the needs of
the majority
Default pension plans that consist of diversified
securities portfolios (Ch 7, Ch 9 )
Mandated choice for organ donation
Laws requiring credit card companies to allow for
automatic payment of full monthly amount due,
not merely minimum payment (p 150)
Default legal rules for couples making a
commitment to each other in the event of
separation
Automatic tax returns, already filled in with
default information (p 229)
Set default charity donation options (pp 26-27;p
227)
42
Nudge 97.
43
Ibid.
Policy goal
To encourage individuals to save more for
retirement
To encourage the choice of drug prescription plans
more suited to each individual’s health needs
To encourage the adoption of health plans more
suited to each individual’s healthcare needs
To encourage higher and more secure pension
plan returns for employees
To encourage higher organ donation rates
To protect individuals from highly unfavourable
loan contracts
To protect the interests of the financially weaker
party to intimate relationships (Ch 13)
To encourage timely lodgement of tax returns
To encourage more generous charity donations
44
Thaler and Sunstein’s support for mandatory disclosure laws sits somewhat uncomfortably with their
expressed dislike of coercive policy instruments from which individuals and firms cannot opt-out: see Nudge 5.
Yet they do not elaborate on their vision of the proper role of legal coercion in public policy.
12
Charity Debit Card to encourage tax effective
giving (p 228)
Information campaigns to encourage energy
conservation (p 40; pp 204-5)
Social advertising (Don’t Mess with Texas) pp 645
Advise people that 90% of public comply with tax
laws (p 72);
Advise students that most college students don’t
smoke (p 74)
Non-verbal clues (smells, smiley faces to indicate
approval (p 75); photos of various quality housing
to indicate savings targets on pension plans (p 138)
Prime people before they make decision (pp 76-8)
To encourage more tax effective charitable giving
To encourage energy conservation
To discourage littering on public highways
To encourage timely lodgement of tax returns
To discourage college students from smoking
To encourage the action for which non-verbal
approval is designated (eg higher retirement
savings, energy conservation)
To encourage people to act in the manner primed
B. Physical Design
Automatic safety mechanisms
No-bite nail polish (p 233)
Disulfiram (for alcoholics) which causes a
person to throw up and experience a
hangover on consuming any alcohol (p 233)
Painted road lines to create the illusion of
speed (p 41)
Glow ball energy meters
Car gas cap attached to car (p 97)
Diesel fuel nozzles do not fit standard fuel
tanks (p 97)
Non-luer compatible connectors for drug
delivery (p 97)
Paint ‘look right’ on roads in UK (p 99)
Shutter click on digital cameras to indicate
that image has been captured (p 99)
Arrangement of rings and knobs on 4-burner
stove top (p 92)
The ‘eco-pedal’ for car accelerators, that
indicates to the driver how much fuel
efficiency could be saved by easing the
pressure (p 208)
Motor cycle helmets not mandatory with
those with a special driving qualification (p
231)
Automatic warning if foul language detected
in the contents of an email which has not yet
been sent (the civility check– pp 233-4)
To reduce the risk and severity of accidents
To discourage nail biting
To discourage alcohol consumption
To encourage drivers to slow their speed around
dangerous road bends
To encourage energy conservation
To prevent drivers leaving petrol caps at petrol
stations after re-fuelling
To prevent drivers from pouring diesel fuel into
petrol fuel tanks
To prevent wrong route drug administration
To reduce the risk of pedestrians from countries
which drive on the right from walking in front of
vehicles approaching from the right
To alert the photographer that an image has been
digitally captured
To reduce accidents from errors in the use of
stovetop rings
To encourage energy-conserving driving practices
To protect motorcycle drivers
To encourage civility in email communication by
forcing reflection before sending an email
C. Deliberation Tools
Mandate information disclosure in specified
formats deemed comprehensible (provides
information that could be useful to informed
To enhance individual’s understanding of complex
information before engaging in potentially
burdensome transactions
13
decision-making). eg mandatory simplified
reporting of borrowing costs (pp 146-7)
Cooling off periods following certain kinds of
transactions
To encourage more reflective decision-making
Naming and shaming of those found in
violation of specified norms
To discourage norm violation
Source: R Thaler and C Sunstein, Nudge (2008)
3.2
Are Nudges Libertarian?
Taken together, the simplicity and appeal of nudge strategies becomes readily apparent. This appeal is
enhanced by their “libertarian” credentials, albeit of a paternalistic hue, with Thaler and Sunstein
claiming that,
The libertarian aspect of our strategy lies in the straightforward insistence that, in general, people should be free to
do what they like – and to opt out of undesirable arrangements if they want to do so....We strive to design policies
that maintain or increase freedom of choice. When we use the term libertarian to modify the word paternalism,
we simply mean liberty-preserving. And when we say liberty- preserving, we really mean it. Libertarian
paternalists want to make it easy for people to go their own way; they do not want to burden those who want to
exercise their freedom’.45
Yet the idea of ‘libertarian paternalism’ which underpins Thaler and Sunstein’s policy proposals appears
to be a contradiction in terms, which the authors themselves acknowledge.46 Although the concept of
paternalism and its boundaries are contested issues, Gerald Dworkin provides a helpful analytical
framework by suggesting that paternalism involves at least three elements: first, some kind of limitation
on the freedom or autonomy of some agent; second, that limitation is imposed without the consent of
the agent; and thirdly, that this limitation is imposed for a particular class of reasons, that is, with the
aim of improving the welfare of the agent (where this includes preventing the agent’s welfare from
diminishing) or in some way promotes the interests, values or good of the agent.47 Since paternalism
involves some intervention with the freedom or autonomy of the agent without the latter’s consent, it
appears to be, by definition, anti-libertarian in character (see below).
Thaler and Sunstein side-step this apparent contradiction in three ways, firstly, in the emphasis which
they place on the purpose of nudge policies in seeking to promote outcomes which are in accordance the
45
Nudge 5.
46
Nudge 5.
47
G Dworkin, ‘Paternalism’, in Stanford Encyclopaedia of Philosophy, available at
http://plato.stanford.edu/entries/paternalism/ , accessed on 23 August 2011.
14
agent’s own understanding of his or her self interest, secondly, in the nature of the intervention which
choice architecture entails, and thirdly, in the freedom of the agent to ‘opt out’ of the nudge simply and
cheaply. The first two of these strategies is reflected in their definition of paternalism:
‘In our understanding, a policy is ‘paternalistic’ if it tries to influence choices in a way that will make choosers
better off, as judged by themselves’ (emphasis in original).48
By emphasising policies that seek to influence choices that would accord with agents’ understanding of
their own welfare, Thaler and Sunstein’s conception of paternalism constitutes what is often referred to
in the philosophical literature as ‘soft’ paternalism. Soft paternalism involves intervention to prevent an
agent from doing X, where the paternalist judges that, relative to the agent’s own views of his or her selfinterest, the doing of X is not in the agent’s interests. In contrast, hard paternalism involves
intervention to prevent the agent from doing X where the paternalist judges that, relative to the
paternalist’s view of what is in the agent’s best interest, the doing of X is not in the agent’s interest. Since
soft paternalism can be interpreted as consistent with displaying respect for the agent’s view of her own
welfare, it is less objectionable from a liberal perspective than hard paternalism because it does not entail
overriding the agent’s preferences. Hence, laying out food in a cafeteria to encourage me to choose the
fruit rather than the chocolate cake may be understood as consistent with my own preferences in at least
two ways: either by reflecting my ‘deeper’ preference for a healthy diet rather than my ‘shallow’ desire
to satisfy my immediate craving for chocolate, or, if I am unaware of the adverse health effects of
consuming the chocolate cake relative to the fruit salad, then a cafeteria layout that encourages me to
select the salad can be understood as consistent with my “true” preferences: had I been properly
informed of the consequences of the available options at the time of choosing.
Yet difficulties arise in attempting to classify any intervention (including but not limited to nudges) as a
form of ‘hard’ or ‘soft’ paternalism due to the heterogeneity of preferences. Consider the Save More
Tomorrow scheme, which automatically commits employees to contributing a portion of their salary to
a retirement savings plan unless they return an opt-out form to their employer. For individual X, who
would like to save more for retirement but suffers from weakness of the will, the default scheme can be
understood as helping to secure her ‘true’ or ‘deep’ preferences. But for individual Y, who has a strong
preference for current consumption over future consumption (ie she is what economists refer to as a
‘hyperbolic discounter’), then the scheme will tend to override her preferences.49 And for individual Z,
who holds unstable preferences, sometimes wanting to save more for retirement but at other times
willing to sacrifice future savings in favour of current consumption, the scheme will sometimes be
48
Nudge, 6.
49
D. N. Husak, 'Legal Paternalism' in H. La Follette (ed) The Oxford Handbook of Practical Ethics (Oxford
University Press, Oxford 2005).
15
consistent with her preferences, whilst at other times overriding them, depending upon her preferences
at the relevant point in time.
The pervasive and inescapable heterogeneity of preferences implies that, at least for some individuals, on
at least some occasions, the policy preferences reflected in any given nudge will be contrary to that
individual’s preferences and thus constitute a form of hard rather than soft paternalism. To avoid this
problem, and thereby maintain the ‘libertarian’ label, Thaler and Sunstein insist that to qualify as a
nudge, the formal range of choices available to any individual must not be restricted or otherwise
altered. Rather, each person remains free to opt out of the default arrangement without appreciable
costs in terms of time, trouble, social sanctions or so forth. 50
3.2.1
Liberty, Autonomy and Rational Self-deliberation
By restricting interventions to those involving the deliberate shaping of choice architecture without
altering the formal range of choices available, combined with insistence on preserving the agent’s
freedom to opt out of the choice architect’s preferred outcomes, Thaler and Sunstein’s policy proposals
lay claim to their libertarian label. Hence the diner is free to reach over the fruit to reach the chocolate
cake, the traveller can choose to direct his aim away from the housefly etched onto the urinal, and the
employee can opt out of the default savings scheme established by her employee. If liberty is
understood simply as the preservation of formal ‘freedom of choice’, then these policy measures can
properly be described as libertarian. But this is a very thin understanding of liberty, failing to identify
the value of individual freedom. Liberty can also be understood in a thicker, richer sense, incorporating
an understanding of the value which liberty occupies within our moral and political framework. On
this view, liberty is understood as respect for individual autonomy rather than simply freedom of choice.51
Individual autonomy is an idea that is generally understood to refer to the capacity to be one’s own
person, to live one’s own life according to reasons and motives that one takes to be one’s own and not
the product of manipulative or distorting external forces.52 My actions are autonomous when they are
guided by reasons that I can underwrite. As Isaiah Berlin put it
50
Similarly, Ogus claims that ‘because the cost of opt out is trivial, such forms of intervention can be
regarded as consistent with traditional notions of individual autonomy’: A. Ogus, 'The paradoxes of legal
paternalism and how to resolve them' (2010) 30 Legal Studies 61-73.
51
See generally J. Raz. The Morality of Freedom (Oxford University Press, Oxford 1986).
52
Ibid 407.
16
When I say that I am rational, at least part of what I mean is that it is my reason that distinguishes me as a human
being from the rest of the world. I wish, above all, to be conscious of myself as a thinking, willing, active being,
bearing responsibility for my choices and able to explain them by reference to my own ideas and principles.53
The centrality of reason as a basis for individual action is critical, revealing the fundamental weakness of
nudge’s claimed libertarian character, at least in relation to nudges that can be understood as
‘irrationality-exploiting’. The paradigm autonomous decision is that made by a mentally competent,
fully informed agent, arrived at through a process of rational self-deliberation, so that the agent’s chosen
outcome can be justified and explained by reference to reasons which the agent has identified and
endorsed. Yet for many nudges, the causal mechanism through which choice architecture is intended to
work deliberately seeks to by-pass the individual’s rational decision-making processes in order to
channel behaviour in the direction preferred by the choice architect. The entire basis upon which such
policies are is constructed rests on the premise that, due to various cognitive ‘defects’, individuals
frequently fail to exercise their powers of reasoned self-deliberation, and this failure can be exploited to
‘nudge’ choices in a particular direction. Hence default schemes from which individuals may positively
opt out of easily and cheaply seek to harness status quo bias and the inertia effect in order to nudge
individuals into adopting the default preferences. As Bovens explains:
‘there is something less than fully autonomous about the patterns of decision-making that nudge exploits. Such
actions entail not letting my actions be guided by principles that I can underwrite, so these actions are not
autonomous. They can be said to be irrational in so far as what is driving my action does not constitute a reason for
my action (ie. not a feature of the action that I endorse as a feature that makes the action desirable).54
Seen in this light, nudges of this kind entail a subtle form of manipulation by taking advantage of the
human tendency to act unreflectively and, to that extent, are inconsistent with demonstrating respect
for individual autonomy. Consider, for example, two contrasting ways in which a garment shop owner
might seek to increase sales. He might, through sensitive and appropriate use of flattery and charm,
persuade me that a suit that I have tried on in his shop makes me look tall and slender. I know full well
that he is flattering me, given my short stature and rather dumpy figure. Nonetheless, his technique is
one of rational persuasion, playing on my emotion and ego no doubt, but I am nonetheless fully aware of
his purposes and the means by which he seeks to effect them. So if, partly as a result of the shop
owner’s transparent attempts to influence my purchasing decisions, I decide to purchase the suit, I do so
based on a process of rational self-deliberation. On the other hand, he might fit the shop with false
mirrors, creating the illusion to those gazing into them that their appearance is taller and more elongated
than it is in fact. This latter technique clearly amounts to invidious manipulation, creating false beliefs
that form the foundation of an individual’s decision-making deliberations. While both techniques seek
53
I. Berlin, 'Two Concepts of Liberty' in H. Hardy and R. Hausheer (eds), The Proper Study of Mankind
(Pimlico, London 1998) at 233.
54
L. Bovens, 'The Ethics of Nudge' in T. Grune-Yanooff and S. Hansson (eds), Preference Change: Approaches
from Philosophy, Economics and Psychology (Springer, Berlin and New York 2008).
17
to prey upon my personal vanity with the aim of eliciting the same behavioural response, the former is a
legitimate sales technique, while the latter is not. The use of false mirrors is objectionable because it
involves a form of deception of which I had no knowledge and hence to which I could not consent. The
mere fact that I can choose not to buy the suit does not alter the invidiousness of the manipulation.
Granted, if I would have bought the suit in any event, then the consequences of the shop owner’s
techniques of covert manipulation are less serious for me, but this fact alone does not alter the
invidiousness of the technique itself. If I am deceived into accepting a date with a stranger, the mere fact
that I can get up and leave at any time does not alter the legitimacy of his or her action.
The analogy between irrationality-exploiting nudges and that of false mirrors is, however, an imperfect
one.55 False mirrors involve the active distortion of information that forms the basis for individual
decision-making, but displaying the fruit more prominently than the chocolate cake does not. Hence the
manipulation entailed by the nudge is more subtle and less objectionable in moral terms than that of false
mirrors. On the other hand, some architectural nudges, such as Painted Road Stripes, deliberately seek
to create an optical illusion, creating the impression that the vehicle’s speed is increasing in order to
prompt the driver to slow down. Thus both false mirrors and Painted Road Stripes seek to create false
beliefs in the mind of the decision-maker that are deliberately intended to influence her actions in the
choice architect’s preferred direction. Each strategy can therefore be understood as involving a similar
form of manipulation. While the purpose of these two interventions is of considerable importance in
our assessment of their legitimacy, nonetheless one of the reasons why manipulation of any kind is
objectionable lies in the way in which the person thereby manipulated becomes a willing participant in
action that she may not have genuinely wished to participate in. By utilising a form of manipulation,
irrationality-exploiting nudges express contempt and disrespect for individuals as rational beings capable
of reasoned decision-making concerning their own affairs, an objection which Thaler and Sunstein fail to
acknowledge.56
On the other hand, some nudges seek to correct or eliminate some kind of cognitive defect or bias to
promote more informed individual decision-making, or are intended rationally to influence individual
decisions.57 Unlike ‘defaults and anchors’, which purposefully seek to exploit the tendency of
individuals to choose passively and unreflectively, ‘deliberation tools’ such as mandatory cooling off
55
I am indebted to an anonymous reviewer for helping to clarifying my thoughts on this issue.
56
D. M. Hausman and B. Welch, 'Debate: To Nudge or Not to Nudge' (2010) 18 Journal of Political
Philosophy 123-136, 134; Raz, supra n.51, 418; M. D. White, 'Behavioural Law and Economics: The Assault on the
Consent, Will and Dignity' in G. Gaus, C. Favour and J. Lamont (eds), New Essays on Philosophy, Politics & Economics:
Integration and Common Research Projects (Stanford University Press, 2010).
57
Amir and Lobel, supra n.15.
18
periods or mandatory disclosure laws) seek to encourage individuals to make active, reflective decisions
and can therefore be readily characterised as liberty-preserving. Nudges that take the form of physical
architecture may take either form. When the built environment is designed to provide visible cues
which rationally influence individual choice (such as a speed hump, which invites the driver to consider
whether to reduce her speed to avoid damage to her vehicle and the discomfort of driving swiftly over
the hump, or a turnstile which renders physical entry difficult without a valid ticket), they do not
involve any attempt to harness cognitive irrationalities in aid of desired social policy outcomes. Such
‘non-exploitative’ nudges can properly be regarded as libertarian, understood in its thicker, richer sense
as autonomy-respecting. In contrast, the fly etchings in Schipol Airport urinals and Painted Road Lines
are designed to alter behaviour by taking advantage of cognitive irrationalities, rather than appealing to
individual reason. This distinction between exploitative and non-exploitative nudges is by no means
clear-cut, and there is likely to be a substantial grey area, often reflected in social marketing campaigns,
such as the ‘Don’t Mess with Texas’ anti-littering campaign cited by Thaler and Sunstein.58 Although
some social advertisements are concerned primarily with providing new information which can then be
taken into account in the course of individual decision-making, many are intended to affect our emotions
so that a situation is presented with such force that we change our behaviour. The latter kind of social
advertisement relies upon appeals both to our emotional responses combined with the anchoring
heuristic and our rational faculties of deliberation, and is therefore partly irrationality-exploitative in
character.59
4.
Are Nudges Illegitimate?
Despite the fuzziness of the distinction between irrationality-exploiting nudges and those which do not,
it is of considerable importance in helping to evaluate their legitimacy. Although nudges in the form of
deliberation tools and those which appeal to the individual’s rational decision-making capacities can
properly be characterised as libertarian, those which seek to exploit some pattern of irrationality in
individual decision-making do not60.
58
Nudge 64.
59
Bovens, supra n.54.
60
Ibid; J. Christman, 'Autonomy in Moral and Political Philosophy' in Stanford Encyclopedia of Philosophy
(2009) available at <http://plato.stanford.edu/entries/autonomy-moral/> accessed 9 September 2010. As
Hausman and Welsh explain, ‘although nudges leave freedom of choice, understood in the sense of the
available alternatives, virtually unaffected, the extent to which they have control over their own
evaluations is diminished. The policy-maker is attempting to bring about something against the
individual’s will. To the extent that nudges attempt to undermine the individual’s control over her own
deliberation, as well as her ability to assess for herself the alternatives, they are prima facie threatening
to liberty, broadly understood as overt coercion.’: Hausman and Welch, supra n.56.
19
4.1
The Need for Context and Consequence-Sensitive Evaluation
Does the autonomy-diminishing character of irrationality-exploiting nudges and the contempt and
disrespect which they express for individuals imply that they should be strenuously avoided? Although
these are serious objections, they do not necessarily render all such techniques illegitimate. Rather, each
nudge proposal must be evaluated on a case-by-case basis in light of the broader context in which it is
proposed. The need for context-sensitive evaluation arises because an assessment of the legitimacy of
the means which we employ is frequently contingent upon its relationship to the ends which we seek to
secure. It is more helpful to ask if a three year prison sentence could legitimately be imposed on those
who park their cars on double yellow lines, than to ask whether three year prison sentences are
legitimate. In other words, since the legitimacy of a given means entails consideration of the proper
relationship between ends and means, in many cases such assessments can only meaningfully be made in
light of their intended purpose.61
This is not to suggest, however, that meaningful generalisations about legal and social policy instruments
cannot be made for a variety of analytical purposes.62 So, for example, there may be considerable
analytical value in seeking to understand how a particular class of legal or policy instrument is intended
to work, by identifying their salient features for the purposes of comparative examination, or to assess
the extent to which they promote or detract from a given set of values.63 Thus it can be fairly and
meaningfully claimed that criminal sanctions are coercive and therefore require special justification.
Nor is it to suggest that there some techniques that can properly be regarded as illegitimate regardless of
the particular social purpose for which they are employed. Just as there are some ends which are
illegitimate irrespective of means (such as the intentional killing of an innocent person), so also are there
some means that are illegitimate regardless of ends. So, for example, torture is illegitimate for all
purposes, however serious and urgent the social purpose which is claimed to justify its use. But in my
view, the autonomy-diminishing character of irrationality-exploiting nudges does not, in and of itself,
61
Hood poses the question thus: ‘Are moral judgments to be applied to instruments themselves, or only to
the applications of such instruments?’ in C. Hood and H. Margetts. The Tools of Government in the Digital Age
(Palgrave Macmillan, Basingstoke 2007) 163. Even if there is agreement that the moral dimensions of a particular
instrument can only be evaluated in light of specific social applications, there might nevertheless be disagreement
about how the social purpose of the particular instrument is appropriately understood and its implications for
legitimacy. Such contestation is particularly pronounced in debates concerning the legitimacy of using medical
technologies for the purposes of human ‘enhancement’ rather than the ‘treatment’ of disease or illness. See for
example E. Parens, 'Is better always good? The enhancement project' in E. Parens (ed) Enhancing Human Traits
(University of Georgetown Press, Washington 1998).
62
The literature is considerable. See for example C. Hood and H. Margetts. The Tools of Government in the
Digital Age (Palgrave Macmillan, Basingstoke 2007); L. Salamon. The Tools of Government: A Guide to the New
Governance (Oxford University Press, Oxford 2002); E. Vedung, 'Policy Instrument: Typologies and Theories' in
M.-L. Bemelmans-Videc, R. C. Rist and E. Vedung (eds), Carrots, Sticks and Sermons: Policy Instruments and Their
Evaluation (Transaction Publishers, New Brunswick, New Jersey 1998);
63
Indeed, I have done just that in relation to nudges as a class of policy instrument.
20
warrant rejecting all nudge proposals as illegitimate. Although respect for individual autonomy is of
critical importance in any assessment of a policy instrument’s legitimacy, there are a range of other
values which also require consideration. Hence, some kind of consequentialist analysis is required, in
which the harm arising from the manipulation of individual self-deliberation that the nudge entails, and
its associated expression of contempt for individuals as rational agents, is weighed against the benefits
arising from nudging individuals in the direction deemed desirable by the choice architect.64
In particular, consideration must be given to the nature, character and severity associated with the
problem of mistakes that may arise when defaults run contrary to any particular individual’s preferred
choice, but that individual fails to opt out of the default. Consider the possibility that I choose the salad
rather than the lasagne (which I would have genuinely preferred) as a result of the health-promoting
layout of the cafeteria. Although my choice has been manipulated, and fails to reflect my true
preferences, the harm is relatively minor, and the associated level of insult arising from the manipulation
fairly trivial. On the other hand, if the benefit is considerable, particularly in circumstances where I live
in a community in which population level obesity is a serious and increasingly urgent problem, and there
is clear evidence that improved diets can help alleviate the problem, then the nudge is arguably justified.
Likewise, if I slow down when driving around corners because I fall pray to the optical illusion created
by Painted Road Lines contrary to my preference for driving around dangerous corners at speed, then
the consequences of my mistake are very minor indeed, as is the associated seriousness of the insult
associated with the underlying manipulation, whilst the benefit is considerable: significantly reducing the
risk of very serious harm to myself and to other road users.
Although Thaler and Sunstein acknowledge that evaluating nudges requires consideration of their
effects, their framework for evaluation is based on pragmatic concerns relating to the complexity of the
decision in question:
Our basic conclusion is that the evaluation of nudges depends on their effects – on whether they hurt
people or help them. Sceptics might argue that in some domains, it is best to avoid nudges altogether.
But how can firms do this? It is not possible to avoid choice architecture, and in that sense it is not
possible to avoid influencing people. We agree that in some cases, forced choosing [ie. requiring an active
choice, without any default] is best. But it is often not feasible, and sometimes it is more trouble than it is
worth.
True, some kinds of nudges are not inevitable. Education and advertising campaigns are optional, and
they can be avoided. Should governments educate people about the risks of smoking and drinking,
unprotected sex, trans fats, spike-heeled shoes? Should employers offer educational campaigns about
similar topics? To answer these questions, we need to know something about the Nudgers and the
Nudgees. One question is whether the outside agent (the Nudger) is likely to be able to help an
individual (the Nudgee) make a better choice. Part of this depends on how hard the choices are for the
64
Hausman and Welch, supra n.56, 134; Bovens, supra n.54.
21
Nudgees. As we have seen, people are most likely to need nudges for decisions that are difficult,
complex, and infrequent, and when they have poor feedback and few opportunities for learning.65
Although the difficulty and complexity associated with informed decision making may well be relevant
and important when evaluating the legitimacy of nudges, Thaler and Sunstein nevertheless fail to provide
due respect to the role and significance of rights.
4.2
Rights and Wrongs
In some contexts such as Cafeteria and Painted Road Lines, nudge strategies may offer considerable
benefits whilst the consequences of mistakes that occur as a result of nudge strategies are relatively
minor. But there are many contexts in which the consequences of mistakes may be very serious. For
example, one default rule which lawyers cherish is the presumption that those accused of crimes are
innocent until proven guilty. On Thaler and Sunstein’s logic, for policy-makers who wished to lower
public expenditure on the administration of justice (at least if we exclude the cost of providing legal aid
to criminal defendants) and raise conviction rates, it would be eminently sensible to alter this
presumption in favour of the prosecution so that those charged of crimes are presumed guilty until they
prove their innocence. But for lawyers, such a suggestion is preposterous. This example highlights why
the legitimacy of any particular form of choice architecture depends critically on the nature and
significance of the choice in question. Of particular importance is the need to ensure that the rights of
individuals are properly respected. The presumption of innocence is rooted in the fundamental right of
an accused to have the prosecution prove the case against the accused at trial.66 The wrongful conviction
of an innocent person is a grave injustice, inflicting serious moral harm on the person thereby
convicted.67 Accordingly, the state’s obligation to ensure that such harm is avoided is of paramount
importance. For liberal democratic states, this obligation is reflected in the individual’s presumption of
innocence and the right to a fair trial. At the same time, the accused is free to plead guilty to the charges
in court and thereby waive her right to insist upon putting the prosecution to proof and, in that sense,
the presumption of innocence can be understood as a default rule which applies unless and until the
accused opts out of the default. However, in the absence of an explicit waiver of this right by the
accused person, the burden of proof of guilt lies firmly with the prosecution. While altering this default
by reversing the presumption of innocence would invariably lead to an increase in the number of
criminal convictions, this would clearly constitute an unjustified interference with the accused’s right to
65
Nudge 247.
66
A Ashworth, (1998) The Criminal Process: An Evaluative Study, 2nd edn (Oxford, Oxford University Press)
286-297.
67
R. Dworkin. A Matter of Principle (Harvard University Press, Cambridge, 1985), Chapter 3.
22
a fair trial, and such a violation is unlikely to be adequately remedied by allowing accused persons the
freedom to opt out of the default by filling in and returning a form to a public authority within a
specified time period.
Thaler and Sunstein demonstrate some awareness of the need to respect rights-based limitations to the
use of nudges. For example, they acknowledge that nudges must be avoided in relation to the design of
ballot papers in order to favour a particular candidates running for political office because ‘sometimes
people have a right, even a constitutional right, to government neutrality of a certain kind’.68 But
beyond stating that rights to the free exercise of religion or the right to free speech precludes the
government from encouraging individuals to join a ‘Pray to Jesus More Tomorrow’ plan, or a ‘Dissent
Less Tomorrow’ plan, rights receive no further discussion.69 While Thaler and Sunstein
“enthusiastically agree that required (or strongly encouraged) active choosing is sometimes the right
route”, their enthusiasm is tempered by the claim that “forcing people to choose is not always wise, and
remaining neutral is not always possible”.70 Yet no serious attempt is made to identify the conditions in
which nudges ought to be avoided. Moreover, the importance of rights cannot be equated with the
importance of government neutrality. The right to a fair trial and the presumption of innocence, for
example, requires that the state should not be neutral in setting the ‘default’ baseline from which the
guilt of accused persons should be assessed. Rather, it demands that the state should display partiality in
favour of the accused and in which the onus of proof of guilt lies firmly with the prosecution.71
Rights are not, however, the only constraints on the proper use of nudges.72 Consideration must also be
given to the possibility that a default decision that fails to reflect an individual’s genuine preference may
involve the commission of a wrong, even if it does not entail a violation of rights. One well-known
context in which nudges are often advocated concerns post-humous organ donation. In particular, it is
often suggested that an increase in the number of organ donors could be easily achieved by a system of
presumed consent, which Thaler and Sunstein describe thus:
‘A policy that can pass libertarian muster by our standards is called presumed consent. Presumed consent
preserves freedom of choice, but it is different from explicit consent because it shifts the default rule. Under this
68
Nudge 246.
69
Ibid.
70
Nudge 243.
71
R. Dworkin. A Matter of Principle (Harvard University Press, Cambridge, 1985), Chapter 3.
72
The proper role, status and scope of rights in moral, political and legal discourse are heavily contested.
See R. Dworkin. Taking Rights Seriously (Duckworth, London 1977), Chapter 7.
23
policy, all citizens would be presumed to be consenting donors, but they would have the opportunity to register
their unwillingness to donate, and they could do so easily.73
Such a proposal was recently considered by the Organ Donation Taskforce, established by the British
government, to consider the potential impact of an opt-out system for organ donation in the UK. In
recommending against the introduction of such a system, it drew attention to the possibility that such a
scheme might result in the commission of wrongs against deceased persons, even if they experience no
direct harm as a result, by failing to display due respect for their body (including failure to respect
appropriate cultural or religious practices) or to honour their wishes regarding donation:
We do wrong if the potential donor’s wishes are not fulfilled, and this wrong is compounded if the organs are
removed when they would not have wanted them to be. On the other hand, if the organs are not removed, even
though the person wanted them to be, the potential donor has been wronged and the potential recipients harmed
unnecessarily’74
The Taskforce’s Legal Working Group also concluded that a presumed consent system for organ
donations in which those who had not indicated their contrary wishes via an opt out register would be
presumed to have consented to the removal of their organs was unlikely to comply with the European
Convention on Human Rights because it would result in the state taking organs from the disorganised,
regardless of that person’s views, and would therefore be vulnerable to a successful legal challenge
under the Human Rights Act 1998. While Thaler and Sunstein also shy away from advocating a default
‘opt-out’ system in favour of a regime of ‘mandated choice’ to encourage organ donation, they do so on
the grounds that the former would be a “hard sell politically” owing to the “sensitivity of the subject”.75
The thinness of their reasoning, particularly when contrasted with the analysis provided by the
Taskforce, is indicative of the rather light-handed treatment which the authors devote to potential
objections to their proposals. 76
73
Nudge 187.
74
It reached this conclusion for a variety of reasons, including the real possibility that such a regime might
in fact lead to a reduction in the number of organs donated owing to its potential to undermine the concept of
donation as a gift, to undermine trust in NHS institutions, to distract attention from the need to improve the
essential infrastructure for supporting organ donation and the urgent need to improve public awareness and
understanding of organ donation. Most compelling of all, it concluded that “we found no convincing evidence that
it would deliver significant increases in the number of donated organs”: Organ Donation Taskforce, 'The Potential
Impact of an Opt Out System for Organ Donation in the UK' (London 2008) 7, 341; cf Thaler and Sunstein claim
that, “careful statistical analyses by the economists Abadie and Gay (2004) find that, holding everything else
constant, switching from explicit consent to presumed consent increases the donation rate in a country by roughly
16%. Johnson and Golstein obtain a slightly smaller but similar effect. Whatever the precise figure, it is clear that
the switch would save thousands of lives every year”: Nudge 188-189.
75
Nudge 189.
76
See also M. B. Gill, 'Presumed Consent, Autonomy, and Organ Donation' (2004) 29 Journal of Medicine
and Philosophy 37-59; C. Cohen, 'The Case for Presumed Consent to Transplant Human Organs After Death'
(1992) 24 Transplantation Proceedings 2168-2172; R. M. Veatch and J. B. Pitt, 'The Myth of Presumed Consent:
Ethical Problems in Organ Procurement Strategies' (1995) 27 Transplantation Proceedings 1888-1892.
24
4.3
The Importance of Transparency
In Chapter 15, Thaler and Sunstein outline five possible objections to their proposals. Of these, one is
based on objections to state-based redistribution,77 and three are primarily objections to paternalism
rather than objections to nudge as a means for securing policy goals, ie fears that paternalism in the
modest form advocated will lead to slippage in the direction of more intrusive forms of paternalism,78
concerns that policy-makers will act for self-interested motives, rather than in the public interest,79 and
a belief that individuals are entitled to make wrong (or unwise) decisions.80 Only one of the anticipated
objections is concerned with nudge as a means for implementing public policy goals, and it is here that
Thaler and Sunstein demonstrate their own unease with nudges’ autonomy-diminishing potential. In
particular, they note that certain kinds of nudges might resemble subliminal advertising,81 observing
that:
‘a general objection to libertarian paternalism, and to certain kinds of nudges, might be that they are insidious –
that they empower governments to manoeuvre people in its preferred directions, and at the same time provide
officials with excellent tools by which to accomplish this task.’82
Recognising the potential for abuse, Thaler and Sunstein endorse what they refer to as John Rawls’s
“publicity principle” which, in their view:
‘bans governments from selecting a policy that it would not be able or willing to defend publicly to its own
citizens. We like this principle on two grounds. The first is practical. If a government adopts a policy that it could
not defend publicly, it stands to face considerable embarrassment, and perhaps, much worse, if the policy and its
grounds are disclosed....The second and more important ground involves the idea of respect. The government
should respect the people whom it governs, and if it adopts policies that it could not defend in public, it fails to
manifest that respect. In this sense, the publicity principle is concerned with the prohibition on lying. Someone
who lies treats people as means, not as ends.’83
On this basis, Thaler and Sunstein conclude that subliminal advertising is unacceptable in ways that
cannot be cured by an open announcement by government that it will employ such techniques.84 The
77
Nudge 241-2.
78
Nudge 236.
79
Nudge 238-241.
80
Nudge 240-241.
81
Nudge 244.
82
Nudge 244.
83
Nudge 244-5.
84
Nudge 245-6.
25
analogy with subliminal advertising is revealing. Such practices are objectionable not merely because we
have not been informed of their use, but because they involve a form of deception which we could not
detect (at least, not without the use of special equipment). By highlighting the importance of
monitoring government activity, Thaler and Sunstein effectively concede that the principle of
transparency operates as an important limitation on the use of nudges by the state. If a nudge is not
sufficiently transparent so as to allow meaningful monitoring of its use by government, then such a
measure would be highly vulnerable to abuse and therefore illegitimate.85
Given the diversity and variety of nudges, and their varying levels of transparency, each nudge must be
assessed on a case-by-case basis, to determine whether it can be meaningfully scrutinised. In this
respect, Bovens helpfully contrasts the ways in which subliminal advertising and nudges differ in
transparency by distinguishing between public announcements by government that it will use certain
types of instrument (such as subliminal images, or nudges) to solve social problems (‘type interference
transparency), and each occasion when that particular instrument is employed (‘token interference
transparency’). He argues that if there is no way we could notice (other than the use of special
equipment) each occasion when a particular instrument is employed, then a public announcement by
government that it will use instruments of that type will not provide adequate transparency, for it does
not overcome the barriers to effective monitoring. Thus, if the government announced in advance that
it proposed to undertake a subliminal advertising campaign to encourage individuals to brush their teeth,
this would not be sufficient to overcome the objections from transparency. However, in relation to
some kinds of nudges, it might be possible to recognise that this is occurring on each occasion when it is
employed – eg to recognise that the food is arranged in a particular manner, or that pension savings
forms are sent earlier. Bovens therefore suggests that if a watchful person would 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, then this might constitute adequate transparency.86 In this respect, nudges
differ from subliminal advertising: token interference transparency is not possible for subliminal images
– if allowed, we would be ‘signing a blank cheque and could only hope that the government was not
abusing its power when using subliminal advertising’, but this is not the case for all Thaler and Sunstein
style nudges. 87 However, even if the state confines itself to using nudges which are sufficiently
transparent to enable watchful individuals to identify and monitor their use, this may not provide
85
Hausman and Welch, supra n.56, 132.
86
Bovens, supra n.54.
87
Ibid 14.
26
adequate transparency unless there are institutional mechanisms for ensuring that individuals can raise
objections to such techniques to ensure that the state is accountable for their use.88
4.4
Infantilisation or Policy Pragmatism?
It is not difficult to see why policy makers have been easily seduced by the logic underpinning nudge
techniques, which appear to provide simple and effective ways to shape individual behaviour. But even
if we assume that nudges are effective in altering individuals’ short-term preferences, whether these
behavioural changes prove enduring is far from certain. Several commentators have warned that the
short term success of nudging might be consistent with long term failure.89 One reason why nudging
techniques might ultimately produce counter-productive effects arises from the danger of infantilisation.
As Klick and Mitchell point out, the deliberate shaping of choices in the policy-makers’ preferred
direction might reduce opportunities through which individuals learn and develop competence to
identify which options are most likely to produce desirable outcomes and to compile, rank and select the
most favourable option.90 In other words, exploitation of decision-making foibles by government may
ultimately diminish people’s rational decision-making capacities.91 Whether or not nudges have this
effect is an empirical question, and is likely to vary across policy-domains. So, it may be that persistent
nudging that prompts me to choose the salad rather than the lasagne that I lose my appetite for lasagne
and develop a taste in favour of salad so that over time a genuine shift in my preferences occurs.92 But it
might be that if I choose the salad only on the basis of the nudge, then without the ‘help’ of choice
architecture, I return to my lasagne-eating ways.93 In other words, nudges may shape an individual’s
choices, but they might not lead to a genuine alteration of preferences.94
88
Concerns about the difficulties of contesting subtle spatial power in comparison with more overt
disciplinary techniques are often expressed by Foucauldian scholars. See R. Jones, J. Pykett and M. Whitehead,
'The Geographies of Soft Paternalism in the UK: The Rise of the Avuncular State and Changing Behaviour after
Neoliberalism' (2011) 5 Geography Compass 50-62, 57.
89
Amir and Lobel, supra 15; Bovens, supra n.54; Hausman and Welch, supra n.56.
90
J. Klick and G. Mitchell, 'Government Regulation of Irrationality: Moral and Cognitive Hazards' (20056) 90 Minnesota Law Review 1620-1663.
91
Hausman and Welch, supra n.56.
92
Bovens, supra n.54.
93
Thaler and Sunstein briefly note that one objection to their proposals is based on the ‘the right to be
wrong’ on the basis that it is sometimes helpful for us to make mistakes, since that is how we learn: Nudge 240.
94
Buckley, supra n.18.
27
Concerns about the possible infantilisation effect of nudge techniques have resonance in critique
provided by criminologists concerning the use of so-called ‘situational crime prevention techniques’
which seek to channel behaviour in ways that reduce the occurrence of criminal events through the use
of situational stimuli to guide conduct towards lawful outcomes, preferably in ways that are subtle,
unobtrusive and invisible to those whose conduct is effected.95 In particular, criminologists have drawn
attention to the implications of such approaches for individual responsibility, signifying that people are
incapable of responding to appeals to moral reason or to exercise self-control and restraint.96 Similar
concerns, albeit in a rather different context, have been expressed by those concerned with the use of
human ‘enhancement technologies’.97 One claim often made by opposed to the use of such technologies,
even by mentally competent individuals, is that it entails the cultivation of inauthentic virtues. These
concerns reflect a fear that the use of such technologies will result in a loss of moral responsibility, such
that it is the technology, rather than the moral character of the individual, that is responsible for the
resulting display of virtue, and hence the individual is not an appropriate candidate for moral praise. Yet
what distinguishes these critiques from concerns about the possible infantilisation effects of nudging is
not that they may reduce the effectiveness of such interventions, but what they signify and express about
individuals and their capacity for autonomous and responsible decision-making.
These critiques can be understood as largely ideological objections to policy pragmatism. Yet the
pragmatists cannot be easily dismissed, particularly given that experiments in cognitive psychology
clearly demonstrate that, in reality, many of our decisions are made as a result of processes that bear
little resemblance to the paradigm model of reflective, reasoned decision-making processes of the
autonomous individual upon which liberal thought is premised. Not only does our pervasive use of
decision-making heuristics helps us cope with the enormous volume of decisions that we confront
routinely in our daily life but what has been described as the ‘second wave’ or ‘counter-revolution’ of
behavioural psychology’ demonstrates that many apparently irrational decisions are only irrational at a
superficial level: our decision-making heuristics are much more efficient and sophisticated than the first
wave of behavioural law and economics suggests.98 In other words, even if we accept that our policy
95
D. Garland, 'Ideas, Institutions and Situational Crime Prevention' in D. Garland (ed) Ethical and Social
Perspectives on Situational Crime Prevention (Hart Publishing, Oxford 2000).
96
R. Duff and S. Marshall, 'Benefits, Burdens and Responsibilities: Some Ethical Dimensions of Situational
Crime Prevention' in A. von Hirsch, D. Garland and A. Wakefield (eds), Ethical and Social Perspectives on Situational
Crime Prevention (Hart Publishing, Oxford 2000).
97
The literature is enormous. Some well-known discussions include President's Council on Bioethics,
Beyond Therapy: Biotechnology and the Pursuit of Happiness (Washington DC 2003); F. Fukuyama, Our Posthuman Future
(Profile Books, London 2002); J. Habermas, The Future of Human Nature (Polity Press, Cambridge 2003); M. J.
Sandel, The Case Against Perfection (Harvard University Press, Cambridge, MA 2007); E. Parens (ed) Enhancing
Human Traits (University of Georgetown Press, Washington 1998).
98
Buckley, supra, n.18; See N. Levy, Neuroethics (Cambridge University Press, Cambridge, 2007) 29-32.
28
responses should reflect a proper respect for liberal principles, there might be little point in staunch
adherence to principle if it generates policies that are seriously out of step with the realities of our
decision-making behaviour.
5. Nudge as Fudge
Doubts about the effectiveness of nudge techniques can be linked to deeper objections to employing the
laboratory findings of cognitive psychologists as the basis for broad-based social and legal policy
formation. For example, several scholars point out that it is not yet established that empirical findings
from carefully controlled laboratory experiments concerning human decision making provide an
accurate portrayal of the decision-making behaviours of individuals in real-life situations.99 Cognitive
paternalists can point to experiments where our hunches seem to misfire. But, as Buckley points out,
they lack a grand theory that tells us what hunch will be employed and when.100 We simply do not
know which hunch a person will employ for any particular choice, and often different hunches seem to
tug in opposite directions, so that the paternalist’s policy prescriptions are indeterminate. Acceptance
of these limitations implies that the most valuable contribution of these experimental studies lies in
providing an incremental, highly contextual approach to developing new legal and policy initiatives
intended to alter people’s behaviour, experimenting with small interventions whose outcomes can be
carefully studied in order to identify whether they might be applied to other similar situations.101 It is
here that nudge can make a genuine and helpful contribution, through an analysis of a range of smallscale, experimental interventions that can be carefully studied in order to identify their behavioural
effects and provide a stronger evidence base for the development and implementation of legal and social
policy.
Yet Thaler and Sunstein pursue a much more ambitious agenda, portraying their proposals as a coherent
regulatory philosophy that claims to offer a middle way between advocates of strong and weak
regulation.102 As a theory of the legitimacy of state action, Nudge and the ‘new movement’ which Thaler
and Sunstein describe as ‘libertarian paternalism’, is seriously wanting. What Nudge provides is an
attractively packaged approach to the use of choice architecture as a means for pursuing legal and social
policy goals. But it provides no meaningful guidance on the ends which the state can legitimately pursue,
99
Buckley, supra, n. 18, 37; M. J. Rizzo and D. G. Whitman, 'Little Brother is Watching You: New
Paternalism on the Slippery Slopes' (2009) 51 Arizona Law Review 685.
100
Buckley, supra n. 18, 48-49.
101
Rostain, supra n.31, 989.
102
Nudge 14-15.
29
beyond trite claims that nudges can help people make ‘better’ decisions. To use a rather prosaic
example, it seems eminently sensible that local authorities adopt nudge proposals to improve the
cleanliness and hygiene of public conveniences, by installing fly etching in urinals for example. But this
tells as nothing about whether the state and other governmental authorities should be considered
responsible for providing such facilities in the first place. As Schlag puts it, ‘if you’re going to have a
cafeteria, then you might as well have a well-administered one. Nudge helps. Considerably. But does
politics come down to cafeteria?’103
In making over-inflated claims that Nudge provides a coherent regulatory philosophy, what Thaler and
Sunstein demonstrate is that government policy instruments have inescapably ideological dimensions.
Hence claims that any form of intervention can lay claim to ‘neutrality’ should be approached with
considerable scepticism.104 Given that in many contexts, some form of choice architecture is inevitable
(and therefore prone to influencing the resulting choices which individuals make) to what extent can a
liberal state actively seek to shape individual choices? Some suggest that framing problems might be
cured through ‘de-biasing’ techniques that seek to encourage people to act reflectively and consciously
through a process of active choice, by helping individuals to identify their deepest preferences by
reframing problems in a way that requires people to deliberate over the choice.105 Such strategies have
various benefits, forcing individuals to learn about the options, helping individuals to reflect and more
accurately identify what they really want. Yet forcing people to make active choices is often costly,
particularly when the information needed to choose is not easy to identify and collate. And if most
people would want the same thing in any event, then perhaps a default regime would be preferable to a
de-biasing strategy. In a world of bounded rationality, where information and deliberation is not
costless, even libertarians might prefer default rules that nudge them toward the same choices they
would make if they had invested the time and energy to reflect upon them.
In other words, given the reality of human decision-making, combined with the ubiquity of choice
architecture, then why not a nudge by the state? Since our choices are inevitably and inescapably a
function of the way in those choices are framed, there may be no clear way for legal and policy-makers
to respect what people really want.106 One of Nudges’ great strengths lies in highlighting the ubiquity of
choice, providing vivid illustrations of the ways in which choice architecture influences individual
103
P. Schlag, 'Nudge, Choice Architecture and Libertarian Paternalism' (2010) Michigan Law Review 913
104
B. Morgan and K. Yeung. An Introduction to Law and Regulation (Cambridge University Press, Cambridge
2007) 140-148.
105
C. Sunstein ‘Behavioural Law and Economics: A Progress Report’ (1999) American Law & Economics
Review 115-157, 150-151; Buckley, supra n. 18, 48-49.
106
C. A. Hill, 'The Rationality of Preference Construction (and the Irrationality of Rational Choice)' (2008)
9 Minnesota Journal of Law, Science & Technology 689.
30
decision-making and, in so doing, provokes consideration of a series of more fundamental questions
about the nature and legitimacy of paternalism. I have suggested that the deliberate shaping of choice
architecture interferes with the liberty or autonomy of agents when it explicitly seeks to exploit
cognitive irrationalities, but such an interference might nevertheless be justified in particular policy
contexts, taking into consideration the importance of rights, the potential for wrongs, and paying careful
consideration to the consequences of mistakes. In reaching this conclusion, I have argued that deliberate
attempts to circumvent an individual’s processes of reflective self-deliberation amount to an interference
with individual autonomy. But as Gerald Dworkin points out, identifying whether a particular measure
entails an ‘interference’ with the liberty or autonomy of agents is often tricky. Clear cases include
threatening bodily compulsion, lying, withholding relevant information that the person has a right to
have, or imposing requirements or conditions. But what of the alteration of choice architecture which
does not involve any deliberate attempt to circumvent an individual’s processes of reflective selfdeliberation, particularly when it does not alter the formal range of choices available? And in what
circumstances does any individual have a right to the presentation of choices in a particular form?
Although there are good grounds for criticising Thaler and Sunstein’s rather simplistic portrayal of
‘libertarian paternalism’, they have nonetheless succeeded in opening up, and making accessible, a set of
important and difficult questions about the proper role of the state in shaping the choice environment
that we inhabit and for which further research and reflection of both an empirical and normative kind is
necessary and important. Seen in this light, Nudge is a fudge in both senses of the term. On the one
hand, it fails to deal adequately with the autonomy-diminishing character of nudges which deliberately
seek to exploit the frailties of human decision-making and their potential for abuse. Yet it presents
choice architecture in a highly readable, readily accessible and digestible form offering superficial
satisfaction of policy-makers’ cravings for quick, simple and effective short-term solutions but which,
ultimately, might not provide much in the way of serious and enduring sustenance.
15.9.2011
31