One of the great academic debates of our time revolves around how people make choices. On the one side, neoclassical theory assumes that individuals generally act in sensible ways in order to advance their individual self-interest. They are motivated to control aggression and monopoly, and to let private parties in competitive markets strike what bargains they like. In recent years, this neoclassical approach has come under attack from the field of behavioral economics. Its proponents argue that the neoclassical model of behavior, premised on the fact that human beings are rational decision-makers, does not sufficiently account for the many false heuristics and biases that lead people astray as they make decisions.
The two most prominent leaders in this movement are Richard Thaler of the University of Chicago, winner of the 2017 Nobel Prize in economics, and the Harvard Law Professor Cass Sunstein, who have advanced—most notably in their book Nudge--what they problematically call Libertarian Paternalism. This involves using both public and private institutions to “nudge” people to improve their lives without forcing them to do so, supposedly preserving their personal liberty. Resting on the foundational scholarship of Daniel Kahneman and the late Amos Tversky, Thaler and Sunstein deny that individuals are as rational as neoclassical theory holds: People often operate under the influence of systematic cognitive biases that prevent them from making sound decisions. In order to nudge people in the right direction, Thaler and Sunstein propose that the legal system set its “default” rules to induce them, without coercion, to act in ways that better advance their own welfare. In some cases, the switch is as simple as a move from “opting in” to “opting out.” People are permitted to reverse the default position if they prefer, so that their freedom of choice is thereby preserved.
In principle, libertarian paternalism consciously hopes to preserve freedom of contract by eschewing mandatory rules and relying instead on a framework of default rules. In practice, however, its approach is deeply flawed for both conceptual and practical reasons. Its initial conceptual mistake is to interpret paternalism far too broadly. In some cases, paternalism is absolutely necessary. Consider parenting. Parents, of course, nudge children to say “thank you” when they receive a present. But such nudges are only a small part of any legitimate system of parental control. Parents also defend their children against outsiders and impose highly coercive sanctions against children who misbehave. The logic in favor of this paternalism is that there is no superior alternative. Children cannot fend for themselves, and, for good biological reasons, parents are more invested in the well-being of their children than any stranger. Accordingly, in a free society, the state intervenes in parent-child relationships only in cases of abuse or neglect—when it is clear that the parents’ interests are no longer aligned with those of the child. That same level of state intervention may also make sense in extreme cases of suicidal and self-destructive behavior. But as libertarians of all stripes recognize, it is a dangerous premise for the decisions of everyday life—like where to work, how to save money, whom to marry, whether to have children, and how to raise them.
The first error of libertarian paternalism is its failure to vindicate its libertarian piece—namely a full-fledged endorsement of freedom of contract. Before fixing default rules, libertarian paternalism should work actively to dismantle the vast set of coercive legislation that hangs over competitive labor markets, including key bodies of law banning age, race, and sex discrimination in employment, or those imposing minimum wage and maximum hour regulations, or rules that otherwise ban unjust dismissal. But Nudge is silent on minimum wage laws, and it contains only this fleeting reference to employment discrimination based on race, sex, and religion: “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 explored here.”
No way. Generally speaking, the correct justification for antidiscrimination laws is that they serve as a counterweight to monopoly power—people have nowhere else to go. It is for that reason that monopoly unions should be (if allowed to exist) subject to duties of fair representation. In competitive markets, these laws chiefly undermine labor mobility by creating a massive administrative apparatus. If some vague reference to “Human errors” overcomes the libertarian opposition to antidiscrimination law in labor markets, authoritarian paternalism has found an all too comfortable home. Needless to say, Thaler and Sunstein do not advocate an antidiscrimination norm as a default rule that employers can explicitly override.
Unfortunately, libertarian paternalism does no better with its soft forms of intervention. In some instances, they conflate sensible market forms of innovation with paternalism. No private party can walk away from government regulation, even if he or she can quit a job if the terms are too onerous. Conceptually, it is a simple mistake for Thaler and Sunstein to count giving advice to clients and friends as a form of paternalism, libertarian or otherwise, when it is in fact one of the most powerful market mechanisms for correcting error. No one is omni-competent, so advice-giving allows voluntary markets to flourish without state coercion: People will eagerly pay for personalized advice that improves decisions even after they are bombarded with warnings and disclosures. Indeed, in virtually all contexts, any system of joint decision-making reduces error rates, so that the object of a legal regime should be to ease the path to advice-giving by removing onerous rules. This was done recently in Chamber of Commerce v. U.S. Department of Labor, when Judge Edith Jones struck down elaborate and misguided Obama administration regulations that would have prevented traditional brokerage houses from offering much-needed advice over their retirement programs and annuities.
In keeping with its general approach, however, libertarian paternalism obsesses about setting the appropriate default rules for retirement decisions. These rules can address the minimum amounts that should be set aside or the allocation of resources across different classes of stocks and bonds. Those default rules are best understood as implicit recommendations on how to proceed, but this overconfident agenda is subject to serious risks in implementation. First, the single default will not be suitable for everyone. The amount to save in retirement plans depends in part on how long a person expects to work and live, family obligations to spouse and children, assets accumulated in other forms, such as home equity, and current needs for the same income. It is a rash overgeneralization to assume otherwise.
In addition, the preoccupation with default rules misses an important distinction in libertarian theory: who sets the default rules? When employers introduce auto-enrollment plans, those are part of an employment contract. Any ostensible paternalist element disappears, just as it does when an employer mandates employee participation as a condition of employment. The key feature of these plans is that the distribution takes place with funds that are exclusively allocated to each individual employee. The plans are designed to keep workers satisfied.
Public retirement plans, in contrast, deeply offend libertarian principles. The pension plan disasters in states like California arise because the legislature, with a stroke of the pen, can increase pension benefits while leaving taxpayers holding a very expensive bill. Social Security, for its part, flunks on every conceivable standard. Its disclosure forms make it impossible for any person to figure out the present value of their plan benefits, or the rate of return on their individual contributions. The government wishes to conceal that information so that it can engage in massive forms of redistribution without facing political heat. The traditional economic theories of public choice help explain these serious risks of government misbehavior, to which behavioral economics adds nothing. This is a blind spot in libertarian paternalism which emboldens some of its critics to throw the individual retirees further into the fire by strengthening government mandates—and perhaps private retirement plans as well.
The same preoccupation with default rules threatens to derail the central freedom of contract principle in employment contracts—the employment at will that allows an employer to fire (or an employee to quit) for any reason at all. This principle (which is violated by the anti-discrimination and labor laws) is essential to preserve labor mobility. But Sunstein in particular thinks that it is advisable to set the default provision in favor of a for-cause dismissal rule, which is found virtually nowhere in unregulated private markets. The rule is supposed to protect employees who underestimate the risk of dismissal.
But again, it is a mistake to concentrate on one side of the labor market while ignoring the other. It is an even bigger mistake to overlook the common employment practice of offering a severance pay package to workers who are dismissed. That fixed sum serves several functions: it reduces the probability of a dismissal; it cushions the employee’s blow from a dismissal; it reduces administrative costs from separation; and it gives workers the right incentives to minimize their losses after dismissal while looking for new work. At this point, yet another vice of libertarian paternalism asserts itself. By focusing on cognitive glitches, it ignores all the other relevant features of an employment contract that are only mastered by taking a close look at each institutional arrangement that libertarian paternalists seek to tweak by default rules. By overlooking these issues, the theory has led to much misguided paternalism, and too little libertarian thinking. Both public policy and private markets would profit by junking this supposed behavioral approach and sticking with more traditional economic models.