The Wrong Way to Get People to Do the Right Thing

May 12, 2015

The Wrong Way to Get People to Do the Right Thing

By Alfie Kohn


[This essay is adapted from The Brighter Side of Human Nature, which contains the complete references to research cited here.]

Our culture may sing the praises of the heroically selfless, yet it also seems to disdain the very idea of helping. We view prosocial behavior with suspicion, and even when we do commend it to our children and to each other, we tend to do so by invoking the idea of “enlightened self-interest.” Helping is almost always justified in terms of what the helper may get out of it.

Fairy tales, for example, are replete with magical creatures or royalty who pose as unfortunates and reward charitable acts. The same moral is found in magazine ads for charities that carry headlines like “How Helping a Poor Child Will Make You Feel Rich.” And on a billboard that stands not far from my home: “Lend a hand! A friendly act is rarely forgotten.” Many religious traditions promise that those who act rightly will be rewarded with a pleasant afterlife or a state of enlightenment, sometimes by earning the equivalent of cosmic merit badges. In a newer version of this egoistic appeal — one not restricted to children or the religious — we are told that volunteering our time for good causes will yield a healthier immune system, lower cholesterol levels, more sex, and longer life.

What happens when this same frame of mind drives public policy? Here are excerpts from three newspaper articles, representative of hundreds of others, that happened to appear within the span of a single month:

A Federal commission today called for a comprehensive policy to combat infant mortality, saying that it costs far less to support health care for pregnant women and newborns than to pay later for critical and long-term care for infants.

“Family violence costs employers millions of dollars annually in lower productivity, turnover, absenteeism and excessive use of medical benefits,” said Deborah Anderson, executive vice president of Responses [to End Abuse of Children, Inc.]

Those [illiterate people] who cannot find work end up costing the Government billions of dollars in welfare and unemployment compensation. Some end up turning to crime.

The message here is that we ought to try to reduce infant mortality, family violence, and illiteracy because it is in our own best financial interest to do so. If we do not act now to help suffering people, it will only cost us more down the road. The purpose of such declarations, of course, is to create or mobilize a constituency to support a given program. I don’t mean to deny the need for such support in light of scarce resources and other contenders for funding. The point, rather, is to emphasize that the criteria for disbursement invariably relate to financial self-interest. If we are to choose between a campaign to end homelessness and some other budget item, we’ve already accepted that the first (and perhaps last) question to ask is: Which will benefit us taxpayers more?

Our economic system is based on just such calculations, and on reducing non-economic issues, including such “externalities” as human misery, to dollars and cents. A corporation exists for only one reason, and that is to provide a favorable return on investment to those who own it. From the corporate sector one therefore fully expects to encounter a worldview in which wife-beating is disapproved mostly because it leads to the excessive use of medical benefits. But since the business of America is business (as the well-known Marxist Calvin Coolidge once put it), this perspective seeps out of downtown skyscrapers and suburban industrial parks and becomes the lifeblood of public policy, too.

Thus our treasury is tapped to reduce infant mortality if and only if this can be shown to be a wise investment. (We do not ask: If preventing babies from dying is not a sufficient reason to tap the public treasury, then what’s the point in having one?) Soon we’re reasoning this way in our personal lives, justifying each decision about how to treat others in terms of what we’ll get out of it, or following Tom Lehrer’s wry musical advice to Boy Scouts: “Be careful not to do/Your good deeds / When there’s no one watching you.”

Policy analyst Lisbeth Schorr has declared (in a reporter’s paraphrase) that “the political will and economic resources to expand [programs to reduce teenage pregnancy and child abuse] will come not from altruism but from concerns about productivity and the high cost of crime and unemployment.” If that’s true, let’s pause to reflect on just how profound an indictment we have thereby leveled against the society in which we live.

While we’re at it, let’s also pause to ask whether this assumption is even true. The issue is generally framed so that those who would make an appeal to something like altruism — who say we should stop domestic violence or provide care to newborns just because it’s the right thing to do — are perceived as naive. Those who appeal to the self-interest of the donor, on the other hand, are seen as shrewd pragmatists, willing to trade lofty moral appeals for references to the bottom line in order to get the job done.

The problem is that, from a long-range perspective, this may be a self-defeating strategy. To understand how this is true, we must take a step back and recall the line of research finding that rewards tend to undermine the very things they’re used to promote. Remove the reward and there’s no longer any desire to continue performing the task. Assuming that we seek to nurture an attraction to the given activity (learning, painting, inventing, etc.), this evaporation of interest is disturbing enough in itself, but it also has the consequence of yielding poorer quality work in many instances, as scores of studies have confirmed.

One implication is the folly of appealing to self-interest by offering rewards for prosocial behavior. Consider some of the experimental evidence:

* Second and third graders who received rewards for donating to another child — or fines for not helping — were less likely to explain their own behavior in words suggesting an intrinsic motivation to help than were children who received no rewards or punishments.

* Among children who gave away some of their game winnings after watching a model do so, those who were told they had made the donations “because you’re the kind of person who likes to help other people” — or who were simply told they had “shared quite a bit” (without any attribution) — were subsequently more generous than those told they had donated because they were expected to do so.

* The likelihood of donating increased in one study both when children were praised and when they were led to think of themselves as helpful people. But in a follow-up experiment, it was the latter who turned out to be more generous than those who had received verbal reinforcement. In other words, praise promoted generosity in a given setting but ceased to be effective outside of that setting, whereas children with an intrinsic impulse to be generous continued to act on that motivation in other circumstances.

* Adults who were offered money for agreeing to help a researcher rated themselves as less altruistic than those who either received no payment or were told after they said they would help that they’d be paid. “Extrinsic incentives can, by undermining self-perceived altruism, decrease intrinsic motivation to help others,” the researchers concluded. “A person’s kindness, it seems, cannot be bought.”

* Women offered money for answering a questionnaire over the telephone were less likely to agree to a similar request two or three days later in which no money was involved as compared with women who had not been paid for their participation in the first survey.

* Repeated blood donors who were reminded of altruistic reasons for giving blood indicated that they were more willing to do so again than those who were reminded of the personal benefits of what they were about to do. Focusing attention on personal benefits actually reduced the motivation to go through with the donation.

The lessons of these studies, when taken together, are straightforward. When we are rewarded for prosocial behavior, we tend to assume the reward, and not altruism, accounts for our having acted as we did — and that’s also true of verbal rewards (praise). If we don’t see ourselves as altruistic, we’re less likely to act prosocially once the extrinsic reward for acting that way is withdrawn. Conversely, encouragement to think of oneself as a generous person — an appeal not to self-interest but to genuine altruism — seems to be the most reliable way to promote helping and caring over the long haul and in different situations.

It’s hard to imagine that anyone familiar with this line of research would ever dangle some sort of prize in front of students as an inducement to participate in a charity drive. In fact, the only thing more counterproductive than offering a reward for raising money (or collecting food) for the needy is offering an award — that is, setting students or classes against one another in a contest to see who can triumph over everyone else in being the most charitable. You can almost watch kids’ concern for others evaporate before your eyes.

There is a lesson here, too, for those who propose to clean up corporate scandals and rein in the naked greed of the business world by proclaiming that “ethics pays.” Such a strategy serves principally to legitimate personal gain as the only relevant standard for one’s behavior. If a situation arises in which immoral conduct really is more lucrative, we have offered no basis for rejecting it.

More broadly, this discussion points to the existence of two distinct self-fulfilling prophecies. First, we tend to “live down to” the assumption that we are basically selfish, or live up to the assumption that we are inclined to act prosocially. These orientations feed on themselves. Helping is like lying: it’s hard to stop with just one generous act. Second, we also live down to the view that when we do help others it’s only for egoistic reasons, or we live up to the view that we are basically altruistic. What we believe to be true about ourselves and others affects how we behave. And that, in turn, affects our assumptions about human nature.

Consider a likely developmental scenario: A young child attributes altruism to others on the simple basis of how many helpful acts they perform. Unless he becomes a behaviorist, he will at some point grow out of the tendency to just count behaviors; instead he will begin to think about why people act to benefit others. Where he sees such behavior being rewarded in one way or another, he assumes that the reward accounts for the inclination to help. Since he himself is offered extrinsic incentives, he figures that he, too, is generous only because of what he gets out of it. In fact, he eventually concludes, whenever people seem to act altruistically, there must be egoistic reasons for it.

Thus there is no need to call on the idea of “human nature” to explain widespread selfishness. The use of rewards, and invocations of self-interest, to promote generosity creates a powerful, self-reproducing framework that explains this sad state of affairs all too well.

To be notified whenever a new article or blog is posted on this site, please enter your e-mail address at