We can harness peer pressure to uphold social values
People are natural followers, so use “behavioural contagion” to improve lives, says Robert Frank of Cornell University
“It’s the situation, not the person,” goes a mantra of social psychology. That is the idea that people are influenced as much by their surroundings as by their deliberative judgements and free will. If lots of people around us overeat and snacks are close by, we may too; if we’re in a circle that exercises, we might do likewise. Attitudes are as communicable as viruses: the environment we are in, the culture around us, matters.
The implications of this “behavioural contagion” is the subject of Robert Frank’s latest book, “Under the Influence: Putting Peer Pressure to Work,” (Princeton University Press, 2020). An economist at Cornell University in New York, Mr Frank considers what this means for incentives on activities and government regulation. Society almost never considers the “behavioural externalities” of policies. Hence, we justify smoking restrictions by the harm inflicted on others from passive smoking, not that more smokers subtly lead others to take up tobacco.
Mr Frank believes that society should acknowledge these externalities and harness behavioural contagion in the cause of broad social goals, such as reducing pollution and increasing the use of renewable energy. Ever the good economist, he also sees a role for tax policy to funnel people’s activities to account for behavioural externalities.
As part of The Economist’s Open Future initiative, we are publishing an excerpt from the book, on how smoking rates were dramatically reduced as it became less socially acceptable, less easy and more costly. We then ask Mr Frank about how this might be a model for other societal values, such as overcoming social polarisation.
Behavioural contagion Excerpt from “Under the Influence: Putting Peer Pressure to Work” (Princeton University Press, 2020) by Robert H. Frank.
When I started smoking at age fourteen in 1959, many of my friends had already been smoking for several years. My parents didn’t want me to smoke, but as smokers themselves, their objections rang hollow. In those days, more than 60 percent of American men were smokers, and almost as many women. Smoking was just something that most people did.
Yet even then, most people who smoked didn’t seem happy about it. Today, roughly 90 percent of smokers say they regret having started, and about 80 percent express a desire to quit. Almost half of all smokers try to quit each year, but fewer than 5 percent succeed. Several of my own attempts to quit failed. So I count myself fortunate to have abandoned the habit before leaving for college.
I have four adult sons, none of them a smoker. I once told a friend that if they’d grown up when I did, at least two of them would have taken up the habit. The reason I succeeded in raising nonsmoking children and my parents did not is that today’s environment is different from the earlier one. By far the strongest predictor of whether someone will smoke is the percentage of her closest friends who smoke. If that number rises from twenty to thirty, for example, the probability that she will smoke rises by about 25 percent. Whereas most of my teenage friends were smokers, relatively few of my sons’ friends were. Fewer than 15 percent of American adults now smoke.
Today’s environment is different mostly because of the taxes, prohibitions, and other measures we have taken to discourage smoking. In the 1950s, a pack of Camels could be had for as little as twenty-five cents in some parts of the country (about $2.15 in today’s dollars). But in many areas today, taxes have pushed that price north of $10, and in New York City a pack of cigarettes cannot be sold legally for less than $13. In the intervening years, we have also banned smoking in restaurants, bars, and public buildings. Some jurisdictions have prohibited smoking even in outdoor public spaces. We have spent billions of dollars on media campaigns to discourage smoking.
Given the long-standing American hostility to social engineering, each of these steps faced heavy pushback. When called on to justify them, regulators offered their time-honored response: restricting individual freedom is often the only way to prevent undue harm to innocent bystanders. By a wide margin, the example of harm to others they have most often cited has been that secondhand smoke causes injuries that bystanders cannot easily avoid.
Even strict libertarians concede the legitimacy of this rationale in principle. As John Stuart Mill, perhaps the Western world’s most eloquent champion of individual freedom, memorably wrote in On Liberty, “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. His own good, either physical or moral, is not sufficient warrant... Over himself, over his own body and mind, the individual is sovereign.”
That a desire to parry libertarian objections influenced regulators to invoke secondhand smoke in defense of antismoking measures is also consistent with their insistence that their aim is not to protect smokers from harming themselves. And health hazards from exposure to secondhand smoke have in fact been conclusively documented.
But are those hazards sufficient to justify extreme measures to discourage smoking? Unless you worked in a crowded bar with no ventilation, the damage caused by secondhand smoke was extremely small compared to that from being a smoker. Others defended regulation as necessary to protect taxpayers who bear the cost of smokers’ illnesses. But because smokers die younger than others and are less likely to die of lingering illnesses, they actually reduce government outlays on balance.
The narrow focus on secondhand smoke and fiscal effects greatly understates the harm that smokers impose on others. By far the greatest injury caused by someone’s decision to become a smoker is the harm caused by making others more likely to smoke.
When someone becomes a smoker, every friend of that person will have one more smoker in his or her peer group. Every member of every one of those groups will then become more likely to smoke. Those who take up the habit will then make each member of their own peer groups more likely to smoke, and so on. Each of the new smokers will suffer personal harm and each will inflict the real, albeit smaller, harms associated with secondhand smoke.
In short, when a regulation discourages someone from smoking, the harm to others that would have been caused by that person’s secondhand smoke or by pressure on government health-care budgets is only a minuscule percentage of the total harm actually prevented.
Many opponents of regulation are quick to argue, however, that behavioral contagion is not a proper justification for government intervention. It is one thing, they say, to protect someone whose asthma is aggravated by secondhand smoke, but quite another to penalize people merely because their behavior makes others more likely to smoke. People have agency, they insist, and it is the individual’s responsibility, not the state’s, to decide whether to smoke.
I applaud the sentiment that motivates this objection. People faced with the decision of whether to smoke do indeed have greater agency than those who are damaged by secondhand smoke. And all else equal, the regulators’ burden of proof clearly should be heavier in the first case than in the second. Yet smoking that results from behavioral contagion also harms many people who have no recourse—including not just those injured by the second-hand smoke emitted by new smokers, but also the additional parents who fail in their goal of raising their children to be nonsmokers.
When legitimate aspirations are in conflict, people’s freedom to do as they please will be limited no matter which way we turn. The claim that behavioral contagion constitutes a legitimate rationale for regulatory intervention against smoking is thus difficult to dismiss with slogans about individual rights and agency. Clearer thinking about behavioral contagion requires careful analysis of the trade-offs between competing types of freedom, which in turn requires difficult conversations about free will and other thorny philosophical issues.
Are these conversations worth having? This question becomes easier to answer once we examine contagion’s central role not just in smoking but also in excessive drinking, sexual predation, bullying, cheating, obesity, and other costly behaviors. More important, the behavior of others profoundly shapes our energy-use and spending patterns, in both cases aggravating already-profound threats to the environment.
As social psychologists like to say, “It’s the situation, not the person.” That brief sentence captures the longstanding scientific consensus that we can better predict what someone will do by examining the social environment than by looking at individual traits of character or personality. But the causal arrows run in both directions: the social environment is itself a consequence, in the aggregate, of our own choices.
Yet since each individual’s influence is so small, we tend to ignore this second pathway. Because the social environment influences us so strongly—sometimes for good, but often for ill—we have compelling reasons to shape it to our advantage.
The good news is that there are non-invasive policy measures that can promote that goal without demanding painful sacrifices from anyone.
Robert Frank: The term refers to the tendency for ideas or behaviours to spread from person to person in much the same way as an infectious disease. Like living organisms, ideas influence the environments they inhabit. Some are more adept at surviving than others and the most successful ones may or may not be good for us. Like a spreading disease, behavioral contagion can be injurious, as when a rise in the smoking rate causes additional people to take up smoking. But it can also be positive, as when someone’s decision to install rooftop solar panels stimulates neighbors to take similar steps.
The Economist: What are the upsides and downsides?
Mr Frank: Many view our tendency to be influenced by others as a regrettable human frailty, and often it is, as in the smoking example. Yet someone who was completely oblivious to others would be significantly handicapped. I have only a small share of the information needed to navigate safely through a complex and dangerous world. Others are ignorant too, of course, but collectively they know vastly more than I do. So when I see people confidently taking an unfamiliar action, it is prudent for me at least to investigate whether what they’re doing may make sense. But as most parents are careful to teach their children, people who exercise careful judgment in these situations will fare much better than others who follow blindly.
The Economist:Are behavioural remedies to problems a form of manipulation? It seems to undermine the idea of liberalism, that people are free to make their own decisions so long as it doesn't harm others. Does this not winnow human freedom, agency and responsibility?
Mr Frank: A good starting point for thinking about this question is indeed John Stuart Mill’s harm principle, which holds that the only justification for government to restrict someone is to prevent harm to others. (Mill must have meant “undue harm,” since it’s impossible to do anything without causing some harm, real or imagined, to someone, somewhere.) The biggest harm caused when someone takes up smoking is not the oft-cited injury from secondhand smoke; rather, it is causing others to become more likely to smoke.
People have agency, of course, and many successfully resist this peer influence. But what about parents who have done everything possible to discourage their children from smoking? When smoking rates increase, millions more parents will fail to achieve this laudable goal. Shouldn’t this count as undue harm? And does anyone regret that cigarette taxes and regulations have created a social environment in which our children are now 70% less likely to take up smoking?
No one disputes that the social environment shapes our behavior profoundly, sometimes for good but often for ill. I argue that we should adopt a similar pragmatic framework for thinking about whether it is permissible to employ policy measures that encourage us to behave as if we cared about how our own actions affect the social environment. Failure to permit a non-invasive measure like a tax on cigarettes to encourage a more benign social environment would be to protect one freedom (the smoker’s right to inflict harm on others) at the cost of destroying a much more valuable freedom (the right not to be harmed by smokers).
People who want to play the liberty card in this debate need to show why the particular liberty they favor should trump the competing freedoms that would have to be sacrificed to defend it.
The Economist: How can we harness behavioural contagion in the service of human dignity and values of truthfulness, respect for others, good governance, etc? After all, if policy-wonks got Americans to give up smoking over the past three decades, surely they can get them to give up their guns and xenophobia, no?
Mr Frank: Examples set by leaders are often the most effective way to promote supportive social environments and discourage destructive ones. But in many other cases, economists have long agreed that taxes and subsidies are the fairest and most efficient to encourage behavioral changes. My claim in “Under the Influence” is that behavioural externalities—actions that affect social environments—are exactly analogous to conventional externalities like pollution, noise and congestion.
Whenever practical, we have strong reasons to prefer taxation to prescriptive regulation as a remedy for behavioral externalities. Relative to a ban, it preserves the freedom of those who would find it most difficult to curtail an activity. And the revenue from such taxes also enables us to reduce current taxes on beneficial activities—like business hiring, which the payroll tax in America discourages.
The Economist: So what would that look like in terms of reducing gun ownership or increasing tolerance to people different than oneself? How can we practically apply what has been learned from smoking to other social values?"
Mr Frank: An intelligent answer to the question of what rights society should accord to smokers depends on a reasoned comparison of the costs and benefits of the alternatives under consideration. The same should govern how we think about the rights of gun owners. Their right to do as they please should be restricted only if it can be shown that failure to do so would result in undue harm to others. Considerable evidence bears on this question.
America, which grants extremely liberal gun rights, experiences more than 50 times the gunshot death rate as Britain, where gun rights are more highly restricted. Because simple regulations, such as requiring background checks and banning the sale of military-grade weapons, would reduce the US death rate substantially without requiring difficult sacrifices from gun owners, failure to have enacted such regulations is difficult to defend.
Statutes and regulations can also do much to defend the rights of people of different religious traditions and racial or ethnic backgrounds. In time, such measures also lead to more tolerant attitudes generally. Survey data, for example, reflected a marked decline in hostility to the LBGQT community in the wake of the Supreme Court’s 2015 decision to legalise same-sex marriage. But the example set by community leaders is also an important determinant of attitudes.
The Economist: If behavioural contagion is taken seriously, doesn't this suggest that American society is doomed, since it has cleaved into two tribes with their own leader, own media, own facts, own values—with less and less interaction with the other side. How could it be used to restore a common, mainstream ethos when "social influence" points towards further polarisation?
Mr Frank: On the contrary, taking behavioural contagion seriously would actually bring warring camps closer together. Like conventional externalities, behavioral externalities make the economic pie smaller. Any step to mitigate that problem therefore creates the possibility to divide a larger pie so that everyone gets a bigger slice than before.
A revenue-neutral tax on CO2 emissions is a case in point. Most of its revenue would come from top earners, who use much more energy than others. If that revenue were then returned to low- and middle-income households, as many as 90% of voters would receive monthly rebate checks that exceeded what they would have pain in carbon taxes. And since the wealthy stand to benefit most by slowing the planet’s warming, they too would come out ahead on balance. It’s political malpractice of the highest order that leaders have not made the effort to explain why taking steps like these would benefit virtually everyone.
The Economist: Certainly shared interests exist. But the more germane feature of society is that these “macro” ties have to compete with the “micro”: we are in tight subgroups that are our behavioural social circles. Red and blue states; Brexiters and Remainers. Can we break through that and if so, how?
Mr Frank: I take up this question in the last chapter, where I describe studies demonstrating the futility of most conversations between political opponents. These studies find that progress in such conversations becomes more likely if you can pose the right question, one that leads your counterpart to think about things in a new way.
An example: Most conservatives liked the Affordable Care Act’s requirement that insurers cover people with pre-existing conditions at affordable rates, but hated the requirement that everyone buy insurance. My initial attempts to explain why the mandate was essential were almost invariably futile. I began to make progress only when I began to ask the following question: What do you think would happen if the government required home insurers to sell fire insurance at affordable rates to people whose homes had already burned down?
People don't become defensive on hearing this question. On the contrary, they quickly recognise that the person whose home has burned down is exactly analogous to the patient with a pre-existing condition. And at that point they can explain to you why such patients wouldn’t be insurable without the mandate.