The Mystery of Motivation
The most carefully crafted incentives, from cash rewards to social props, routinely backfire. What does it really take to get people to do the "right" thing?
By Gary Drevitch published January 3, 2017 - last reviewed on October 25, 2019
Beginning in 2011, Wells Fargo bank employees opened more than 2 million sham bank and credit card accounts in their customers' names, violating their trust, as well as charging them hundreds of thousands of dollars in fees. What drove so many agents to act so unethically? Incentive structures that tied a substantial piece of their compensation, not to mention their continuing employment, to steep sales targets.
When it became unrealistic for some workers to meet those targets solely by signing new or existing customers to legitimate checking accounts and credit cards, they pushed tractable clients to open separate accounts for groceries, pet care, and birthday parties. College students, senior citizens, and undocumented immigrants were frequent targets. Before eventually turning to outright fraud, some sales reps begged their family members to open "ghost accounts." The stress, one banker told The New York Times, led her to drink alcohol-based hand sanitizer at the office to cope. Another said he frequently cried in the bathroom and eventually landed in the hospital with an anxiety attack.
After the scandal was exposed last fall, the bank was fined $185 million, and it agreed to refund consumers at least $2.6 million. More than 5,000 sales reps were fired, and new business dropped precipitously. Given that most of the sham accounts held virtually no funds, the bank had actually profited little from their creation. The company's new head of community banking said that her first act would be to remove incentives that could promote bad behavior.
Decades of research and centuries of anecdotal evidence show that incentive systems like that of Wells Fargo can backfire, even when they are launched with the best of intentions. Poorly thought-out incentives to motivate people to be healthier, kinder, or greener can actually lead them to do the opposite, and incentives designed to spur workers to do their best can push them to do their worst.
"Research shows that if goals are unrealistic, but you can achieve them by cheating, then people will cheat. They will commit fraud to obtain the incentive," says George Loewenstein, a professor of economics and psychology at Carnegie Mellon University who is considered one of the founders of behavioral economics.
What Will Get You to Do What's Right?
Communities need their members to act in a civic-minded fashion—to obey traffic laws, to donate blood, to recycle or conserve water and energy—but when too few individuals are willing to take such actions on their own, policy makers may try to craft incentives to raise participation.
The ongoing challenge for behavioral economists is determining whether, and under what conditions, people will respond to economic incentives—bonuses, tax rebates—or to social pressure—tangible or intangible reminders to follow social norms and avoid opprobrium. The experts continue to discover that when it comes to getting people to do what they ought to, our first instincts are often wrong.
"The most common response is to design a set of incentives so that selfish people will act as if they were concerned about others and therefore do the right thing," says Samuel Bowles, an economist at the Santa Fe Institute and the author of The Moral Economy: Why Good Incentives Are No Substitute for Good Citizens. "It's a big mistake."
Purely economic incentives for doing the right thing, Bowles argues, are based on the outdated notion of motivating homo economicus, a rational, self-interested (and theoretical) individual. But in the real world, he says, such people represent a small minority of the population. "Economists came to think that prices could do the work of morals. But we know that there are some problems for which we cannot devise prices that will get people, if they're entirely selfish, to do the right thing." In fact, recent studies suggest that economic incentives may not only fail to achieve community goals, they may actually sideline people's moral instincts by replacing them with financial calculations. This is the "crowding out" phenomenon, and it appears to be such a powerful social reaction that it's been registered even in toddlers.
"When people are intrinsically motivated by altruism or by wanting to signal that they're good people, giving them even small incentives can backfire," Loewenstein says. In one study, when students raising money to fight cancer without incentives were offered commissions for bringing in more cash, they actually began collecting less. "They could no longer signal that they were virtuous people by collecting the money."
Poorly deployed incentives can poison the work of even the most altruistic professionals. Oncologists, for example, have been found to prescribe more chemotherapy drugs when their compensation is tied to the volume they administer. "If you take a group of people who are intrinsically motivated and start introducing all sorts of incentives for accomplishing different goals, then they focus on the incentives and lose track of why they are in the profession in the first place," Loewenstein says.
In December 2001, the Boston Fire Department responded to an increase in sick days taken by firefighters on Mondays and Fridays by eliminating its unlimited sick-day policy and imposing a new annual limit of 15. The predictable result: Over the next year, the total number of sick days taken more than doubled. When department leaders treated their workforce as homo economicus, the firefighters responded in kind by prioritizing their own self-interest; both sides' sense of duty to the citizens of Boston had been crowded out. "It sent a message that we don't trust you, we don't think you're actually a dedicated firefighter," Bowles says.
Putting a Price on Bad Behavior
In its power to instantly change one's mental equation, for better or worse, a monetary incentive is in a class of its own. If you asked a friend to help move your new couch, for example, you might offer him or her a beer when you're finished, but handing over $10 instead would be extremely awkward. The monetary offer suddenly turns your friendship into an exchange relationship, which we're comfortable with when we pay a store clerk for groceries but not in our intimate lives. A couple of beers may in fact have the same value as the money, but a cash exchange shifts a relationship from social to economic, just as giving a date a $20 bill instead of flowers with the hope of incentivizing her to share a goodnight kiss would instead be an offensive turnoff.
Even in relationships that are not personal, monetary incentives can backfire, says behavioral economist Uri Gneezy of the University of California, San Diego, co-author of The Why Axis: Hidden Motives and the Undiscovered Economics of Everyday Life. Consider an invitation to join a focus group trying a new headache pill. If the offer comes with a $50 incentive, you might accept out of a desire to help and a sense that the money is fair value for your time. But if the offer is $5,000, Gneezy says, you might be extremely reluctant: "'Why are you paying me $5,000 for an hour? There must be something wrong with this pill.' The fact that they're paying you this much sends you a signal that the drug may be dangerous. You might agree to participate, but for sure you'll be more worried than when they offered $50." For the same reason, municipalities struggle to locate public facilities like waste-treatment plants, even when they offer large financial incentives to nearby homeowners. The simple fact that the sums are large convinces residents of the danger and strengthens their resistance.
For one widely-cited study, Gneezy surmised that day-care centers that urged parents to pick up their kids on time, but without a specified penalty, would find that more parents, and not fewer, would be late once the sites imposed a modest monetary fine for tardiness.
The centers and the families had what economists call an "incomplete contract": You agree that you must pick up your kids at 4 p.m., but you haven't agreed on what will happen if you don't. Families may imagine that the day-care staff needs to rush home to their own kids, for example, and out of courtesy, push themselves to arrive on time. "But if I introduce a fine of $3, then you can say, 'Now I know that it's not that bad to be late.'" Researchers observed 10 day-care centers in Israel, six of which introduced a fine for late pick-up and then eliminated it 12 weeks later. The rate of late pickups at centers with fines more than doubled within a few weeks and remained at that level even after the fine was removed. In the parents' minds, the contract had been completed, the value of tardiness had been set, and it was not especially high.
The results mirrored Gneezy's own experience. Before his children's day-care center enacted a similar fine, he says, "I would drive like crazy to pick them up on time because that's what I was supposed to do. After that, I didn't, and I thought that other people would not be different from me in that respect."
Gneezy acknowledges that a higher fine—say, $10 a minute—might have had a much different impact. But the point he proved, as reflected in his paper's title, "A Fine Is a Price," was that "the introduction of the fine changes the perception of people regarding the environment in which they operate." Just like the awkward gift of cash to a friend for helping with a move, the fine redefines the exchange from social to economic.
"The presence of an incentive leads us to think differently about a problem," Bowles says, citing brain-scan studies indicating that when people are told about a fine being imposed on antisocial behavior, blood flow relocates from brain regions associated with social reward to those involved in cost-benefit analysis. "Putting a price on violating a social norm reframes an action," he says, turning it into something more akin to shopping.
Everybody's Doing It
Strictly social incentives can also backfire in practice, but in general, says psychologist David Rand, director of Yale University's Human Cooperation Laboratory, "if you can make people's behavior observable to others, it creates social pressure to behave well."
When temperatures soar in the summer, overuse of air conditioners can lead to blackouts. Local governments and utilities can implore residents to turn their systems down or off, but with no one watching them, not nearly enough people respond to the call. Utilities, however, can install a device in homes that enables the company to remotely raise the temperature setting on clients' air conditioners when the power grid is stressed. Studies show that few people ever notice the change. But even when residents are offered a $25 incentive to allow power companies to install the devices, the adoption percentage remains stuck in the single digits.
Rand and three colleagues tried a different approach in apartment houses: Residents received a standard mailing from their power company about the devices, but instead of being asked to opt in online or by mail, they were directed to sign-up sheets posted in their building lobby. With participation now public, enrollment rates tripled. To attract the same participation with a financial incentive alone, Rand's team calculated, the utility would have had to offer $170.
"Trying to maintain a good reputation for yourself is a deeply ingrained part of our social interactions," Rand says, especially when we're potentially being judged by people, like neighbors or coworkers, with whom we know we will have repeated interactions. "That creates a strong social incentive to cooperate."
The chance to be seen by peers as someone who does the right thing can be a powerful incentive. So is being convinced that everyone else in your social circle is already on board with a certain behavior—whether or not it's true. In one series of studies, a team led by Robert Cialdini, a professor emeritus at Arizona State University and a researcher on influence and persuasion, explored the best way to get hotel guests to reuse their towels. The goal was not trivial: It would support the public good by conserving water and energy and benefit the hotel's bottom line by reducing its labor and energy costs.
Hotels typically place a card in rooms encouraging people to reuse towels with a straightforward appeal to preserve the environment, but the cards are largely ineffective. Cialdini believed a focus on descriptive social norms—telling people that everyone else is doing it—would get better results. His team tested three cards: one a straightforward appeal to preserve the environment; one informing guests that the hotel would donate a percentage of its energy savings to environmental causes; and one specifying that 75 percent of guests—that is, people just like you—who were asked to help had reused their towels. The latter card was the most successful, driving cooperation to nearly 50 percent.
But the reality is that, despite the proven success of strategies like these, they have a collectively minor effect on global warming: Incentives to decrease home energy use tend to reduce consumption in communities by no more than 2.5 percent. Blunter, more significant economic incentives, such as a carbon tax or higher gas taxes, may be less popular but, global experts in climate science agree, would almost certainly have a larger influence on behavior by incentivizing people to consume less. At the same time, purely social strategies pose the risk of becoming a crutch for politicians fearful of angering voters with new taxes but eager to show that they are working to nudge citizens in the right direction, if only more would comply.
"We need a hybrid of traditional economic and psychological interventions," Loewenstein says. "To me, that's the sweet spot of public policy—behaviorally informed substantive interventions."
One successful model of economic and social incentives working in tandem was launched in Ireland in 2002, when the government imposed a small tax on the use of plastic shopping bags at grocery stores (today the equivalent of about 33 cents per bag), with all proceeds going to the environmental ministry. The authorities were stunned by their success—an almost complete elimination of plastic grocery bag use in a matter of weeks, as consumers began to carry reusable cloth bags.
Irish shoppers already knew, as do consumers around the world, that plastic bags clog sewer systems and are a hazard to ocean life. But why did this economic incentive take hold there in a way that, at least at the time, it hadn't in other locales? For one, the tax was universal and unavoidable. Also, its purpose was clear and accompanied by an ad campaign promoting the nation's beautiful countryside. But there was more: "Carrying a plastic bag turned into something like wearing a fur coat," Bowles says. "It became a social-norm violation. If you pick up your kids late from day care, there are probably very few other parents to observe you, but your decision about what kind of bags to carry your groceries home in is public because you have to ask for it and other shoppers see you."
However, when grocery chains or municipalities have tried to incentivize the exact same behavior with subsidies—in the form of small rebates for customers who bring their own bags—the programs have generally failed.
"Traditional economics would predict that a five cent tax for using plastic bags is the same as a five cent subsidy for not using them," Loewenstein says, "but it turns out, there's a load of difference between the two. Field research shows that the tax has a huge impact and the subsidy almost none." The difference is caused by "loss illusion"—the idea that we feel more pain when we lose something than we feel pleasure in gaining something.
The Greatest Challenge to Change
Behavioral economists are concerned not only with how institutions can incentivize individuals to do what's right, but also with how we can incentivize ourselves. There may be no greater challenge. "We're quite bad at changing our own behavior, and we invest a huge amount of effort in it," Gneezy says. Imagine a 55-year-old man with a sedentary lifestyle whose doctor has told him that a half hour of walking per day could save his life. "The incentives for him are huge," he says, "yet he still doesn't change."
In general, it's not so difficult to incentivize people to complete a one-time task: Take this survey and get a Starbucks gift card. Incentives that seek to establish long-term lifestyle habits are much trickier to craft.
"We know we ought to exercise and study hard and save for retirement, but in the heat of the moment, we rarely want to do those things," says Katherine Milkman of the University of Pennsylvania's Wharton School of Business. "We are much more interested in binge-watching TV, drinking beer with friends, or splurging on new appliances." This is impulsivity, or the "present bias," and it's a well-studied tension of which most of us are consciously aware. Some of us try to reward ourselves for going to the gym, eating right, or meeting a deadline, in an effort to enforce some self-control. "It's a natural response," Milkman says, and it creates an opportunity for researchers like her.
For a 2013 study, which was the first of its kind, Milkman and two colleagues deployed the concept of "temptation bundling," or getting people to simultaneously tackle a "want" and a "should." Specifically, they theorized that people with a desire for self-improvement, but limited will, would agree to push their future selves to meet their goals. The team recruited participants on college campuses who wanted to go to the gym more often (the "should") and gave each an iPod with his or her choice of popular audiobooks like The Hunger Games (the "want"). The catch: They could access the audiobook only at the gym—if they didn't work out, they wouldn't find out what happened next. The audiobook incentive quickly drove people to go to the gym 50 percent more frequently than participants who did not rely on cliffhangers.
Unfortunately, the program ran into a roadblock after seven weeks: Thanksgiving. Most participants returned from their holiday break with their new habit substantially broken. The study, however, showed that temptation bundling may be a promising way to incentivize healthy behavior. "In contexts where we're self-motivated, if you can align incentives with what people already want, you add extrinsic motivation to intrinsic motivation," Milkman says, "and there's nothing you can do to kill my intrinsic motivation to not be fat."
Cobras, Rats—and Kids
Growing up, Kevin Zollman's family lawn was overrun with dandelions, so his parents decided to pay him for pulling the weeds out. He knew the right way to do the job. "But I was old enough to figure out that if I pulled them up by the roots, there would be fewer dandelions for me the next week," says Zollman, now a professor of philosophy and a game-theory researcher at Carnegie Mellon. "So I engaged in the sustainable harvesting of dandelions. I would just pluck off the flowers and turn them in for money so that they'd be there again the next week."
His strategy was an example of what the late German economist Horst Seibert called "the cobra effect," based on an episode from colonial India in which British authorities, concerned about the number of cobras in Delhi, offered residents a bounty for cobra skins. Unfortunately, the cash incentive was large enough that it became profitable for some people to begin farming snakes for the bounty. When the British realized what was happening, they eliminated the payments. But with no further incentive to raise and skin cobras, the farmers let them loose, and the city's problem was worse than before. An almost identical perverse incentive was launched by the French colonial governors of Hanoi in 1902, when a bounty on rat tails to eliminate a pest problem led residents there to raise and harvest the animals for profit.
Within families, it is disturbingly easy for incentives to backfire, says Zollman, a co-author of The Game Theorist's Guide to Parenting. Parents who give their child an allowance but also consistently supply extra cash for treats when the initial funds run out have not only failed to incentivize their son or daughter to be frugal, they've also done the opposite, incentivizing them to spend the entire sum. And a mother or father seeking to disincentivize an especially unwelcome behavior by making a grandiose threat fails to have an impact once a child figures out the parent has no real interest in canceling the family trip to DisneyWorld or staying home all weekend to make sure the kids don't use any screens—and thus the unwelcome behavior may be reinforced.
"If you don't think through a set of incentives," Zollman says, "kids will figure out a way to manipulate them, and you'll end up worse off."
When it comes to academic performance, parents face the same dilemma that employers do: How to tell when children are doing their best, and how to properly incentivize them if they're not. "This is one of the most profound challenges in economics," Zollman says. "The bad news is that you may never solve it." Parents who think a child is not achieving the grades he or she should might offer $20 for each A on the next report card. But if their child is sincerely struggling with the material, or facing other challenges, that incentive will seem unattainable—and as likely as not lead the child to stop trying altogether. Successful employers and parents instead land on a system of progressive but attainable incentives that keep targets motivated without discouraging them, Zollman says.
But parents should also never forget children's innate resistance to doing what they're told. "I once offered to pay my son for better grades, and it was entirely unsuccessful," Loewenstein says. "Later on, when he decided he wanted to do it, he got the good grades without any incentive." Rather than taking credit for putting the idea to improve in his son's head, though, he surmises that the child's intrinsic motivation to achieve was, if anything, just delayed by his father's incentive program.
Submit your response to this story to email@example.com if you would like us to consider your letter for publication, please include your name, city, and state. Letters may be edited for length and clarity.
Pick up a copy of Psychology Today on newsstands now or subscribe to read the the rest of the latest issue.
Facebook image: Iakov Filimonov/Shutterstock
LinkedIn image: fizkes/Shutterstock