Journal of Behavioral Economics for Policy, 2, No.

1, 61-67, 2018

Misconceptions about nudges

Cass R. Sunstein1 *

Abstract
Some people believe that nudges are an insult to human agency; that nudges are based on excessive trust
in government; that nudges are covert; that nudges are manipulative; that nudges exploit behavioral biases;
that nudges depend on a belief that human beings are irrational; and that nudges work only at the margins
and cannot accomplish much. These are misconceptions. Nudges always respect, and often promote, human
agency; because nudges insist on preserving freedom of choice, they do not put excessive trust in government;
nudges are generally transparent rather than covert or forms of manipulation; many nudges are educative, and
even when they are not, they tend to make life simpler and more navigable; and some nudges have quite large
impacts.

JEL Classification: D01; D9; B41; H11; H2

Keywords
nudges — behavioral economic — default rules — manipulation
1 Robert Walmsley University Professor, Harvard University

*Corresponding author: [email protected]

Nudges are private or public initiatives that steer people in (Conly, 2012); about whether people in diverse nations ap-
particular directions but that also allow them to go their own prove of nudges, or not (Reisch and Sunstein, 2016); and
way (Thaler and Sunstein 2008; Thaler 2015). A reminder is about when and why nudges fail. Much of what has been
a nudge; so is a warning. A GPS device nudges; a default rule, learned is insistently empirical (Halpern, 2015; Benartzi et al.,
automatically enrolling people in some program, is a nudge 2017). With every month, new knowledge becomes available,
(Ebeling and Lotz, 2015). To qualify as a nudge, an initiative and it is by turns chastening, surprising, confirmatory, and
must not impose significant material incentives (including inspiring.
disincentives). My goal here is not to celebrate what has been learned, or
A subsidy is not a nudge; a tax is not a nudge; a fine or to engage the many productive objections, clarifications, and
a jail sentence is not a nudge. To count as such, a nudge refinements (Goldin and Lawson, 2016; Allcott and Kessler,
must preserve freedom of choice. If an intervention imposes 2015; Goldin 2015; Rebonato, 2012), but more modestly
significant material costs on choosers, it might of course be to catalogue some common mistakes and misconceptions.
justified, but it is not a nudge. Some nudges work because Unfortunately, they continue to divert attention both in the
they inform people; other nudges work because they make public domain and in academic circles, and hence to stall
certain choices easier; still other nudges work because of the progress.
power of inertia and procrastination. A pervasive problem, I suggest, is undue abstraction, as
In the ten years since Nudge was published, policymakers people are misled by large-sounding nouns that do not engage
have shown a great deal of creativity and initiative in using with specific practices. Commenting on Sir Joshua Reynolds,
nudges and other behaviorally informed approaches, often William Blake suggested that “to Generalize is to be an Idiot.
saving money and lives in the process. A full catalogue has To Particularize is the Alone Distinction of Merit”. To be sure,
yet to be produced, but partial accounts are available (White- Blake’s suggestion is itself a generalization (and to that extent
head et al., 2017; Halpern, 2015; Sunstein, 2014). There self-contradictory). But let’s not be too fussy. Blake was right,
has also been an extraordinary outpouring of new academic and I shall try to follow his guidance here.
thinking and research on behaviorally informed approaches,
with particular reference to public policy (for a sampling, see Without further ado1 :
Sunstein and Reisch 2016).
As a result, we now know a great deal more about the 1 I am not going to struggle here over definitional questions, though in
consequences of information, reminders, and default rules recent years, a great deal of work has been devoted to those questions. My
(Johnson and Goldstein, 2013); about how to analyze the com- hope is that the opening sentence of this essay is clear enough, at least if it
is informed by the examples that immediately follow it. In the same vein,
plex welfare effects of nudges (Allcott and Kessler, 2015); see Nudge, p. 8: “a nudge, as we shall use the term, is any aspect of the
about the promise and the limits of mandatory choosing choice architecture that alters people behavior in a predictable way without
and prompted choice; about how to think about paternalism forbidding any options or significantly changing their economic incentives”.
 Misconceptions about nudges — 62/67

1. Nudges are an insult to human agency. In free societies, 2. Nudges are based on excessive trust in government. The
people are treated with respect. They are allowed to go their most intuitive objection to nudging is rooted in fear of gov-
own way. Some people object that nudges are troublesome ernment. To put that objection in its sharpest form: Suppose
because they treat people as mere objects for official control that public officials are incompetent or corrupt. Suppose that
(Waldron 2014). your least favorite leaders are or will be in charge. Would you
The objection is off the mark. One of the main points of want them to nudge?
nudging is to preserve freedom of choice –and thus to maintain Or suppose that you are keenly alert to public choice prob-
people’s capacity for agency (Sunstein, 2016). Many nudges lems, emphasized by James Buchanan and his followers, or
are self-consciously educative, and hence they strengthen that “the knowledge problem”, emphasized by Friedrich Hayek
very capacity; consider calorie labels, or warnings about risks and his followers. If interest groups are able to push govern-
associated with certain products. With information, warnings, ment in their preferred directions, and if public officials lack
and reminders, people are in a better position to choose their crucial information, then you might insist: Do not nudge!
own way. Noneducative nudges, such as uses of healthy Reliance on private markets might seem far better (Glaeser,
choice architecture at cafeterias or in grocery stores, allow 2006).
people to choose as they wish. We should pause here to emphasize that a recognition of
Perhaps it could be argued that if the goal is to promote public choice problems, or the absence of relevant informa-
agency, some nudges, such as default rules, are problematic. tion, should lead first and foremost to an attack on mandates
But because such rules are omnipresent in both law and life, and bans, not on nudges. If we distrust government, we should
and often a great blessing. it is not easy to see how that ar- be most concerned about coercion. For that reason, it is odd,
gument might be made convincing. Would it make sense to and a continuing puzzle, that in recent years, those who dis-
excise default rules from the law of contract? To say that em- trust government have often aimed at nudges –and libertarian
ployers, hospitals, and banks are forbidden from using default paternalism– rather than attacking the modern regulatory state,
rules? In practice, what would that even mean? Those who and perhaps adopting a full-scale program in favor of some
are inclined to reject default rules out of respect for individual version of Thatcherism, Hayekianism, or (much more rad-
agency would do well to ponder the countless contexts in ically) uncompromising enthusiasm for free markets. But
which such rules make life simpler and easier to navigate. (On let us put this point to one side. It is fair to worry that if
the immense importance of navagibility, more in a moment). governments err, official nudging may go badly wrong.
A narrower argument would be that in certain settings, Indeed, behavioral science itself might be taken to put this
those who prize agency should insist on active choosing in worry in bold letters. There is no reason to think that public
preference to default rules. In Nudge, Thaler and I make officials are immune to behavioral biases. In a democratic
exactly that argument in the context of organ donation, urging society, the electoral connection might mean that they will
that when people receive drivers’ licenses, they ought to be respond to the same biases that affect ordinary people (Kuran
asked whether they want to be organ donors. In some settings, and Sunstein 1999). To be sure, structural safeguards might
active choosing is indeed better. help, especially if they ensure a large place for technocrats,
Note, however, that sometimes people cannot easily choose insistent on science and on careful attention to costs and ben-
(because they lack bandwidth or expertise) or do not want to efits. But in any real-world polity, behavioral distortions on
choose (Sunstein, 2015); they consider default rules to be a the part of public officials are difficult to avoid. We have had
blessing. One reason is that people have limited time and them in the past and we will have them in the future.
attention (ibid.; Mullainathan and Shafir, 2015), and they ex- These are fair and important points, but if they are taken
ercise their own agency by relying on default rules. If we as an objection to nudging, they run into a logical problem: a
aim to respect individual agency, we will often be inclined to great deal of nudging is inevitable. So long as government
favor those rules for that very reason (Sunstein 2017b). For has offices and websites, it will be nudging. If the law estab-
people who are busy, or who face poverty, default rules can be lishes contract, property, and tort law, it will be nudging, if
particularly valuable, even essential (Mullainathan and Shafir, only because it will set out default rules, which establish what
2015). happens if people do nothing. (Much of the law is a form
To be sure, it is a complex question when active choosing of choice architecture). If the government provides informa-
should be preferred to default rules, or vice-versa. A simple tion, it will be nudging. As Hayek himself wrote, the task of
framework, on which much more would have to be said: In- establishing a competitive system provides “indeed a wide
quire into the costs of decisions and the costs of errors. That and unquestioned field for state activity”, for “in no system
framework helps to explain and organize sensible intuitions that could be rationally defended would the state just do noth-
about when a large menu is an excellent idea (because people ing. An effective competitive system needs an intelligently
like to choose, and because a default option would produce designed and continuously adjusted legal framework as much
errors) and when it is a terrible blunder (because people do as any other” (Hayek, 1943).
not wish to choose, and a default rule will serve them well As Hayek well understood, a state that protects private
enough). property and that enforces contracts has to establish a set
 Misconceptions about nudges — 63/67

of prohibitions and permissions, including a set of default is hidden. (If it is, there is a problem; the right to opt out
entitlements, establishing who has what before bargaining should be clear). In the United States, nudges have generally
begins. For that reason, it is literally pointless to exclaim, “do been adopted after an extended period of public comment.
not nudge!” –at least if one does not embrace anarchy. They have been on the front pages of newspapers. They
The second answer to those who distrust government is have benefited, and been improved, by continuing scrutiny
that because nudges maintain freedom of choice, they insist on (Sunstein 2014). In these circumstances, the objection seems
a safety valve against official error. Those who favor nudges a bit daft.
are keenly alert to the public choice problem and the knowl- Why, then, have intelligent people argued that nudges
edge problem, and to the possibility that public officials will are covert? Is there anything at all to that objection? One
show behavioral biases. Recall that if one distrusts govern- possibility is that when people participate in a randomized
ment, the real focus should be on mandates and bans (as well controlled trial, they may not be informed of that fact. (A
as subsidies and taxes). Nudges ought not to be free from randomized trial might not work if people are told about the
scrutiny, but they should be a relatively low priority, because various conditions). But I suspect that the real answer is that
they are less intrusive, and because they insist that people some nudges work, in practice, even though people do not
must be allowed to go their own way. focus on them, or even think about them (Rebonato, 2012)
It is true, of course, that some nudging is optional. Gov- While they are hardly hidden, the people who are influenced
ernment can warn people about smoking, opioid addiction, by them may be unaware of them, or at least unaware of their
and distracted driving, or not. It can seek to protect con- purposes and effects.
sumers against deception and manipulation, or not. It can For example, a cafeteria might be designed so that the
undertake public education campaigns, or not. If you think healthy foods are most visible and placed first, and people
that government is entirely untrustworthy, you might want it might choose them for that very reason. Such a design is
to avoid nudging whenever it can. In the abstract, that position hardly hidden; it is entirely transparent. People can see where
cannot be ruled out of bounds. On highly pessimistic assump- foods are. At the same time, people may not be aware that
tions about the capacities and incentives of public officials, their cafeteria has been specifically designed so as to promote
and highly optimistic assumptions about the capacities and healthy choices. To be sure, they know that the fruits are more
incentives of those in the private sector, nudging should be visible than the brownies, but they might not know why, and
minimized (Glaeser, 2006). But private actors nudge, and their decision to select a fruit might be quick and automatic
sometimes it is very much in their interest to exploit cognitive rather than reflective. Or people might not think much about
biases (Akerlof and Shiller, 2016). Consider, for example, the default rules that come with (say) an agreement with a
efforts to promote purchases of unhealthy foods; to market rental car company. If people are automatically enrolled into
cigarettes; and to encourage poor financial choices with re- some kind of insurance plan and allowed to opt out, they
spect to mortgages and credit cards (ibid.). Would it be a might say, “yeah, whatever”, and simply go along with the
good idea to forbid public officials from taking steps to reduce default. (By the way, Nudge identified only one new heuristic,
smoking and distracted driving? and it’s that: the “yeah, whatever” heuristic).
To be sure, nudges, like other interventions from such In that sense, it is correct to say that some nudges can
officials, should be constrained by democratic requirements, work even if or perhaps because people are unaware that they
including transparency, public debate, and independent moni- are being nudged. Note, however, that emerging evidence
toring (including, critically, ongoing evaluation of how they finds that the effects of such nudges are not diminished even
work in practice). Public choice problems, and the knowledge if people are told that nudging is at work. Transparency about
problem, are real and important. Constraints of this kind can the existence and justification of default rules appears not to
reduce the risks (without eliminating them). The fundamental reduce their impact in general (Bruns et al. 2016; Loewen-
point is that those risks are far larger with other tools, above stein et al. 2015). For some people, such clarity may even
all mandates and bans, which remove freedom of choice. increase that impact, by amplifying the informational signal
that some default rules offer (Mackenzie et al 2006). On plau-
3. Nudges are covert. Some people have argued that man- sible assumptions, drawing attention to the healthy design
dates, bans, and taxes have one advantage: They are trans- of a cafeteria will actually increase the effect of that design,
parent. People know what they are. No one is fooled. By because it will convey valuable information. (To be sure, it
contrast, nudges are covert and in that sense sneaky, a form may produce “reactance” in some consumers).
of trickery (Glaeser, 2006). They affect people without their
knowledge. 4. Nudges are manipulative. In a variation on the claim
For countless nudges, this objection is hard to understand. that nudges are covert, some people have objected that nudges
A GPS device nudges, and it is entirely transparent. Labels, are a form of manipulation (Conly 2010). But return to the
warnings, and reminders are not exactly hidden; if they are, hazards of abstraction and in particular (!) to the points I
they will not work. When an employer automatically enrolls have just explored: If people are reminded that they have a
employees into a savings plan, subject to opt out, nothing doctor’s appointment next Thursday, no one is manipulating
 Misconceptions about nudges — 64/67

them. The same is true if people are given information about of retirement planning.
the caloric content of food or if they are warned that certain Similarly, default rules work in part because of inertia,
foods contain shellfish or nuts, or that if they take more than which undoubtedly counts as a behavioral bias. Relevant
the recommended dosage of Benadryl, something bad might words from Samuel Beckett’s Waiting For Godot:
happen.
Vladimir: Well? Shall we go?
To be sure, we could imagine a graphic warning about
Estragon: Yes, let’s go.
opioid addiction, or about the use of cell phones while driving,
(They do not move).
that would create immediate fear or revulsion, or intensely
engage people’s emotions; it might be objected that nudges of Many human beings are like Vladimir and Estragon (who
this kind count as a form of manipulation. To know whether were, by the way, unrealistically optimistic. Nudging can be
they do, we need a definition of manipulation. To make a effective because they do not move. But it is misleading –a
(very) long and complex story short, philosophers and others form of rhetoric, in the not-good sense– to suggest that nudges
have generally converged on the view that an action can be “exploit” behavioral biases.
counted as manipulative if it subverts people’s capacity for
rational deliberation (Barnhill 2014; Sunstein 2016). On any 6. Nudges wrongly assume that people are irrational. Some
view, most nudges do not qualify. True, some imaginable critics object that nudges are based on a belief that human
nudges might cross the line, but that is very different from beings are “irrational”, which is both insulting and false2 . (It
saying that nudges are manipulative as such. is also unclear what, exactly, is meant by the term “irrational”,
which behavioral economists usually abhor, and almost never
5. Nudges exploit behavioral biases. Some people object use). This objection takes different forms.
that nudges “exploit” or “take advantage of” behavioral biases. In one form, the objection is that while people rely on sim-
Indeed, some people define nudges as exploitation of behav- ple heuristics and rules of thumb, nothing is wrong with that;
ioral biases (Rebonato 2010). That does sound nefarious. But those heuristics and those rules, work well, and so nudging
the objection is mostly wrong, and while people can define is not needed, and can only make things worse. In another
terms however they wish, this particular definition is a recipe form, the objection urges that the whole idea of nudging is
for confusion. (One of its least attractive qualities is its use based on weak psychological research and on an assortment
of framing; the word “exploit” suggests something untoward, of supposed laboratory findings that do not hold in the real
whereas the term “take account of” does not –and is more world. In yet another form, the objection is that people can
accurate). and should be educated rather than nudged. In what seems
Many nudges make sense, and help people, whether or not to me its best form, this objection urges that people’s utility
behavioral biases are at work. A GPS is useful for people who functions are complex and that outsiders may not understand
do not suffer from any such bias. Disclosure of information them; what seems to be “irrationality” may be the effort to
is helpful even in the absence of any bias. A default rule trade off an assortment of goals (Rebonato 2010). A mundane
simplifies life and can therefore be a blessing whether or not example: People might eat fattening foods not because they
a behavioral bias is involved. As the GPS example suggests, suffer from present bias, but because they greatly enjoy those
many nudges have the goal of increasing navigability –of foods. A less mundane example: People might fail to save for
making it easier for people to get to their preferred destination. retirement not because they suffer from optimistic bias, but
Such nudges stem from an understanding that life can be either because they need the money now.
simple or hard to navigate, and a goal of helpful nudging is to These objections raise many questions, and I shall be brisk
promote simpler navigation. with them here. No one should doubt that heuristics generally
I wish that Nudge had made this point clearer, and had work well (that is why they exist), but they can also misfire.
connected nudging to the central idea of navigability. (I devote When they do, a nudge can exceedingly helpful. For example,
a separate paragraph to the point, not because that is a good people might use the availability heuristic, or something like
use of the paragraph break, but because it is worth singling it, in deciding whether a risk is serious. For those who lack
out). statistical knowledge, it might be reasonable to do exactly that.
At the same time, it is true that some nudges counteract be- But even if a risk has come to fruition in the recent past, it
havioral biases, and that some nudges work because of behav- may nonetheless be low as a statistical matter. The availability
ioral biases. For example, many human beings tend to suffer heuristic can lead to severe and systematic errors, and a nudge
from present bias, which means that they give relatively little might serve as a corrective to either hysteria or complacency.
weight to the long term; many of us suffer from unrealistic 2 The least lovely, and the most peculiar, version of this claim comes from
optimism, which means that we tend to think that things will a German psychologist: “The interest in nudging as opposed to education
turn out better for us than statistical reality suggests. Some should be understood against the specific political background in which it
emerged. In the US, the public education system is largely considered a
nudges try to counteract present bias and optimistic bias –as, failure, and the government tries hard to find ways to steer large sections of
for example, by emphasizing the long-term risks associated the public who can barely read and write. Yet this situation does not apply
with smoking and drinking, or by suggesting the importance everywhere” (Gigerenzer 2015).
 Misconceptions about nudges — 65/67

It is true that many nudges build on well-established behav- traction from what might actually help, or help most. Some
ioral findings, demonstrating that people depart from perfect skeptics think that with an understanding of nudging, we
rationality. For example, default rules work in part because of might have some fresh ideas about how to tweak letters from
the power of inertia (Johnson and Goldstein 2013); reminders government to citizens, producing statistically significant in-
are necessary and effective in part because people have lim- creases in desirable behavior. But that is pretty small stuff. If
ited attention; information will be more likely to influence behavioral economists want to make a contribution, shouldn’t
behavior if it is presented in a way that is attentive to people’s they focus on much more important matters?
imperfect information-processing capacities. These and other It is true and important that behaviorally informed ap-
claims are based on evidence, both in the laboratory and the proaches are hardly limited to nudges; mandates, bans, and in-
real-world. Let us underline the latter point; much of the evi- centives may well have behavioral justifications (Thaler, 2017;
dence comes from the real world and not from the laboratory Loewenstein and Chater, 2017; Conly 2010). The policy pro-
at all (Benartzi et al. 2017; Sunstein and Reisch, 2016). To gram of behavioral science is hardly exhausted by nudges
be sure, it is important to test whether a finding in one setting (Thaler 2017). Mandates, bans, and incentives might turn
will generalize to others. out to be the best and most effective tool. It is also true that
Those who embrace nudges do not use the term “irrational- some nudges produce only modest changes. But sometimes
ity”. On the contrary, they abhor it; “bounded rationality” is they do not, and in multiple domains, nudges have proven far
much better. Nor does anyone doubt that education can work. more cost-effective than other kinds of interventions, which
As I have emphasized, many nudges are educative and target means that per dollar spent, they have had a significantly
deliberative capacities (System 2, in contrast to the more au- larger impact (Benartzi et al., 2017).
tomatic System 1, as elaborated in Kahneman 2014) More By any measure, the consequences of some nudges are not
ambitious educative efforts, such as efforts to teach statisti- properly described as modest. One reason is that a seemingly
cal literacy, are usually complements to nudges, and rarely small initiative, directed at a very large population, may have
substitutes or alternatives. Whether they work is an empirical massive effects. Another reason is that some initiatives have
question, and no one doubts that in some settings, such efforts such effects all by themselves. Consider three examples:
may have real promise.
1. As a result of automatic enrollment in free school meals
It is also true (and important) that third parties might not
programs, more than 11 million poor American children
be able to understand people’s utility functions, which may
are now receiving free breakfast and lunch during the
reflect (for example) delight in high-calorie foods, alcohol,
school year.
and very late nights out; that is one reason that nudgers insist
on preserving freedom of choice. To the extent that nudging 2. Credit card legislation, enacted in 2010, is saving Amer-
is inevitable, it is pointless to contend that because of the ican consumers more than $10 billion annually; signifi-
complexity of people’s utility functions, nudging should be cant portions of those savings come from nudges and
avoided. To the extent that nudging is optional, it should nudge-like interventions (Agarwal et al. 2013).
be undertaken with an appreciation of the risk of error and
with careful efforts to ensure that it promotes, and does not 3. With respect to savings, automatic enrollment in pen-
undermine, people’s welfare. A GPS device does not decrease sion programs has produced massive increases in par-
welfare. In general, information about health risks and po- ticipation rates (Chetty et al., 2012; Thaler, 2016).
tential financial burdens should increase welfare (Agarwal
et al., 2013). To be sure, the concept of welfare requires New nudges, now in early stages or under discussion,
specification (Adler, 2011), but most nudges can be tested could also have a major impact. If, for example, the goal is
without getting into murky philosophical waters; we can have to reduce greenhouse gas emissions, automatic enrollment in
an “incompletely theorized agreement” that some approaches green energy can have large effects (Ebeling and Lotz, 2015;
promote welfare, and others do not, without resolving disputes Pichert and Katsikopoulos, 2008). The Earned Income Tax
among Jeremy Bentham, John Stuart Mill, and Amartya Sen. Credit is probably the most effective anti-poverty program
Of course nudges must be tested to ensure that they are in the United States, but many eligible people do not take
doing what they are supposed to do (Halpern 2015; Thaler advantage of it. Automatic enrollment would have large con-
2015). Some nudges fail. When they do, the right conclusion sequences for the lives of millions of people.
may be that freedom worked –or that we should nudge better It is true, of course, that for countless problems, nudges
(Sunstein 2017a). are hardly enough. They cannot eliminate poverty, unemploy-
ment, and corruption. But by itself, any individual initiative
7. Nudges work only at the margins; they cannot achieve –whether it is a tax, a subsidy, a mandate, or a ban– is un-
a whole lot. If experts are asked to catalogue the world’s ma- likely to solve large problems. Denting them counts as an
jor problems, many of them would single out poverty, hunger, achievement.
unemployment, corruption, premature deaths, terrorism, and For nudging, the last decade has been one of extraordi-
climate change. On one view, nudges are an unfortunate dis- nary achievements –not only because of the creation of many
 Misconceptions about nudges — 66/67

“nudge units”–, but more fundamentally because of the incor- Ebeling, F., and Lotz, S. (2015). “Domestic uptake of
poration of nudges in policymaking at the very highest levels green energy promoted by opt-out tariffs”. Nature Cli-
and the acceleration of new findings, in academic circles and mate Change 5 (9), 868-871. Retrieved from: Doi:
in government, that have produced conceptual and empirical 10.1038/nclimate2681.
breakthroughs (Mullainathan and Shafir 2015; Kling et al.
2012). Money has been saved; so have lives. Fryer Jr, R. G., Levitt, S. D., List, J., and Sadoff, S. (2012).
There is much more to do, and much more to learn. Mis- “Enhancing the efficacy of teacher incentives through
conceptions are a terrible distraction. Unburdened by them, loss aversion: A field experiment” (No. w18237). Na-
let’s get to work. tional Bureau of Economic Research. Retrieved from:
Doi: 10.3386/W18237

Acknowledgments Gigerenzer, Gerd (2015). “On the supposed evidence for

libertarian paternalism”. Review of Philosophy and
I am grateful to Richard Thaler for friendship, collaboration, Psychology 3: 361.
and countless nudges, some of which have made this essay
much better. Thanks also to Michelle Baddeley and Lucia Glaeser, Edward (2006). “Paternalism and policy”. Univer-
Reisch for superb suggestions. sity of Chicago Law Review 73: 133-56.

Goldin, Jacob and Nicholas Lawson (2016). “Defaults, man-

References dates, and taxes: Policy design with active and passive
decision-makers”. American Law & Economics Review
Adler, Matthew (2011). Welfare and fair distribution. Cam- 18: 438.
bridge: Cambridge University Press.
Goldin, Jacob (2015). “Which way to nudge?”. Yale Law
Agarwal S., Chomsisengphet S., Mahoney N. and Stroebel J. Journal 125: 227.
(2013). “Regulating consumer financial products: Evi-
dence from credit card”. National Bureau of Economic Halpern, David (2015). Inside the nudge unit. London:
Research, Working Paper. Ebury Press,

Hayek, Friedrich (1943). The road to serfdom. Chicago:

Akerlof, George and Robert Shiller (2016). Phishing for
University of Chicago Press.
phools. Princeton: Princeton University Press.
Homonoff, T. A. (2013). “Can small incentives have large
Allcott, Hunt and Judd B. Kessler (2015). “The welfare effects? The impact of taxes versus bonuses on dis-
effects of nudges: A case study of energy use social posable bag use”. Industrial Relations Section, Prince-
comparisons”. National Bureau of Economic Research, ton University Working Paper, (575). Retrieved from:
Working Paper 21671. http://arks.princeton.edu/ark:/88435/dsp014q77fr47j.

Barnhill, Anne (2014). “What is manipulation?” in Manip- Kling, Jeffrey and Sendhil Mullainathan, Eldar Shafir, Lee
ulation: Theory and practice 50, 72 (Christian Coons Vermeulen, and Marian V. Wrobel (2012). “Compari-
and Michael Weber eds.). son friction: Experimental evidence from prescription
drug plans”. Quarterly Journal of Economics 127: 199-
Benartzi, Shlomo et al. (2017). “Should Governments invest 235.
more in nudging?”. Psychological Science, available at
http://journals.sagepub.com/doi/abs/10.1177/09567976 Kuran, Timur and Cass R. Sunstein (1999). “Availability
17702501. cascades and risk regulation”. Stanford Law Review 51:
683-762.
Bruns, Hendrik et al. (2016). “Can nudges be transparent and
Johnson, Eric and Daniel Goldstein (2013). “Decisions by
yet effective?”. Available at https://papers.ssrn.com/sol
Default”. In: Eldar Shafir (Ed.). Behavioral founda-
3/papers.cfm?abstracti d = 2816227.
tions of policy, Chapter 24, pp. 417–427. Princeton:
Princeton University Press.
Chetty, Raj, John N. Friedman, Soren Leth-Petersen, Torben
Nielsen and Tore Olsen (2012). “Active vs. passive deci- Loewenstein, George and Chater, Nick (2017). “Putting
sions and crowdout in retirement savings accounts: Evi- nudges in perspective”. Behavioral Public Policy 1:
dence from Denmark”. Available at http://www.nber.org 26-53.
/papers/w18565.
Loewenstein, George, Cindy Bryce, David Hagmann and
Conly, Sarah (2013). Against autonomy. Oxford: Oxford Sachin Rajpal (2015). “Warning: You are about to be
University Press. nudged”. Behavioral Science & Policy 1: 35-42.
 Misconceptions about nudges — 67/67

McKenzie, C. R., Liersch, M. J., and Finkelstein, S. R.

(2006). “Recommendations implicit in policy defaults”.
Psychological Science 17(5), 414-420. doi: 10.1111/j.1467-
9280.2006.01721.x
Mullainathan, Sendhil and Eldar Shafir (2014). Scarcity.
New York: Picador.
Pichert, Daniel and Konstantinos V. Katsikopoulos (2008).
“Green defaults: Information presentation and pro-envi-
ronmental behaviour”. Journal of Environmental Psy-
chology 28: 63–73.
Rebonato, Riccardo (2012). Taking liberties. London: Pal-
grave.

Reisch, Lucia and Cass R. Sunstein (2016). “Do Europeans

like nudges?”. Judgment and Decision Making 11: 310-
25.
Sunstein, Cass R. (2017a). “Nudges that fail”. Behavioral
Public Policy 1: 4-25.

Sunstein, Cass R. (2017b). Human agency and behavioral

economics. London: Palgrave
Sunstein, Cass R. (2015). Choosing not to choose. Oxford:
Oxford University Press.

Sunstein, Cass R. (2014). Simpler: The future of Govern-

ment. New York: Simon and Schuster.
Sunstein, Cass R. and Lucia Reisch (2016). The economics
of Nudge (four volumes). Routledge.
Thaler, Richard (2017). “Much ado about nudging”. Behav-
ioral Public Policy Blog, available at https://bppblog.
com/2017/06/02/much-ado-about-nudging/.
Thaler, Richard (2015). Misbehaving. New York: Norton.
Thaler, Richard H. and Cass R. Sunstein (2008). Nudge: Im-
proving decisions about health, wealth, and happiness.
New Haven: Yale University Press.
Waldron, Jeremy (2014). “It’s All For Your Own Good”.
New York Review of Books, available at http://www.ny
books.com/articles/archives/2014/oct/09/cass-sunstein-
its-all-your-own-good/.
Whitehead, Mark, Rhys Jones, Rachel Lilley, Jessica Pykett
and Rachel Howell (2017). Neuroliberalism: Behavioural
Government in the Twenty-First Century. New York:
Routledge.