Nudge: Improving Decisions About Health, Wealth, and Happiness
Richard H. Thaler, Cass R. Sunstein
Penguin Books, Revised & Expanded edition (2009) (public library)
Summarized by Erick Rabin

Nudge-cover.jpg

Introduction

Choice Architecture involves organizing any context in which people make decisions. Urinals, stairwells, alarm clocks, school cafeterias, retirement accounts—all of these can be made more efficient and conducive to health and wellbeing. Though no design can be perfect, choice architects are always faced with tradeoffs and must make a decision of some sort. Thus the question is not whether to go about engaging in choice architecture, but how best to do so.

Libertarian Paternalism is the answer to this normative question. Rather than choose randomly between different possible frameworks, or endeavor to track individuals’ pre-existing preferences as closely as possible, the authors advocate for a form of paternalism whereby choice architecture ought to strive to make people best off all things considered (“to make their lives longer, healthier, and better”). However, they also recognize free choice to be of fundamental importance, and so they insist that choice architecture must be as liberty-preserving as possible. The net result is a “relatively weak, soft, and nonintrusive type of paternalism” that refuses to prohibit or significantly burden people from exercising their free will.

Nudges provide the key to predictably altering peoples’ behavior without forbidding any course of action outright or significantly changing their incentive structures. Nudges are low-cost tools that alert, remind, or mildly warn people by exploiting human psychology, thereby molding behavior in powerful ways. Whereas classical economics conceptualizes man as perfectly rational (“homo economicus”), nudges are crafted in light of evidence from behavioral psychology suggesting that, in the real world, human beings are flawed in myriad ways. Examples abound, but among the most important is the status quo bias, which leads people to follow default options.

Part I of this book lays out in greater detail the behavioral psychology upon which the concept of choice architecture rests. Parts II-IV explore the application of this method in the spheres of money, health, and freedom, respectively. Part V rebuts potential objections and provides several of the authors’ own recommended nudges.

Application to Ethical Systems

This book applies cutting-edge insights from behavioral economics to several of our most important political and economic institutions. It makes a vital contribution to the ongoing debate about whether, when, and how we ought to use our growing knowledge of human psychology to maximize human well-being without undermining respect for human freedom. In just a few short years, it has had a vast impact on the way politicians and regulators in the United States and abroad think about their work, and it ought to be required reading for anyone interested in the intersection of psychology, economics, and the law.

Part I: Humans & Econs

Chapter 1: Biases & Blunders
The human mind is incredibly powerful, but it is also susceptible to a variety of systematic, predictable mistakes. We process information via two cognitive systems: System 1 is automatic, uncontrolled, effortless, associative, fast, and unconscious; System 2 is reflective, controlling, effortful, deductive, slow, self-aware, and rule-following. System 1 evolved earlier and is generally more powerful than System 2.

The interplay between these systems helps explain the great extent to which our thinking and behavior is influenced by experience-based, time-saving rules of thumb for problem solving, learning, and discovery. Classic examples of heuristics include:

  • Anchoring: In making estimations, we start with some known figure (the “anchor”) and then adjust in the appropriate direction. This results in bias when our adjustments are insufficient or when obviously irrelevant anchors creep into the decision-making process.
  • Availability: We assess the likelihood of risks by how readily examples come to mind: the more available the examples, the more salient the risk. This can lead us to overestimate the probability of recent or dramatic harms and to underestimate the probability of subtle or unfamiliar harms.
  • Representativeness: We judge how likely it is that A belongs to category B by thinking about how similar A is to our stereotypes of B. Biases creep in when things seem representative but are in fact infrequent, e.g. perceiving a basketball player to be “heating up.”

We also have a tendency toward optimism and overconfidence. Regardless of whether the stakes are high or low, people tend to have self-serving biases through which they focus on their strengths while overlooking or rejecting their faults. This protects the ego from threat and injury but also perpetuates illusions and error that can lead to unwise risk-taking, especially in the domain of risks to life and health.

When it comes to losses and gains, people are much more sensitive to the former than the latter—losing something makes people roughly twice as unhappy as gaining something of equal value. Loss-aversion makes people inclined to avoid making changes, even when it would be in our self-interest to do so.

The inertia we experience to preserve what we have is also explained in part by the status quo bias, which leads to the “yeah, whatever” heuristic, whereby we stick with our current situations rather than make changes. This tendency is due largely to lack of attention, and when combined with mindless choosing it can produce radically inefficient results. This helps to explain the great power of defaults as nudges.

Framing also makes a big difference to how we process information. People tend to be somewhat mindless, passive decision makers, so we react very differently to the information that “ninety of one hundred are alive” than to the information that “ten of one hundred are dead”—even though the content of the statements is the same.

Chapter 2: Resisting Temptation
Our preferences evolve over time and can be “dynamically inconsistent,” such that we prefer A>B at T1, B>C at T2, and C>A at T3. This is a useful way to understand temptation: we may be tempted by X in the moment, but also regard X to be not choice worthy in general. Self-control is difficult because of the hot-cold empathy gap: when we are thinking in a cold, calm, rational way, we fail to appreciate the extent to which our desires and behaviors will be altered by arousal.

This can lead to seriously bad outcomes, especially when combined with “mindless choosing” (e.g., people decide what to eat based on what is easily available, and how much to eat based on the size of the containers in which the food is presented).

While self-control strategies often fail, carefully designed precommitment plans can enable us to reign in our automatic System 1 responses. For example,

  • Clocky is an alarm clock that runs away and hides if you don’t get out of bed, making it impossible to groggily push the sleep button ad infinitum.
  • Casino Bans into which gambling addicts can self-enroll.
  • Non-Interest Bearing Savings Vehicles which force you to save, at the price of transaction costs, illiquidity, and near-zero returns.
  • Mental Accounting whereby we treat money as non-fungible and reserve some funds for rent, some for education, some for food, etc. – reduces temptation by providing bright-line rules for divvying up income and savings.

Chapter 3: Following the Herd
Social influences exert a powerful effect on human learning and behavior. Social forces can convey information as to what might be best for you to do or think or exert peer pressure to conform in order to curry favor or avoid conflict.

Because of the spotlight effect, we think that others are paying close attention to what we do and spend extraordinary effort conforming to social norms and fashions. Experiments reveal that people pay far less attention to one another than we think, but our misperception nonetheless leads to additional conformity.

Many experiments demonstrate how focusing on others’ thoughts and behavior can lead to conformity effects. For example, people are 20-40% more likely to err in a ridiculously easy visual perception tasks after watching a room full of their cohorts make blatant mistakes. These effects are mitigated when experiments call for anonymous answers, suggesting that people are especially likely to go along with group blunders if their reputations are on the line. Similar findings have been observed in tasks that involved more subtle errors as well.

Moreover these effects endure over time, helping to explain collective conservatism: the tendency of groups to stick to established patterns even as new needs arise. Knowing this, however, makes it possible to nudge people in a positive direction by initiating a bandwagon effect that changes information and behavior for the better.

Conformism also affects cultural and political behavior. Experiments reveal that people are more likely to, e.g., download a song if many other people have done so beforehand, suggesting that past popularity has powerful momentum—thus small interventions and coincidences early on can produce large variations in outcomes. Of course, this often occurs unintentionally (e.g., those who eat with one other person eat about 35% more than they do when they are alone; members of a group of four eat about 75% more; those in groups of seven or more eat 96% more). But this can be exploited, as when advertisements emphasize the fact that “most people prefer” brand X, and thus social nudges are a powerful resource for choice architecture.

Just like these social influences, priming also provides subtle influences that can increase the ease with which certain information comes to mind. For example, peoples’ conduct can be affected by the “mere-measurement effect,” whereby people will behave differently just in response to being asked, e.g., what they intend to do and how they intend to do it. Likewise, people can be primed by exposure to simple and apparently irrelevant cues: people become more competitive and less generous when they see objects characteristic of business environments, and more cleanly when they eat if there is a scent of all-purpose cleaner in the room.

Chapter 4: When do we Need a Nudge?
There are a variety of fraught choices that are prime candidates for nudges. Nudges can combat the tendency to avoid investment goods (with immediate costs and delayed benefits) and over-indulge in sinful goods (with immediate benefits and deferred costs). They can help us solve especially rare or difficult problems that we are likely to ignore or mishandle through confusion and lack of experience. They can increase feedback so as to improve our ability to learn from past experiences, and give us a better sense of how our preferences are likely to turn out in unfamiliar contexts where we have trouble translating the situation into easily comprehensible terms (e.g., choosing among mutual funds or health plans).

Unfortunately, markets are no panacea. Though competition generally helps to ensure that price reflects quality, sometimes there is insufficient competition, and companies are often incentivized to profit from human frailties rather than to try to minimize their effects. Thus there is a legitimate role for government the play in these matters, and nudges are powerful tools for doing so without imposing mandates or crowding out free choice.

Chapter 5: Choice Architecture
Designers strive to incorporate human factors into their work. Everything from doors to stop signs to stovetops can be designed in light of “stimulus response compatibility,” a simple psychological principle which holds that the signal you receive ought to be consistent with the desired action, as inconsistencies lead people to blunder.

Analogously, choice architects can improve our health, wealth, and happiness through thoughtfully designed NUDGES:

  iNcentives

  Understand mappings defaults

  Give feedback

  Expect error

  Structure complex choices

Incentives should be analyzed in terms of four crucial questions: Who uses? Who chooses? Who pays? Who profits? Free markets often address all of these problems, but incentive conflicts may sometimes arise, especially when choosers do not notice the incentives they face. Thus it is important for choice architects to help make incentives more salient—e.g., nudging people to reduce energy consumption via cost-disclosing thermostats may be more effective than increasing the actual price of energy.

Defaults are profoundly important in light of what has already been said about inertia and the status quo bias, and they are even more powerful when they appear to represent what is normal or recommended. Moreover, defaults are unavoidable because in any node of a choice architecture system – private or public – there must be rules for determining what happens to agents if they do nothing. They may be set to preserve the status quo ante, but this may lead to undesirable consequences. Alternatively, they can require agents to make a choice, but this may be perceived as a nuisance and may even be unfeasible when the decision is sufficiently complex. Not all defaults are selected to make the chooser’s life easier or better, but as a general matter the authors think they should be.

Feedback is the best way to help people improve their performance, especially if it warns agents when things are about to go wrong. However, this involves a delicate balancing act because giving too much feedback renders warning systems useless as agents begin to ignore the information altogether.

Errors are inevitable and should be anticipated in advance, yet some error-forgiving innovations are surprisingly slow to be adopted. One representative example is post-completion error: after finishing a task, people tend to forget things relating to previous steps. Thus some (but not all) ATM machines forces users to reclaim their cards before getting their cash; without this design feature, people often forget their cards in the machine. Likewise, Google Mail asks users “Did you forget your attachment?” when they mention the word “attach” in their message but fail to attach documents to the email. Similar thinking applies to “drug compliance” in health care regarding the optimal frequency with which patients are asked to take their medications, etc.

People are well equipped to analyze small numbers of well-understood alternatives, but as choices become more numerous and/or vary on more dimensions, social science research reveals that we adopt simplifying strategies that can get us into trouble. Choice architecture can step in and help people process their options—for example, consider how paint stores organize their products with color wheels, or how Netflix organizes movies by actor, director, genre, or collaborative filtering (which provides recommendations based on the preferences of other movie lovers). Likewise, the authors recommend mild government “RECAP” Regulations (Record, Evaluate, and Compare Alternative Prices) for products with complex, non-transparent pricing schemes, such as credit cards, mortgages, cell phone plans, and insurance policies.

Part II: Money

Chapter 6: Save More Tomorrow
Classical economics suggests that people calculate how much they will earn over the course of their careers and how much they will need after they stop working. Then, they save up just enough to enjoy a comfortable retirement without sacrificing too much while they are still working. While this provides excellent advice for how people should think about saving, it utterly fails to accurately describe how people do think and behave when it comes to financial planning.

This account fails to capture two crucial constraints. First, people have bounded rationality with respect to understanding and solving the mathematical problems involved in figuring out what to do. Second, people have bounded willpower with respect to implementing these plans. How can choice architects address the first part of saving for retirement, i.e. joining a plan and deciding how much to invest?

Most public social security systems provide defined-benefit plans that entitle participants to a benefit based on their salary and time worked under the plan. These are well designed in that they are forgiving of mindless human mistakes, requiring participants to decide only when to retire and when to start claiming benefits. The authors identify this as an ideal domain for nudging: because people have to make only one decision per lifetime, choice architecture can help them do a better job of weighing the relevant factors. They endorse automatic enrollment in savings plans unless and until participants decide to opt out. This can be augmented by a Save More Tomorrow program, which provides for automatic increases in contribution rates over time.

Chapter 7: Naïve Investing
How should we invest our money? The switch from defined-benefit plans to defined-contribution plans gives us more control, options, and responsibility over this decision than ever before. Investors must first think rigorously about how much risk to take, and then revisit their decisions periodically based on past performance.

People make many mistakes in this domain, in part because they are unduly influenced by short-term fluctuations (thanks largely to risk-aversion), and in part because their decisions are likely to be based on rules of thumb that are inadequate to the task (e.g., “it’s a bull market,” “when in doubt, diversify”). Fortunately, a wide variety of nudges can help:

  • Public-spirited companies could default their employees into a Save More Tomorrow plan that sells their company stock over a period of time, with the proceeds directed into a diversified portfolio.
  • Defined-contribution plans could offer a set of model portfolios with varying degrees of risk, or plan sponsors could offer participants target maturity funds matched to their expected retirement date.
  • New enrollees could be defaulted into a boilerplate investment plan carefully selected by knowledgeable experts, and participants who want to be more involved could be offered a menu of plans commensurate with their level of sophistication.
  • Anticipating the common error of not revisiting pre-existing investments, automatic rebalancing plans could adjust peoples’ asset allocations.

Chapter 8: Credit Markets
Americans borrow more than they save, in part because of the natural tendency to value present benefits over future costs. Adding more options can make people better off, but only if they can successfully pick the loan best suited to their situation and preferences. Often it’s best to ask people to take care of themselves, but when people borrow, standard human frailties can lead to serious hardship. Government should respect freedom of choice, but with a few nudges people would be far less likely to choose badly.

Mortgage shopping is more complicated than ever before: there are a wide variety of fixed- and variable-rate loans, along with interest-only loans, teaser rates, prepayment penalties, miscellaneous fees, etc. Unsophisticated and uneducated shoppers are especially disadvantaged by this complexity, not only because of predatory lending and the higher interest rates associated with subprime loans, but also through the sheer difficulty of matching people to the appropriate mortgages.

In response, the authors recommend that lenders report lending costs by fee and by interest. All the different types of fees would be reported, but they would be added up into a single salient number. This would ensure that borrowers at least know what their payments will be when the teaser rate ends. Lenders would also have to provide this information to independent third parties who might then offer advice by comparing lenders in a sophisticated way. This would make it easier to shop for mortgages online and would make the mortgage market more competitive. (The authors also analyze problems with student loans and credit cards.)

Chapter 9: Privatizing Social Security: Smorgasboard Style
Using the Swedish privatization plan as a case study, it is clear that the combination of free entry, unfettered competition, and lots of choice seems great, but does not guarantee the best possible social security outcomes. Maximizing choice does not always maximize utility, and for this reason it is crucial to select the default social security fund carefully. The key here is to decide what should be in the default portfolio and whether its selection should be encouraged or discouraged by the government in light of what we know about the population of participants.

Part III: Health

Chapter 10: Prescription Drugs: Part D for Daunting
In 2003, Congress passed a version of President George W. Bush’s proposed Medicare overhaul, creating a half-trillion-dollar federal subsidy for prescription drug coverage called Part D. Part D offered a menu with lots of choices and has done a lot of good, but as a matter of choice architecture it failed by:

  • Placing most seniors into non-enrollment by default.
  • Giving participants little guidance to help them make the best selections.
  • Choosing random defaults for six million automatically enrolled people, while refusing to match people to plans based on their drug histories.

Random assignment is especially objectionable as it can cause random harm to unlucky people placed in the wrong plans, and it is inconsistent with the market-based philosophy that better products and services merit larger market share.

To make matters worse, the Medicare Part D website and associated information sources are very unfriendly to users, and because the prescription drug plans are constantly updating their drug prices, there is no guarantee that the best plan today will be the best plan tomorrow.

Part D is great as a matter of maximizing free choice, but it stands to benefit greatly from better choice architecture, beginning with the intelligent assignment of people and plans via a “RECAP” nudge that helps people to evaluate their own health plan needs in comparison with other possible options.

Chapter 11: How to Increase Organ Donations
The demand for organs greatly exceeds the supply. In the US, most states require explicit consent for people to donate their organs. This rule depresses the number of organs that are harvested each year to rates below peoples’ stated willingness to become donors: many who report interest in donating end up failing to check their driver’s license forms accordingly.

Presuming consent has been proved a very effective way to increase the supply of organs for transplant, and it is completely consistent with libertarianism in that everyone still has the ability to opt out of donating their organs post mortem. But because this is politically controversial in the United States, the second-best alternative is to mandate choice. This could be implemented through a simple addition to the driver’s license registration scheme requiring licensees to actively say “no, not willing to donate” rather than passively defaulting them to that answer.

Chapter 12: Saving the Planet
Governments the world over are concerned about protecting the environment from air and water pollution, pesticides and toxic chemicals, global warming, etc. In this domain, command-and-control regulations (e.g., technological mandates and across-the-board restrictions) have become the norm. While many on the right and on the left agree that such measures are perfectly appropriate, they go above and beyond the recommendations of libertarian paternalism.

It can be helpful to think about the environment as the outcome of a global choice architecture system that fails to align incentives and fails to provide adequate feedback on the environmental consequences of individual actions. Fortunately a wide variety of modest nudges can improve our ability to protect the environment, alongside or even in the absence of command-and-control measures:

  • Improve incentives by taxing greenhouse gas emissions and/or establishing cap-and-trade systems.
  • Improve feedback to consumers on the costs of pollution by implementing disclosure policies and hazard-communication programs—and because the media tends to target the worst offenders, such feedback can be the basis of an additional social nudge, namely a de facto “environmental blacklist” that current or potential polluters will strive to avoid).
  • Voluntary participation programs designed to assist companies in improving the environment (the idea being that, even in a free market, companies sometimes fail to use the latest, greatest products which might very well help save the environment while also saving them money).

Part IV: Freedom

Chapter 13: Improving School Choices
FDR’s notion of a right to a good education, while not in the Constitution, is a firm American cultural commitment. Though allowing vouchers with which to choose between competing schools is not a panacea, it has been shown to improve student performance. School choice can help even more when combined with the following nudges to help parents make sensible decisions for their children:

  • Providing fact sheets with better and simpler information about average test scores and acceptance rates, from highest to lowest, at schools available to a given child (this is effectively a version of the RECAP nudge).
  • Providing a suite of programs to nudge students toward college: requiring high school students to complete an application to the local community college, scheduling meetings with college counselors, allowing students to take standardized admissions exams free of charge, providing financial aid information, and offering tax consultations for parents.

Chapter 14: Should Patients Be Forced to Buy Lottery Tickets?
All healthcare customers in America have a right to sue their doctors for negligence, and this cost gets passed along and built into their medical bills. Many patients and healthcare providers would gladly enter into an arrangement according to which doctors, hospitals, or insurance carriers reduce the price of healthcare in exchange for a waiver of the right to sue for medical malpractice.

However, courts currently treat such waivers to be void as against public policy. The legal justification for tort liability in this context is that it deters negligence, but this argument is undermined by the fact that doctors pay the same premium no matter how many times they have been sued for malpractice. Moreover, most patients don’t sue even if their doctor has been negligent, and those who do sue often end up with favorable settlements even when they don’t deserve the money. Jury awards for pain and suffering tend to be erratic and unpredictable, as does the occasional awarding of punitive damages, making the right to sue look more and more like a “lottery ticket.”

Freedom of contract, alongside a few nudges, might work better in this context. If waiving the right to sue were the default, and retaining it would cost extra, most patients would probably waive. Alongside such waivers, states can also lower the administrative and litigation costs of medical malpractice by capping “noneconomic damages” (e.g., pain and suffering).

Chapter 15: Privatizing Marriage
Marriage confers many major material economic and non-economic benefits:

  • Tax benefits/burdens, especially if one spouse earns much more than the other.
  • Entitlements, e.g. employers must allow unpaid leave for workers to care for their spouses under the Family Medical Leave Act.
  • Inheritance and other death benefits, e.g. avoiding estate taxes.
  • Automatic ownership benefits.
  • Surrogate decision making, e.g. in cases of incapacitation.
  • Evidentiary privileges, e.g. a right to keep marital communications confidential and to exclude adverse spousal testimony.

People have an accurate sense that ~50% of marriages end in divorce in general, but because of the self-serving bias nearly all newlyweds believe that they are almost certain not to get divorced. Thus couples are reluctant to get prenuptial agreements, and it is essential to design good default rules to govern contractual arrangements between domestic partners. This helps give people a clearer sense of their rights and obligations and protects those most vulnerable, namely women and children. For example, such default rules might provide that special help will be provided to those who have been the primary caretakers of the children.

It is crucial that these rules be extraordinarily clear, because people have a tendency to think that objectively “fair” outcomes skew in their favor, especially in difficult or important negotiations like marital disputes. Unrealistic expectations lead people to hard bargaining, making disputes longer, more intense, and more expensive. These costs can be avoided by anchoring peoples’ expectations of the possible outcomes through a narrow range of possibilities within which a judge has discretion to consider other factors, akin to criminal sentencing guidelines. The best approach might be an explicit formula based on such factors as the ages of both spouses, their earning capabilities, the length of marriage, and so forth. Starting with the formula as an anchor, a judge could weigh other considerations such as the standard of living during the marriage, the health of the spouse seeking maintenance, the financial prospects of both sides, and other relevant factors.

Part V: Objections & Recommendations

Chapter 16: A Dozen Nudges

  1. Give More Tomorrow would give people the chance to give a small amount to their favorite charities starting sometime soon, then commit to increasing their donations every year. Opting out would be easy, but this device would help those who want to give in general but forget to make good on their good intentions.
  2. Charity Debit Cards would be issued by banks and accepted only by charities, making donating simpler and more attractive by making it easier to keep track of contributions for tax purposes. Accounts could also reflect non-monetary donations, and statements could be sent directly to the IRS so that the government could automatically process the appropriate deductions.
  3. Automatic Tax Returns for anyone who does not itemize deductions and has no income that is not reported to the IRS. To file, the taxpayer would need only to sign it and mail it (or, even better, go to a secure IRS Website, sign in and click). This is estimated to potentially save 225 million hours of tax preparation time and $2 billion a year in tax preparation fees.
  4. Stickk.com allows users to pre-commit to their goals. It provides them with a platform for publicizing their plans and, if they fail, it triggers automatic donations to charities they love—or, even more provocatively, charities they hate.
  5. Committed Action to Reduce and End Smoking is a savings program in which the participant opens an account with a minimum balance of $1. For six months, any money that would have gone toward cigarettes is deposited into the account. Then, the client takes a urine test and, if no recent smoking is detected, the money is returned. If the user fails, however, the account is closed and the money is donated to charity. Evidence shows that opening an account makes those who want to quit smoking 53% more likely to achieve their goal.
  6. Special licenses for motorcycle riders who don’t want to use helmets, requiring extra driving courses and proof of health insurance—less intrusive than a ban, more preventative than making helmets entirely optional.
  7. Gambling self-bans through which addicts can keep themselves from being admitted to casinos, collecting gambling winnings, etc.
  8. Destiny Health Plans incentivize people to make healthy choices by giving them “Vitality Bucks” when they do things that reduce the probability of medical bills (e.g., working out, getting good results on a blood-pressure check).
  9. Dollar A Day programs pay teenage mothers $1/day for each day they are not pregnant. Because young mothers often have additional babies within a year or two and often lack the resources to cover the expenses involved, these costs are frequently passed on to taxpayers, so such programs are promising and cost-justified.
  10. Red Light Alerts warning users that their air conditioners need new filters.
  11. No-Bite Polish allows users to plan ahead and discipline their Automatic System 1 responses by making nail-biting painful and unpleasant. Likewise, alcoholics can take Disultiram so that any future alcohol consumption, however slight, becomes nauseating.
  12. The Civility Check is a software program proposed by the authors which would detect rude email messages and warn the sender (e.g., “This appears to be an uncivil email. Do you really and truly want to send it?” or “This appears to be an uncivil email. This will not be sent unless you ask to resend in twenty-four hours.”).

Chapter 17: Objections

“Nudges are a slippery slope toward much more intrusive interventions.”

  • This ducks the question of whether the authors’ substantive proposals have merit in and of themselves.
  • These proposals are designed to retain freedom of choice, and libertarian paternalism opposes the most objectionable government interventions.
  • Often nudges are inevitable, so we might as well set them sensibly.

“What mechanisms will prevent evil nudgers / bad nudges?”

  • Always maintaining the possibility of opt-outs preserves free choice and provides an important check on planners’ control.
  • Raising awareness of the impact of choice architecture should make planners more informed and capable overall.

“We have the right to be wrong, and sometimes mistakes help us improve.”

  • Opt-outs protect peoples’ freedom to make mistakes.
  • There is little harm in providing special warnings for novices, unsophisticated parties, and people who are vulnerable to manipulation.

Forced redistribution is wrong; all exchanges should be private and voluntary.”

  • Sometimes the optimal level of redistribution is not zero.
  • We must balance protecting the unfortunate and encouraging self-help.

Nudges are insidious, empowering government to maneuver people covertly.”

  • Nudges should be constrained using Rawls’ Publicity Principle: government should not select a policy that it would not defend publicly to its citizens.

Why stop at libertarian paternalism?”

  • Slippery-slope concerns begin to have some merit with respect to mandates without opt-outs, especially if regulators are heavy-handed.
  • Some such measures may at times be appropriate, e.g. cooling-off periods in contexts where decisions are unusual and emotions are likely to run high.

Chapter 18: The Real Third Way
For too long, our political system has pitted Democrats against Republicans. Democrats’ distrust for private markets leads them to insist on rigid mandates that can be uninformed, counterproductive, and antithetical to citizens’ liberty interests. Republicans in turn have adopted a laissez-faire platform that too often leaves markets under-regulated and results in an absurd, dogmatic antipathy for healthy governmental interventions. The Third Way laid out in this book involves meaningful government involvement while also promoting freedom of choice. The complexity of modern life and the emergence of new and profound insights into behavioral psychology should strengthen both the principled commitment to freedom of choice and the case for the gentle nudge.

Visit the authors’ website to see reader-submitted nudges: www.Nudges.org.