Bob Frank has a problem shared with few economists: he writes so clearly — and often so entertainingly — that people assume he must be a lightweight. In fact, the Cornell University professor has been ahead of the pack for decades, analyzing little-noticed aspects of modern economies that affect us profoundly. For example, in Winner Take All (written with Phil Cook fully a quarter-century ago), he presciently explained how the core institutions of our society and economy were driving us toward ever-increasing income inequality. Here, we excerpt a chapter from Frank’s newly published Under the Influence*, which challenges conventional wisdom on the limited justification for regulating individuals’ behavior in a free market. Most people concede the logic of managing “externalities” — say, by preventing one person’s bad habits from making bystanders sick. Frank goes much further, arguing it is justified to deter bad habits in order to prevent those bystanders from picking up the bad habit themselves. A step too far? Read it and decide for yourself.
— Peter Passell
Illustrations by Adam Nicklewicz
Published January 23, 2020
Robert H. Frank, Under the Influence: Putting Peer Pressure to Work. Princeton University Press (2020). All rights reserved.
Earlier in this book, I argued that behavioral contagion is a far more compelling reason to regulate smoking than is the conventional rationale of protecting innocent bystanders from the dangers of secondhand smoke. Compared to the direct harm experienced by smokers themselves, harm rom secondhand smoke is trivial. And because smoking is highly socially contagious, by far the greatest harm caused by smokers is the injury suffered by others who become more likely to smoke.
In my conversations with others about this proposition, the most consistent pushback I’ve received is that, although each premise of my argument might be true empirically, there remains a deep difference between injuries from secondhand smoke and those suffered by people who smoke because of peer influences. The former is a proper basis for government intervention, critics argue, because the victims have little recourse. In contrast, people have agency over whether they become smokers. They may be more likely to smoke if peers do, but it’s still their decision in the end. My critics worry that if regulators were to acknowledge that it is legitimate for the state to protect us from the negative consequences of peer influences, the result would be to diminish efforts to take responsibility for our own actions.
This objection has obvious rhetorical force. The fundamental attribution error — our failure to recognize that people’s actions are shaped more by circumstance than by traits of personality and character — does indeed raise difficult questions about individual responsibility. My aim here is to consider these questions seriously. But the case for taking behavioral contagion into account in regulatory decisions remains strong on balance.
Critics’ concerns are supported in part by evidence from experiments that probe how behavior changes when people are led to question the concept of free will. The concept is important because many of the laws and institutions of modern societies are grounded on the implicit or explicit premise that people are responsible for their choices and actions.
In most countries, for example, people who are found to have committed serious crimes are deprived of their liberty for extended periods, a step that many would find hard to defend if people lacked free will. The importance of free will is further underscored by the circumstances under which we exempt people from responsibility for acts that harm others. Under extreme conditions — say, when there is clear evidence of profound mental illness — the law refuses to hold even murderers accountable for their actions.
Yet scientists and philosophers continue to debate whether people have free will in any meaningful sense. The consensus among researchers has been moving slowly but steadily against free will in favor of the view that behavior is determined by genetic and environmental forces largely outside individual control. It’s not yet a settled issue, but if the anti-free-will position wins this debate in the end, many worry about the consequences for society.
It’s not an idle fear, for there is at least some evidence that believing in free will promotes a host of beneficial behaviors. As the psychologists Jasmine Carey and Delroy Paulhus have found, for example, those with stronger beliefs in free will also hold themselves and others to stricter moral standards.
The psychologists Kathleen Vohs and Jonathan Schooler have shown experimentally that priming subjects to question free will makes them more likely to steal money and more likely to cheat on a math test. The psychologist Roy Baumeister and his coauthors have shown that similar priming provokes a variety of other undesirable behavioral changes. It makes people less creative, less willing to learn from their mistakes and less grateful toward one another. Irrespective of whether people’s actions are in fact predetermined, then, at least some positive effects appear to accrue from believing that they are not. (Some of these findings, it’s worth noting, have failed to replicate in a recent study.)
The Limits of Free Will
It is clear, in any event, that situational factors affect our choices. Individual traits of personality and character matter, too, but such traits are themselves shaped largely by genetic and environmental factors beyond individual control. These observations would appear to imply at least some limits on the extent to which it is reasonable to hold individuals accountable for their actions.
In short, the deepest worry of those who oppose regulation based on behavioral contagion is that it would undercut people’s sense of responsibility. If their actions are predetermined, how could we blame them for robbing a bank?
Thinking about whether to rob a bank, a potential criminal weighs his lust for the money against not only whatever allegiance they feel toward norms against theft, but also the prospect of being caught and punished. If there were no penalty for robbing banks, many more people would rob them.
Although this concern is widespread, its logical basis is shaky. To deny free will, or to acknowledge that our behavior is influenced by external factors, is not to deny that we are responsible for our actions. It simply means that all actions have causes. When thinking about whether to rob a bank, a potential criminal weighs their lust for the money they hope to obtain against not only whatever allegiance they feel toward norms against theft, but also the prospect of being caught and punished. If there were no penalty for robbing banks, many more people would rob them.
That’s why every sensible society would maintain sanctions against robbing banks, even if none of its citizens believed in free will as conventionally defined. Free-will skeptics believe that every choice is caused by factors that exist prior to the moment of choice. Fear that one will be punished for violating laws or norms is simply one of those causal factors.
Some free-will skeptics acknowledge the possibility that questioning free will may make it more difficult for some people to resist temptation, but go on to argue that it also promises benefits. The philosopher Sam Harris, for example, notes that questioning free will encourages a more open-minded assessment of the environmental causes of behaviors that society wants to discourage. And this, he argues, may lead to more effective ways of discouraging those behaviors.
In the case of smoking, at least, it is difficult to see how acknowledgment of the importance of peer influences would make it more difficult for nonsmokers to refrain from smoking, or make it more difficult for smokers to quit. Even those who haven’t read the relevant evidence have probably either heard about it or else understand intuitively that peer effects strongly influence smoking decisions. Regulators could take peer influences into account without having to broadcast commentary about them.
In any case, accepting that our choices are often influenced by external forces is clearly compatible with the belief that society has a legitimate interest in holding us accountable for those choices. But because people faced with the decision of whether to smoke do have greater agency than those who are damaged by secondhand smoke, it is also fair to require regulators to meet a stiffer burden of proof in the first case than in the second.
What no one questions, however, is that an increase in smoking rates makes each nonsmoker more likely to smoke. It is thus a statistical certainty that if additional people take up smoking, that fact alone will cause further increases in the overall number of smokers, even though each additional smoker was free to have abstained. People who say regulators should ignore peer effects are thus saying, in effect, that regulators should ignore any injuries that people suffer as a direct result of their own choices.
When the benefit of an option occurs right away but the costs come only with significant delay, the option often becomes misleadingly attractive. Those who choose the early reward typically voice regret over having done so. Many awaken the next morning wishing they’d drunk less, while many fewer wish they’d drunk more.
To call this a controversial position would be an understatement. Although libertarians often object to seat-belt laws and bicycle-helmet requirements for adults, for example, many acknowledge the attractiveness of such measures for children. And children are the ones most vulnerable to peer influences. Opposition to attempts to shield children would therefore seem difficult to defend, even for those most inclined to believe that regulators should ignore behavioral contagion.
Many respected scholars have also defended regulations whose apparent aim is to protect adults from the consequences of ill-advised choices. An important class of examples involves decisions that entail self-control. Psychologists have long known that people often succumb to temptation to engage in behavior they would prefer to avoid. When an inferior but earlier reward becomes imminently available, many exhibit a strong tendency to choose it over a much larger reward that occurs with significant delay. Put another way: when the benefit of an option occurs right away but the costs come only with significant delay, the option often becomes misleadingly attractive. Tellingly, those who choose the early reward typically voice regret over having done so. Many awaken the next morning wishing they’d drunk a little less, while many fewer wish they’d drunk a little more.
The extent to which the state should try to shield people from the consequences of insufficient self-control of course remains a matter of spirited debate. When the negative consequences are sufficiently small, the consensus is that people should resist temptation on their own. But when larger outcomes are at stake, that consensus quickly breaks down.
A case in point is the regulation of radiation exposure in the workplace. Workers’ willingness to accept delayed risks to their health and safety for extra current pay is vividly illustrated by practices in the nuclear power industry, where it is occasionally necessary to clean up radiation spills. This task commands a significant and immediate pay premium, but also large and significantly delayed health risks from exposure to ionizing radiation. Even so, there is a ready supply of applicants eager to perform these tasks.
Federal regulations currently limit the amount of radiation to which these workers may be exposed, and workers get a bonus if they burn out, or exceed those limits. The cleanup workers, who are known in the industry as glow worms, invariably do burn out, often deliberately, and, were it not for the federal limits, would willingly expose themselves to even higher doses of radiation.
Do these rules constitute regulatory overreach? Reasonable people can disagree. On their face, they appear to violate John Stuart Mill’s position that “the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others.” But some philosophers have defended seemingly paternalistic regulations by elaborating on what Mill may have meant by the phrase “against his will.” Thus, as Gerald Dworkin at the University of California (Davis) writes:
There is an emphasis on what could be called future-oriented consent — on what the child will come to welcome rather than on what it does welcome. Extensions of paternalism are argued for by claiming that in various respects, chronologically mature individuals share the same deficiencies in capacity to think rationally and the ability to carry out decisions that children possess. Hence, in interfering with such people we are in effect doing what they would do if they were fully rational.
In this view of the problem, regulations that limit exposure to radiation are not really contrary to the will of those affected. Dworkin and others are arguing, in effect, that with the wisdom of hindsight, people would often resent not having been prevented from behaving in self-destructive ways when they were younger. Regulators who respect Mill’s plea must still decide whose will deserves greater weight — that of the current self or that of the future self? True, to rule in favor of the future self is to limit the current self’s freedom. But failure to restrict the current self ignores the future self’s well-considered wish to be protected against being harmed by his current self’s myopic choices.
Empathy is not the Enemy
Behavioral economics has carried this line of thinking much further. Scholars in this rapidly growing field work primarily at the intersection of psychology and economics, but also draw on insights from other disciplines. One of their most robust findings is that, quite apart from issues involving self-control, people often violate the fundamental assumptions of rationality, choosing in ways that fail to serve their interests. Collective action can often steer people toward choices they prefer without significantly restricting their freedom to choose as they see fit. Often this can be accomplished by simply rearranging the way choices are presented, without restricting behavior in any way.
But for the sake of discussion, suppose we adopt the extreme libertarian view that regulators should ignore injuries that people suffer as a direct result of their own choices. Even that view does not imply that regulators should ignore behavioral contagion, because smoking caused by behavioral contagion harms not only the new smokers themselves, but many others who have no practical means to avoid injury.
Consider, for example, those parents who have already taken every reasonable step to discourage their children from smoking. Given what we now know about the health consequences of smoking, could anyone second-guess their pursuit of this goal? Perhaps they have attempted to achieve it by refraining from smoking themselves, by choosing friends who do not smoke, by steering their children away from public spaces where others smoke, and by repeatedly advising their children about the dangers of smoking. Perhaps they could have done even more. But evidence from adolescent psychology suggests the possibility that more extreme measures might well have proved counterproductive. Push teenagers too hard, and they become more likely to smoke, not less.
What is certain is that in any large group of such parents, many more will fail to achieve their goal in environments in which a higher proportion of their children’s peers are smokers. These parents, like the victims of secondhand smoke, have no recourse. And although the harm they suffer may be hard to quantify, it is surely no less worthy of consideration on that account. And it is not just parents who are harmed. When a smoker dies prematurely, hundreds of other relatives and friends suffer, too.
Free-market enthusiasts may underestimate this harm because of their tendency to emphasize self-interest as the most important human motivation. Opponents of regulation, who are overrepresented in this group, often cite Adam Smith’s “invisible hand” approvingly. In their telling, market forces channel individual self-interest to serve the broader interests of society. It’s a powerful narrative, to be sure, yet Smith himself understood that self-interest alone wouldn’t create a just community. He believed that markets could function adequately only in the context of an elaborate foundation of laws and ethical norms of the sort he described in The Theory of Moral Sentiments, published almost two decades before The Wealth of Nations.
Individual interest and society’s interest coincide much of the time, but not always. For example, the individual business owner’s interest dictates erecting a sign that stands out relative to those of neighboring businesses. But when all owners pursue this interest without restriction, we get visual cacophony.
But even the most carefully crafted regulations aren’t sufficient. People must also be motivated not to violate them even when no one is looking. In Adam Smith’s view, empathy was a primary source of such motivation. As the primatologist Frans de Waal has written:
[A] fundamental yet rarely asked question is why natural selection designed our brains so that we’re in tune with our fellow human beings and feel distress at their distress, and pleasure at their pleasure. If the exploitation of others were all that mattered, evolution should never have got into the empathy business. But it did.
Empathy motivates good behavior even when narrow self-interest weighs against it. Some people pour unwanted pesticides down their drains rather than take the trouble to dispose of them safely. But most people don’t take this shortcut. Some people don’t leave tips after dining in restaurants they don’t expect to visit in the future. But most people do. Empathy is an important source of such restraint. People exercise restraint out of concern that failure to do so would cause pain to others.
Empathy’s most powerful effects are seen in the behavior of parents. People who have never had children sometimes wonder how they could endure several hours trying to comfort a colicky baby. Once they become parents, however, the child’s discomfort affects them more powerfully than their own, and failure to help is not even an option.
Some utilitarian moral philosophers insist that in a decision between saving the life of one’s own child and saving the lives of two strangers, the right choice is to save the strangers. But if parents were equipped with a psychology compatible with that choice, few children would receive the care required to emerge as well-functioning adults. Far better to live among people raised by parents who would save a child’s life over those of two strangers than to live among those raised by strict utilitarians.
Given the power and importance of parental empathy, it is hardly a mystery that parents hope their children don’t grow up to be smokers. But again, if many of their children’s peers are smokers, a large proportion of those parents are destined for disappointment. Do those who insist that peer effects are not a proper basis for regulating smoking believe that preventing the injury suffered by these parents is less important than protecting the right of people to smoke without restraint? If so, what arguments might support such a belief in the face of evidence that more than 90 percent of smokers wish they’d never started?
Any such arguments deserve careful scrutiny because peer effects are strongly implicated in many other forms of harmful behavior. Most parents want their children to grow up to be healthy adults, for example, but many more of them will become obese if their childhood peers regularly drink 32-ounce sugared soft drinks. Few parents want their children to become bullies, either, but here, too, the influence of peer behavior looms large.
Smoking causes great harm to others — most of it indirect — and the overwhelming majority of smokers acknowledge that their habit is personally harmful. Restricting smoking easily meets the efficiency test.
Even in the absence of persuasive arguments that regulators should ignore peer effects, lawmakers have largely confined their attention to behavior that causes significant direct harm to others. They have ignored the fact that when people suffer harm because of their own choices, others who care about those people also suffer harm. That asymmetry may help explain why regulators have been quick to cite the dangers of direct exposure to secondhand smoke as their defense for taxing cigarettes. As noted, however, that particular harm falls well short of the threshold necessary to justify the relatively extreme measures we have taken to discourage smoking. But if we also consider smoking’s indirect harm via behavioral contagion, even stronger antismoking measures could easily be justified.
Again, it’s the situation, not the person. That brief sentence captures the longstanding consensus among social scientists that we can better predict what someone will do by examining the social environment than by looking at individual traits of character or personality. Because social environments influence us so strongly — sometimes for good, but often for ill — we have compelling reasons to use public policy to shape them to our advantage.
The Long Arm of the Law
Yet the dangers of regulatory overreach are also evident. For example, in a 2015 New Jersey law prohibiting church organizations from selling memorial headstones. The law was championed by the Monuments Builders Association of New Jersey, whose members had seen their share of sales of headstones decline significantly after the Archdiocese of Newark, began selling them in 2013.
As the journalist Tanya Marsh described this law’s history, it “was adopted at the behest of a group of private market participants for a reason no more noble than to protect themselves from competition. This blatantly anti-competitive effort is even more stunning because the products at issue — headstones and memorial tablets — are not regulated.”
Laws like these result from “regulatory capture,” by which regulation is directed away from the public interest and toward the interests of the regulated industry, by the industry itself. Regulatory capture is widespread, but it is hardly the only source of bad regulation.
Consider again the question of safety regulation. The mere fact that a market failure may exist is no guarantee that regulations will reliably improve matters. Markets are imperfect, but so are bureaucrats. One of the most vivid examples of ineffective state intervention I have come across remains the following passage from the Occupational Safety and Health Administration’s 30 double-columned pages of safety requirements for ladders in its 1976 manual of workplace safety standards. A sampling:
The general slope of grain in flat steps of minimum dimension shall not be steeper than 1 in 12, except that for ladders under 10 feet in length the slope of grain shall not be steeper than 1 in 10. The slope of grain in areas of local deviation shall not be steeper than 1 in 12 or 1 in 10 as specified above. For all ladders, cross grain not steeper than 1 in 10 are permitted in lieu of 1 in 12, provided the size is increased to afford at least 15 percent greater calculated strength than for ladders built to minimum dimensions. Local deviations of grain associated with otherwise permissible irregularities are permitted.
But we can oppose regulatory capture and ineptly executed regulation without endorsing the view that all regulation is counterproductive. Adam Smith was correct that market forces often channel self-interest for the common good. But he also was a firm advocate of regulation as a remedy for the market’s shortcomings. Individual interest and society’s interest coincide much of the time, but not always. For example, the individual business owner’s interest dictates erecting a sign that stands out relative to those of neighboring businesses. But when all owners pursue this interest without restriction, we get visual cacophony.
This has led most cities to enact zoning regulations that limit the placement, height, surface area, brightness and other characteristics of signs, often with the enthusiastic support of business owners who are restricted by those restrictions. In the same spirit, zoning ordinances in most cities do not permit my neighbor to operate a pig farm in a densely settled residential area like ours.
Not even the most extreme libertarians are willing to argue that restrictions on individual freedom are never justified. To embrace that position would be to rule out traffic lights and laws against homicide. As even a moment’s reflection makes clear, the state’s refusal to limit anyone’s freedom of action would entail a wholesale reduction in other highly valued freedoms for everyone.
Scholars who work at the intersection of law and economics have argued that, over the millennia, the law has tended to evolve in the direction of efficiency. Although exceptions abound, laws and regulations that address behavior that causes harm to others have been revised in a clearly identifiable way: they are more likely to restrict people from acting as they please when the costs to those restricted are smaller than the harm to others thus prevented. There are cogent reasons to have expected this tendency, for as Nobelist Ronald Coase and others have emphasized, when the activities of two parties interfere with one another, it is in the shared interest of both to resolve the problem in the least costly way.
This perspective about the law does not imply that all cases involving behavior with negative side effects will be easily resolved. Costs and benefits are often difficult to measure, not least because of strategic posturing. But the same perspective also identifies many cases that are not close calls. Consider an activity that causes considerable harm to others, yet is not highly valued by those who engage in it. Smoking is exactly such an activity. It causes great harm to others — most of it indirect — and the overwhelming majority of smokers acknowledge that their habit is personally harmful. Restricting smoking easily meets the efficiency test.
But we have taken far more stringent steps to discourage smoking than would be warranted if our only aim were to protect people from secondhand smoke. Once we acknowledge behavioral contagion as a legitimate basis for regulation, though, it becomes clear that the regulatory steps we have taken have not been nearly strong enough. And smoking is not the only activity that fits this description.
Corporations are People, Too?
Few would deny the value of encouraging people to accept responsibility for the consequences of their own behavior. That belief has, as noted, encouraged many to oppose restrictions whose aim is to foster more supportive peer environments. It is each individual’s responsibility, these critics insist, to identify which peer examples are worthy of emulation and which are best to avoid.
Yet many who hold this view are also sympathetic toward analogous restrictions on corporate behavior. One reason for this asymmetry could be a general belief that, whereas our peers are seldom actively trying to harm us, the same cannot be said of corporations. Firms generally try to persuade us to serve their ends, even when those ends conflict with our own. More troubling, corporations also have enormous power to bend us to their will. When it suits them, they deploy every weapon in the modern marketing arsenal to induce us to engage in deeply self-destructive behavior. Such observations are plausibly cited as justifications for government regulation of firm behavior that harms others.
But the same observations also support regulating consumer behavior because of its effects on peer environments. Corporate marketers now understand, after all, that the surest way to boost demand for their products is to invest in making it more likely that potential buyers’ peers will recommend them. Billions of dollars once spent on ads touting a product’s features to consumers directly are now spent on social media campaigns aimed at launching viral conversations about those products. Businesses have fully grasped the power of peer influences and are heavily investing in new ways to harness them.
As a purely descriptive matter, it follows that to regulate business behavior is often to regulate for the explicit purpose of discouraging negative peer influences. How, then, can those who accept the legitimacy of such regulation consistently oppose the regulation of individual behavior for similar ends? Consistency, of course, isn’t everything. As Ralph Waldo Emerson wrote, “A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines.”
But not every desire for consistency is foolish. Shouldn’t the burden of proof here rest with those who argue against the legitimacy of regulating individual behavior to discourage harmful peer influences?
I should also note that ham-fisted prohibitions are not the only way to discourage activities that cause undue harm to others. For example, such activities are generally much more efficiently discouraged by taxing them. The tax approach embodies multiple advantages in comparison with prohibition and other prescriptive measures. Not least is that it affords those who especially value the activity to continue engaging in it. Another is that it also helps to pay for public services we value. No one enjoys paying taxes, of course, but every dollar we raise from a tax on harmful activities is a dollar less we need to raise from taxes on useful activities.
Regulators are well-advised to be cautious about using the power of the state to limit people’s freedom of action. But caution, like other virtues, is best exercised in moderation. Social influences are extremely powerful. By declaring them off-limits as a basis for intervention, we have been too cautious by far. We have foreclosed valuable opportunities to foster social environments that would help bring out the best in us. We could seize many such opportunities without demanding painful sacrifices from anyone.