One of the most enduring public policy issues in the United States is how to provide income for the elderly after they retire from the workplace. Many European countries provide workers with tax-supported pensions, while the United States provides only bare-bones support through Social Security, and therefore all employees and employers are encouraged to provide their own nongovernmental retirement programs. From a public policy perspective, the goal is to influence employees and employers to develop their own pensions without relying on political force. Cass Sunstein argues that liberal democracy governments may “nudge” employees to plan for their postretirement future but must avoid mandating contributions. Reference Sunstein1
Cass R. Sunstein’s 2016 book The Ethics of Influence: Government in the Age of Behavioral Science provides several examples to clarify the difference between nudging and shoving. Nudging, according to Sunstein, is gently encouraging people to make choices that are likely to benefit the chooser and society in the long run. Nudging means that the freedom of choice is preserved. Nudging must not involve coercion or deception. For example, a salesperson may encourage someone with a back problem to buy the most expensive mattress in the store. This may benefit the purchaser, but the salesperson is doing so for his or her own monetary self-interest and not necessarily out of a desire to help the purchaser’s back. Shoving denies the freedom of choice, such as a ban or mandate. For example, the Social Security tax is a mandate that does not allow a worker to not contribute.
Sunstein gives an example of a gentle nudge that does not impose direct costs on the nudger or the nudged. Changing the default on a printer to print on both sides of the page is often considered to be good for the environment. The beneficiaries are the public as a whole and maybe the budget of an institution. The administrative assistant who makes the change in the printer does not personally benefit from making the change. A professor who is printing a batch of tests is free to make a change so the pages are printed only on one side but is unlikely to change the default unless he or she has a strong preference. Some of the benefits are internalized by the institution through cost savings, and some of the benefits are external, such as less waste and a cleaner environment. No cost is imposed on the assistant, and there are benefits to others. 2 For a nudge to not be a shove, it must be freedom preserving for the individual. In addition, the nudge must make the individual (and possibly society) as a whole better off. Enticements are not nudges. For example, paying someone to make double-sided prints would be an enticement, whereas changing the default is a nudge.
From the perspective of classical economics, free markets are considered to be the best way to distribute goods, services, and resources. When they come up short, that is often referred to as “market failure”; an externality, cost, or benefit that affects a party who did not choose to incur that cost or benefit is an example. The direct beneficiary of a pension plan is the recipient. A generously funded pension may mean a stable meaningful retirement. If most people have well-funded pension plans, society as a whole benefits. Even wealthy individuals may receive external benefits from widespread pension funding. Less poverty means more prosperity, and the size of the economic pie becomes larger, leaving larger slices for everyone. Less poverty also means that there is less need for charity or taxation. All society benefits if fewer people rely on public assistance or if fewer people are destitute. However, we live in a complex world in which impure paternalism complicates choices regarding retirement investment.
Government in the age of behavioral science
Before the age of behavioral science, a default setting could be set randomly. Now the concept of nudging is taught in social science courses, and it is frequently mentioned in the media. The Age of Influence mentions “nudge units” and behavioral insights teams organized throughout the world. 3 Because these concepts are now so widely known, governments may have a responsibility to consider how their decisions address issues such as welfare, dignity, manipulation, and choice architecture. A printer would usually be shipped with a default of printing on one side of a page, so that is how it stays. Employee retirement plans require employees to opt into contributing to the account. Changing these two defaults may benefit the individual, the organization, and maybe society as a whole. If implemented correctly, the change of default means that individuals do not feel manipulated and their dignity remains intact.
In The Age of Influence, Sunstein lists 34 potential government nudges and their accompanying level of popular support based on a survey of the population. 4 This list shows that all nudges are not equally accepted. One person’s nudge may be another person’s coercion. For example, the most accepted government nudges concern food labeling. Most people welcome the idea of genetically modified food labels and calorie count labels. These nudges steer people away from genetically modified organisms and high-calorie foods without infringing on individuals’ autonomy. In comparison, the default assumption that a person is Christian on the census form received the lowest level of approval. This nudge is more intrusive, as it makes assumptions about individuals’ social identity. The comparison of these nudges raises two considerations: first, how nudges may differ in their preservation of choice and decision-making, and second, that nudges may have unintended social and political consequences.
In an appendix, Sunstein includes a copy of Executive Order 13707, which says that the federal government should use behavioral science insights to better serve the American people. 5 While this order may be well intentioned, it is possible that the government behavioral scientists could be lobbied by special interests. While often well intentioned, personal biases and experiences make it difficult for research teams to be completely neutral. In addition, one administration’s behavioral science team may be vastly different from another administration’s team, creating opportunities for interests to influence the process of development. Consequently, it would be best if research teams limited their focus to developing nudges that are likely to be widely accepted.
According to Sunstein’s survey, automatic enrollment in retirement savings plans (subject to opt out) is the fifth most popular government nudge. A default option that selects participation in a voluntary employment-based retirement program implies that the political authority administering the nudge assumes that future uses of money are more important than satisfying current needs. The nudge unit may view an employee spending money on cigarettes and alcohol as frivolous. The individual may be better off if some of this money was saved for retirement. However, this may not always be the case. The individual may be using the money to pay down debt or to avoid debt entirely. In the case of carrying debt, the best use of the money may be to pay down current debts with high interest rates. Consequently, some individuals may choose to have the money now and spend it now, avoiding future high-interest debt.
Some people may pick the default regardless of what it is, implying that little thought is going into the choice. If the default is to not contribute to a retirement fund, fewer people will contribute. If the default is to contribute, more people will contribute. One explanation for this behavior could be ignorance. However, complexity is also a likely culprit. A new employee is likely filling out lots of paper work in a short span of time. While he or she could take it home and read it carefully, the expectation is to complete it quickly. A nudge buried in paperwork is not a nudge. When an authority uses legal-speak and paperwork to move people toward a particular choice, it is coercion. A paternalistic nudge requires no tricks. The participants must be fully aware of the choice they are making.
Automatic enrollment in a pension plan could be considered an “injunctive norm,” which Sunstein defines as the norm about what people think that people should do. Most people think that they, and others, should save more. Thus, a government default that enrolls government employees into a retirement account encourages people to do what people think should be done. Automatic enrollment is not manipulating people to the desires of the choice architect, it is getting people to do what people collectively think should be done. A National Bureau of Economic Research study indicates that a “contribute” default increased participation by 25%. Reference Beshears, Choi, Laibson and Madrian6
John Stuart Mill was concerned that the manipulator lacks the knowledge to be able to properly decide for the chooser (see Ronald White’s essay in this issue). Even a benign manipulator would not be able to see the world according to the chooser’s tastes, preferences, and values. According to Sunstein, Mill was wrong to assume that people will promote their own welfare. 7 Thus, a default to contribute to a savings account is justified. Such a default would lead to an improvement in welfare for the chooser. The choice is encouraged by the choice architect (government), and the chooser is better off by accepting the nudge.
However, a default to contribution is an encouragement of a particular action. A default to noncontribution is preserving the choice purely for the chooser. The chooser must be knowledgeable about the choice, understand it, and make his or her own choice based on tastes, preferences, and values. The chooser cannot rely on a default to do the right thing. A default to automatic contribution is a signal that someone (perhaps the nation’s nudge unit) thinks that contribution is the correct course of action.
The concept of libertarian paternalism indicates that government should nudge citizens and be freedom preserving. Bans and mandates are not nudges because they are freedom diminishing. Choice complexity diminishes freedom if the complexity results in the chooser being unaware of the nudge. In addition, governmental nudge units may write documents in complex legal terminology. Clear, concise documents are needed to create freedom-preserving nudges.
Sunstein’s book is focused on government nudging. A government using a contribution default for its employees is different from a private firm using a contribution default. Contributing to a pension plan is impure paternalism, as it benefits third parties. Anyone who owns stocks or other investments benefits from others who buy in. A default to a retirement account contribution increases the value of accounts already in existence. A large company may encourage its employees to buy stock in the company. Such a move is transparent since the employee knows why he or she is being encouraged to buy in. However, when a government encourages that same course of action, the third-party beneficiaries are less clear. These third parties could be knowledgeable about the composition of government retirement accounts and use that information for personal gain.
Private companies may also wish to encourage their employees to contribute to retirement plans. The knowledge that a “contribute” default leads to more participation than a “noncontribute” default should be enough to help human resource departments design their forms accordingly. A government mandate or nudge unit recommendation is not needed to encourage this change. Currently, companies are changing their printer defaults without a government mandate or advice from a nudge unit.
Sunstein points out that complacency is a justification for nudging. However, it is possible that too much governmental nudging could lead to a different kind of complacency. If choosers see that the default is to “contribute to the retirement account,” they may now think that action is the best action. If employees sees the default is to “not contribute to a retirement account,” they may judge that the choice is theirs to make. The difference is that the latter is a course of nonaction. The chooser triggers the action. At the end of the book, Sunstein lists 34 potential government nudges. A person who is frequently nudged may start to believe that the choice architects are enlightened experts and that following their advice is the correct course of action. If an individual relies too much on the suggested nudge, that could result in a contribution to a retirement fund when the first best use of the money may have been to pay off high-interest debt.
An unintended consequence of paternalistic nudging is an overreliance on experts. Government experts are unlikely to be benign and fully knowledgeable. They may be captured by lobbyists in much that same way politicians are now. A default to noncontribution to a retirement account is more likely to be freedom preserving. It may result in some people saving inadequately for retirement but, if combined with an endless array of other government nudges, may lead to complacency. Default nudges toward nonaction are more freedom preserving. For example, the least popular nudge on the list is a default assumption of Christianity for census data. This may make the United States look like a Christian nation, but that may not reflect Americans’ true tastes, preferences, and values. A default to “no answer” would be more accurate.
While I have focused on problems associated of retirement savings nudges, there may be some external benefits of a “contribute” nudge for society as a whole. Contributing to a pension fund may mean that the individual is less likely to receive public assistance in the future. It is possible that contributing to a fund will benefit future taxpayers who will not be burdened by the chooser’s lack of contribution. In the United States, Social Security is designed to be the bare-bones safety net to diminish the negative externality. If private retirement accounts were widely funded, government programs could focus on the destitute, not the population as a whole.
Governments may wish to encourage private companies to get more of their employees to contribute to a pension plan because of externality. A default to enrollment (as opposed to a default to not enrollment) on standardized forms may be a simple way of doing this. This is less paternalistic than a Social Security contribution mandate. However, because these funds are often linked to the stock market, they also carry more risk (and potential for a larger return) than Social Security. In addition, third parties will benefit from this default, making it impure paternalism. While government may want greater contributions from private sector employees, there is no need for a government mandate. Private firms are capable of changing the default on their own. A legal mandate may lead to third-party lobbying and overly complex forms.
Summary and conclusions
Behavioral economists have shown that simply changing a default on an employment form from “not contribute” to “contribute” will significantly increase participation in retirement accounts. Public and private employers can use this knowledge to encourage participation in these programs, which (in general) might lead to happier future retirees. People will have more money during their retirement years if they opt in. Proponents of opt-out retirement savings nudges also argue that, over the long run, more participation in retirement savings programs will lead to a more prosperous future economy for all of us. This essay suggests that several issues complicate the ethics of retirement savings nudges. First, not all retirement funds are reliable long-term instruments for savings. Most of us lack the knowledge to make these long-term investments on our own; therefore, we invariably rely on agents, who may (or may not) serve our long-term interests. Second, saving for a distant future may (or may not) be the best use of an employee’s income in all contexts. Third, one must also address the question, at what point does a nudge become a shove (bans and mandates)? Opt-out nudges that require complex, time-consuming opt-out procedures can easily become a push. Fourth, most nudges, including retirement nudges, are “impure” and therefore benefit third parties. Critics of state paternalism observe that there is a natural propensity for political leaders to employ paternalistic intervention in order to advance the interests of those third parties, especially their own family, friends, and political allies.
In sum, in a liberal democracy, governmental nudges cannot alienate a large proportion of the population. There is no evidence suggesting that most Americans subsequently regret being nudged into saving for their retirement. However, we should all question the body of legality, especially tax laws, that mandate how, how much, and where Americans can save for that retirement.
