that the answer depends on some kind of cost-beneﬁt analysis. If the cost
of the switch is low enough and the risk of burns is high enough, then the
answer is yes.
Asymmetric paternalists also endorse a class of regulations requiring
“cooling-off periods.” The rationale is that in the heat of the moment,
consumers might make ill-considered or improvident decisions. Self-con-
trol problems are the underlying concern. A mandatory cooling-off period
for door-to-door sales, of the sort imposed by the Federal Trade Commis-
sion in 1972, provides an illustration.
Under the ftc’s rule, any door-to-
door sale must be accompanied by a written statement informing the
buyer of his right to rescind the purchase within three days of the transac-
tion. The law came about because of complaints about high-pressure sales
techniques and contracts with ﬁne print. Again a cost-beneﬁt test, looking
at the beneﬁts for those who are helped and the costs for those who are
not, could be used to decide when such laws would be imposed. Using
such a test, regulators would want to consider how big the imposition is on
eceive the product, and how often
buyers would want to change their minds. When the costs are low (did
anyone ever really need to buy an encyclopedia right away, even before
Wikipedia was online?) and there are frequent changes of heart, such a reg-
ulation makes sense to us.
For certain fundamental decisions, often made on impulse, a similar
y waiting pe-
riod before a couple may get divorced.
Asking people to pause and think
before making a decision of that magnitude seems like a sensible idea, and
we are hard-pressed to think of why anyone would need to divorce imme-
diately. (True, spouses sometimes really don’t like each other, but is it re-
ally terrible to have to wait a short while before the deed is done?) We
could easily imagine similar restrictions on the decision to marry, and some
states have moved in this direction as well.
Aware that people might act in
a way that they will regret, regulators do not block their choices but do
ensure a period for sober reﬂection. Note in this regard that mandatory
cooling-off periods make best sense, and tend to be imposed, when two
conditions are met: (a) people make the relevant decisions infrequently
and therefore lack a great deal of experience, and (b) emotions are likely to
EXTENSIONS AND OBJECTIONS
be running high. These are the circumstances in which people are espe-
cially prone to making choices that they will regret.
Occupational safety and health laws go beyond asymmetric paternalism;
they impose ﬂat bans, and they undoubtedly do hurt some people.
laws do not permit individual workers to trade their right to (what the gov-
ernment considers to be) a safe work environment in return for a higher
salary, even if sophisticated and knowledgeable people might like to do
that. Social Security programs do not merely encourage savings; they re-
quire it. The laws that ban discrimination on the basis of race, sex, and re-
ligion are not waivable. An employee cannot be asked to trade the right to
be free from sexual harassment in return for a higher wage. These various
prohibitions are not in any sense libertarian, but perhaps some of them can
be defended by reference to the kinds of Human errors that we have ex-
plored here. Nonlibertarian paternalists might like to build on such initia-
tives to do a great deal more, perhaps in the domains of health care and
Many of these arguments have substantial appeal, yet we resist going
further down the paternalistic path. What are the grounds for our resis-
tance? After all, we have already granted that the costs imposed by libertar-
ian paternalism may not be zero, so it would be disingenuous for us to say
that we always and strongly object to regulations that raise the costs im-
posed from tiny to small. Nor do we personally oppose all mandates. But
deciding where to stop, and when to call a nudge a shove (much less a
prison), is tricky. Where mandates are involved and opt-outs are unavail-
able, the slippery-slope argument can begin to have some merit, especially
if regulators are heavy-handed. We agree that ﬂat bans are justiﬁed in some
contexts, but they raise distinctive concerns, and, in general, we prefer in-
terventions that are more libertarian and less intrusive. We are much less
cool about cooling-off periods. Even warm. In the right circumstances,
the gains from such rules can be sufﬁcient to make it worthwhile to take a
few cautious steps down that possibly slippery slope.
THE REAL THIRD WAY
In this book we have made two major claims. The ﬁrst is that seem-
ingly small features of social situations can have massive effects on people’s
behavior; nudges are everywhere, even if we do not see them. Choice archi-
tecture, both good and bad, is pervasive and unavoidable, and it greatly af-
fects our decisions. The second claim is that libertarian paternalism is not an
oxymoron. Choice architects can preserve freedom of choice while also
nudging people in directions that will improve their lives.
We have covered a great deal of territory, including savings, Social Secu-
rity, credit markets, environmental policy, health care, marriage, and much
more. But the range of potential applications is much broader than the top-
ics we have managed to include. One of our main hopes is that an under-
standing of choice architecture, and the power of nudges, will lead others to
think of creative ways to improve human lives in other domains. Many of
those domains involve purely private action. Workplaces, corporate boards,
universities, religious organizations, clubs, and even families might be able
to use, and to beneﬁt from, small exercises in libertarian paternalism.
With respect to government, we hope that the general approach might
serve as a viable middle ground in our unnecessarily polarized society. The
twentieth century was pervaded by a great deal of artiﬁcial talk about the
possibility of a “Third Way.” We are hopeful that libertarian paternalism
offers a real Third Way—one that can break through some of the least
tractable debates in contemporary democracies.
Ever since Franklin Delano Roosevelt’s New Deal, the Democratic
Party has shown a great deal of enthusiasm for rigid national requirements
and for command-and-control regulation. Having identiﬁed serious prob-
lems in the private market, Democrats have often insisted on ﬁrm man-
dates, typically eliminating or at least reducing freedom of choice. Repub-
licans have responded that such mandates are often uninformed or
counterproductive—r diversity of Americans,
one size cannot possibly ﬁt all. Much of the time, they have argued on be-
half of laissez-faire and against government intervention. At least with re-
spect to the economy, freedom of choice has been their deﬁning principle.
To countless ordinary people, the resulting debates seem increasingly
tired, abstract, and unhelpful—pointless sloganeering. Many sensible Dem-
ocrats are fully aware that mandates can be ineffective and even counter-
productive, and that one size may not ﬁt all. American society is simply too
diverse, individuals are simply too creative, circumstances change too
rapidly, and government is simply too fallible. Many sensible Republicans
know that even with free markets, government intervention cannot be
avoided. Free markets depend on government, which must protect private
property and ensure that contracts are enforced. In domains ranging from
environmental protection to planning for retirement to assisting the
needy, markets should certainly be enlisted. In fact, some of the best
nudges use markets; good choice architecture includes close attention to
incentives. But there is all the difference in the world between senseless op-
position to all “government intervention” as such and the sensible claim
that when governments intervene, they should usually do so in a way that
promotes freedom of choice.
For all their differences, liberals and conservatives are beginning to rec-
ognize these fundamental points. No less than those in the private sector,
public ofﬁcials can nudge people in directions that will make their lives
go better while also insisting that the ultimate choice is for individuals, not
for the state. The sheer complexity of modern life, and the astounding
pace of technological and global change, undermine arguments for rigid
mandates or for dogmatic laissez-faire. Emerging developments should
strengthen, at once, the principled commitment to freedom of choice and
the case for the gentle nudge.
THE REAL THIRD WAY
This page intentionally left blank
1. See http://www.coathanger.com.au/archive/dibblys/loo.htm. The example
is also discussed by Vicente (2006).
2. Friedman and Friedman (1980).
3. For a similar deﬁnition, see Van De Veer (1986).
1. Biases and Blunders
For anyone interesearr we rec-
ommend two collections: Kahneman and Tversky (2000) and Gilovich, Grifﬁn, and
1. For a good survey of the research on dual process theories in psychology, see
Chaiken and Trope (1999).
2. Lieberman et al. (2002); Ledoux (1998).
3. See Westen (2007).
4. One reason why adjustment is so often insufﬁcient is that the Reﬂective System is
easily waylaid—it requires signiﬁcant cognitive resources, and thus when resources are
scarce (you are distracted or tired, for example) it can’t adjust the anchor. See Gilbert
5. Strack, Martin, and Schwarz (1988).
6. Slovic, Kunreuther, and White (1974).
7. For more than you ever wanted to know about this subject, including instruc-
tions on how to conduct your own test, go to the Hot Hand Web site: http://thehot
8. See http://www.cdc.gov/nceh/clusters/.
9. Paul Price, “Are You as Good a Teacher as You Think?” 2006, available at http:
10. Mahar (2003).
11.Cooper, Woo, and Dunkelberg (1988).
12.For references for the central ﬁndings in this paragraph, see Sunstein (1998).
13.Kahneman, Knetsch, and Thaler (1991).
14.Tversky and Kahneman (1981).
2. Resisting Temptation
The Planner/Doer model is developed in Thaler and Shefrin (1981). For a review of
recent research on self-control and intertemporal choice see Frederick, Loewenstein,
(1997) and O’Donoghue and Rabin (1999).
1.See Camerer (2007), McClure et al. (2004).
2.See Wansink (2006) for a summary.
3.See Gruber (2002).
4.Thaler and Johnson (1990).
3. Following the Herd
We have drawn for part of this chapter on Sunstein (2003). There is a voluminous lit-
erature on social norms and their impact. Two especially good overviews are Ross and
Nisbett (1991) and Cialdini (2000).
1.See Layton (1999) and Stephenson (2005).
2.See Akerlof, Yellen, and Katz (1996) (teenage pregnancy); Christakis and Fowler
(2007) (obesity); Sacerdote (2001) (college roommate assignment); Sunstein et al.
(2006) (judicial voting patterns).
3.See Berns et al. (2005). Conforming answers are associated with changes in the
perceptual features of the brain rather than with changes in the prefrontal cortex,
merely say that they see things as others do. If all others see things a certain way, we may
actually seethem that way.
4.Ross and Nisbett (1991), 29–30.
5.Jacobs and Campbell (1961).
7.See Crutchﬁeld (1955).
8.For a good account, see http://wwwg/history.php.
9.Gilovich, Medvec, and Savitsky (2000).
10.An outline can be found at http://www.historylink.org/essays/output.cfm?ﬁle
13.See, e.g., Cialdini (1993).
14.See Cialdini, Reno, and Kallgren (2006).
15.See generally Perkins (2003), 7–8.
16.Wechsler et al. (2000).
NOTES TO PAGES 32–67
17. See Perkins (2003), 8–9.
18. See Linkenbach (2003).
19. Linkenbach and Perkins (2003).
20. See Schultz et al. (2007).
21. See Sherman (1980).
22. See Greenwald et al. (1987).
23. See Morwitz and Johnson (1993).
24. See Levav and Fitzsimons (2006).
25. See Kay et al. (2004).
26. See Holland, Hendriks, and Aarts (2005).
27. See Bargh (1997).
5. Choice Architecture
1. Letter of July 2, 2003, to State School Ofﬁcers signed by William Hanse, deputy
secretary of education, and David Chu, undersecretary of defense.
2. Byrne and Bovair (1997).
3. Vicente (2006), 152.
4. See Zeliadt et al. (2006), 1869.
5. Sunstein (2007) explores this point in detail.
6. Save More Tomorrow
This chapter relies heavily on joint research with Shlomo Benartzi and especially Be-
nartzi and Thaler (2007).
1. Gross (2007).
2. Beland (2005), 40–41.
3. Investment Company Institute (2006).
4. We thank David Blake and the U.K. Department of Work and Pensions for pro-
viding us with the data.
5. Choi, Laibson, and Madrian (2004). Duﬂo et al. (2005) ﬁnd a similar unex-
ploited arbitrage opportunity in the context of tax ﬁlers eligible for the saver’s tax
6. Madrian and Shea (2001); Choi et al. (2004), (2002).
7. Choi et al. (2006).
8. Carroll et al. (2005).
9. Carroll et al. (2005).
10. Iyengar, Huberman, and Jiang (2004).
11. Benartzi and Thaler (1999).
12. Benartzi and Thaler (2007).
13. Choi et al. (2002).
14. Duﬂo and Saez (2002).
15. The Pension Protection Act says that if employers offer a 401(k) plan in which
employees’ contributions are matched by the employer, the matches are vested within
NOTES TO PAGES 67–116
matic increases are of at least 1 percent per year for three years or more, then the em-
ployer is presumed to satisfy something called the nondiscrimination rule. (More
speciﬁcally, the employer has to match contributions at a 100 percent rate for the ﬁrst 1
percent of pay that the employee saves, and then at least at a 50 percent rate up to a 6
percent savings rate, for a total match of 3.5 percent if the employee saves 6 percent.)
Nondiscrimination rules limit the proportion of beneﬁts that can be paid to the high-
est-paid workers in the ﬁrm. Since 401(k) plans already have dollar caps for each em-
ployee, it was assumed that this combination of plan features would attract enough en-
rollments from the lower-paid employees to pass the test. Although reasonable people
can quibble with the speciﬁc provisions of the bill (which represent the usual sort of po-
litical compromises), we think that it is an excellent example of nudging. Employers are
not required to change their plans, but if they do, they get a free pass on an annoying
form to ﬁll out.
7. Naïve Investing
This chapter draws heavily on joint research with Shlomo Benartzi, especially Be-
nartzi and Thaler (2007).
1.See Benartzi and Thaler (1999).
2.Quoted in Zweig (1998).
3.See Benartzi and Thaler (2001).
4.Read and Loewenstein (1995). See also Simonson (1990), who originated the
5.Benartzi and Thaler (2001).
6.Benartzi and Thaler (2007).
7.Our sources for this story are Ketchum (2002) and Prestwood (2001).
8.Mitchell and Utkus (2004).
9.Benartzi, Thaler, Utkus, and Sunstein (2007).
11.For more details on this topic see Benartzi, Thaler, Utkus, and Sunstein (2007).
12.Mitchell and Utkus (2004) report that the retailer Sears had a proﬁt-sharing plan
invested laressed Congr
tion plans from the 10 percent diversiﬁcation rule being applied to deﬁned-beneﬁt
plans. Sears was once a highly successful ﬁrm, and its employees had historically retired
with large balances from the company stock plan.
8. Credit Markets
1.Thanks to Phil Maymin for suggesting this one.
2.Karlan and Zinman (2007).
3.Simon and Haggerty (2007).
4.Morton, Zettelmeyer, and Silva-Risso (2003).
NOTES TO PAGES 117–38
5. See Cuomo (2007).
6. Bettinger, Long, and Oreopoulos (research in progress [a]).
7. Bettinger, Long, and Oreopoulos (research in progress [a]).
8. Draut and Silva (2003).
9. Privatizing Social Security
This chapter draws heavily on Cronqvist and Thaler (2004). We thank Henrik Cron-
qvist for calculating some updated results for us.
1. For a discussion of required active choosing, see Carroll et al. (2005).
2. French and Poterba (1991).
3. See Samuelson and Zeckhauser (1988); Ameriks and Zeldes (2001).
4. See Kuran and Sunstein (1999).
10. Prescription Drugs
“D is for Daunting” was a headline from the Pittsburgh Post-Gazette’s special section
on choosing plans in 2005. Huge thanks to Katie Merrill and Marion Wrobel for help
guiding us through this morass.
1. White House (2006).
2. Medicare Prescription Drug Plan (n.d.).
3. McFadden (2007).
4. Quoted in Pear (2006).
5. Cubanski and Neuman (2007).
6. Henry J. Kaiser Family Foundation (2007).
7. Winter et al. (2006).
8. Henry J. Kaiser Family Foundation (2006).
9. Nemore (2005).
10. West et al. (2007).
11. Kling et al. (2007) report similarly low ﬁgures from their own survey of dual eli-
gibles. They found that 6–7 percent of the Medicaid population queried actively chose
a different plan in 2007 from the year before.
12. Hoadley et al. (2007).
13. Quoted in Lipman (2005).
14. The Kling et al. (2007) team conducted a small-scale audit of the 1-800-
medicareservice. They found that the lowest-cost plan was identiﬁed in eight of the
twelve calls they made.
15. Vaughan and Gunawardena (2006).
16. Depending on which state you live in, you’ll have to be ready to deal with differ-
ent plan sponsors: national and regional insurance companies and pharmacy bene-
ﬁt managers, with cosponsors that include drugstore chains, retailers, and aarp.
Monthly premiums range from less than $20 to more than $100. Deductibles are set
between zero and $265 (as of 2007). Some basic plans cover 75 percent of drug bills,
while most plans have tier-leveled coverage. If you use any kind of coinsurance, or if
NOTES TO PAGES 140–70
Documents you may be interested
Documents you may be interested