The Why Axis: Hidden Motives and the Undiscovered Economics of Everyday Life

by Gneezy, Uri

  Other people and organizations may also care about your health—often because it can save them money. Consider the incentives some employers and insurers use to try to encourage employees to exercise. Say you’re called in to get yourself weighed and measured. You’re also asked if you smoke. If you’re deemed to be the right weight, you don’t smoke, and you have normal cholesterol and blood pressure, your company reduces or refunds the co-pays and deductibles you’re paying in your health care premium, saving you $750 a year. Not bad, right?

  This is exactly what Safeway supermarkets tried to do with its Healthy Measures program, which it rolled out with great fanfare to its nonunion workforce (mostly people who work in offices). “By our calculation, if the nation had adopted our approach in 2005, the nation’s direct health care bill would be $550 billion less than it is today,” CEO Steven Burd boasted in a 2009 editorial in the Wall Street Journal.7 Burd claimed healthcare costs for his company stayed flat.

  Following the publication of his article, Burd became a celebrity. Companies and insurers began exploring similar programs. In Washington, D.C., Safeway became a poster child for healthcare reform. President Obama talked of Safeway’s cutting its healthcare spending by 13 percent. The House and Senate worked on a so-called Safeway Amendment, which could save families with average health benefits thousands a year.

  One needs to be careful when extrapolating Safeway’s claims to possible savings for the nation. First, the statistics Mr. Burd reported were problematic.8 It is always hard to conclude what works and what doesn’t when the people providing the data have an interest in the conclusion. Additionally, the Safeway gambit was not a controlled experiment. For example, we don’t know what fraction of the change was due to healthy people deciding to stay on the payroll or to join Safeway as a result of this policy. The less fit people may have simply chosen to work for a different company. Either way, Safeway saves money—which is great—but from a global perspective, the problem just ends up being rolled somewhere else.

  None of this means Safeway’s incentive scheme was bad. But from a practical perspective, designing incentives that truly change behavior is challenging. In recent years, we have been involved in a large project with a major health insurance company that tries to use incentives to help its members. Theirs is a win-win situation: the members get in better shape, and the company saves money. The problem is that the incentives come on top of strong, already existing motivations. Think about the amounts of money and effort that people spend on dieting in frustrated efforts to lose weight. They are already motivated to shed pounds. Could paying them a little help get them to change their exercise habits?

  The trick in using money, of course, is to entice people to change their habits. Here’s an example of an incentive scheme that we designed and tested.9 We wanted to use the simplest incentive possible, so we invited students to our lab, and then randomly divided them into two groups. One group was simply used as a control; we “bribed” the other by offering to pay each participant $100 to go to the gym eight times during a month. Following the principle of “paying enough,” there are very few things you can’t get students to do for the right amount of money. Not surprisingly, the participants came to the lab and went to the gym as required.

  But we weren’t after temporary compliance. The important question was whether the incentive led to habit formation: what would happen once the month was over and we stopped paying the bribe? Would the incentive backfire, as happened in the day-care study? Would it make any kind of difference at all? Or would the students form some kind of habit of visiting the gym—enough so that they might continue to visit the gym after we stopped paying them?

  The results were encouraging. We found that attendance, even after we stopped paying, doubled for the group we’d paid to visit the gym eight times. The incentive seemed to help these people “get over the hump” of exercising regularly. Those who said they didn’t exercise because they didn’t have time “found” the time after we “forced” them (with incentives) to find it, and they continued to find the time afterwards. Others may have noticed that they just felt a lot better. Still others might have looked forward to seeing new friends. Whatever the reasons, the important thing was that they managed to change their habits, and they were rewarded with better health as a result.

  What can we learn from this study? Many of us want to exercise more than we do. The experiment told us that the hardest part of doing this is not the sweating, panting, and changing clothes; it’s adapting to the routine. Adaptation, indeed, is what it’s all about. Think about this for a moment. There are certain routines you probably can’t see yourself living without—your morning cup of coffee, brushing your teeth at night, and so on. So if you give yourself enough time to get over the hump and adapt to a new routine of exercising, it will become a habit.

  Start by committing to going to the gym a couple of times a week for a month. Even if you initially feel the cost of exercising will be higher than the benefit, you’ll find that after just those four weeks, you’ll have gotten used to the effects of exercise. You’ll notice your pumping heart, the psychological lift, the feeling of accomplishment. After this month of exercising you’ll find that the effort of heading to the gym is much less difficult than it was during the first week or two of your personal experiment. In fact, you’ll have become so used to the way you feel afterwards that you’ll start to miss that feeling if you skip a gym visit. At this point, you’ll decide that either the cost of going to the gym is lower, the benefit is higher, or both—such that exercising will become a net positive.

  Now, it’s too simplistic to think we can throw money or other positive incentives at people over time and expect them to do what we hope they will do. Changing deeply embedded habits is difficult for most people. After all, some people keep smoking or eating the wrong things even when they are faced with the prospect of death.

  As you can see, making assumptions about how people react to incentives is pretty risky. We assume that people respond in predictable, knee-jerk ways to incentives like money, but they don’t. Sometimes incentives work in the short term but not in the long run. And sometimes they make people behave in the opposite way you would expect them to. Higher incentives don’t necessarily lead to better performance.

  Here is the truth: if you want people to do something, you really need to understand what motivates them. That is the key: once you understand what people value, then you can use incentives to work in predictable ways, and you can get people (including yourself) to behave in ways that you want them to.

  As economists, it’s our job to look under the hood. We have to learn what can happen in different scenarios. And we must try to understand, as best we can, which incentives work, which ones don’t, and why, so that individuals, businesses, and governments can achieve their goals.

  In the next two chapters, we’ll look into how deeply held cultural worldviews might play into the age-old question of why women continue to earn less than men.

  CHAPTER TWO

  What Can Craigslist, Mazes, and a Ball and Bucket Teach Us About Why Women Earn Less Than Men?

  On the Plains Below Kilimanjaro

  In January 2005, Larry Summers, then president of Harvard University, offered a lunchtime address to participants at the Conference on Diversifying the Science & Engineering Workforce. Introducing his talk as an “attempt at provocation,” he proceeded to lob a heavy grenade into the ancient war of the sexes. Specifically, he wondered aloud whether an innate, gender-related difference in aptitude between men and women was the culprit behind the huge gender disparity observed among hard-core scientists.

  Citing research showing that women make up just 20 percent of US professors in science and engineering, Summers questioned whether, “in the special case of science and engineering, there are issues of intrinsic aptitude, and particularly of the variability of aptitude, and that those considerations are reinforced by what are in fact lesser factors involving socializat
ion and continuing discrimination.” In other words, he wondered whether women might be at an inherent intellectual disadvantage when it comes to getting to the top in the hard sciences.1

  The reaction against Summers’s comment was swift, huge, and harsh. A top biologist from MIT, Nancy Hopkins, exited the room in a huff. “For him to say that ‘aptitude’ is the second most important reason that women don’t get to the top when he leads an institution that is 50 percent women students—that’s profoundly disturbing to me,” Hopkins told reporters. “He shouldn’t admit women to Harvard if he’s going to announce when they come that, hey, we don’t feel that you can make it to the top.”2 The local and national media went wild, and a campaign quickly ensued to fire Summers. The following year, he resigned his post at Harvard—partly because of the reactions to his comments at the conference.

  Summers’s comments—viewed as sexist at worst, tone-deaf at best, and utterly politically incorrect (and he apologized for them several times)—at least did fit with eons of tradition. For millennia, culture and science have colluded to explain why women aren’t as competitive and ambitious as men. In the book of Genesis, Adam’s role was to be Eve’s master. In ancient Rome, women were citizens, but they could not vote or hold public office. Many religions, laws, and cultures around the world persist in subjugating women and forbidding them from competing in a “man’s world.”

  Summers’s comments also bore the stamp of Charles Darwin, who more than 150 years ago proposed that successful males evolved to win the mating race. Since then, Darwin’s theory of natural selection has helped to explain why males are generally more aggressive and violent than females. After all, men had to go out and compete with men from other tribes to kill animals, while women raised and nurtured the young. The main idea is that the females’ cost of producing and raising offspring (such as pregnancy, birth, and nursing) is much higher than for males. Hence, males should compete to have as many offspring as they can, while females need to be choosy in their selection of the right male.

  If, as Darwin suggested, evolution is responsible for a comparative lack of competitiveness in females (Darwin didn’t write only about humans), a few hundred years of cultural changes would not make a difference. Evolution could help explain why the number of women in high-profile jobs still pales in comparison to that of their male counterparts, or why US women still only earn, on average, 80 cents for every dollar earned by a man.

  After citing the research and mentioning his “innate differences” hypothesis, Summers explicitly told his audience: “I’d like to be proven wrong on this one.”

  In this chapter and the one following, we will take up this challenge. In particular, we will examine what part of the gender gap in labor markets is due to culture. We couldn’t just take for granted, in the absence of data, that women were innately less competitive than men. We decided to start collecting evidence by looking at ordinary men and women in their natural habitats and doing things people do every day—say, participating in a gym class or answering job ads on Craigslist—and we used the full gamut of experimental tools at our disposal to answer these questions: To what degree are the differences between men and women (such as levels of aggression, competitive drive, and wage-earning power) truly innate? To what degree are they culturally learned? In the end, we’ve come up with a unique explanation for the persistent differences we observe between men and women, particularly when it comes to competition.

  But first, let’s take a closer look at why women, despite massive advances, still seem to be held back.

  How Much Do Women Compete?

  Our interest in gender roles and competitiveness began with the births of our own children. Soon after they were born, we began to notice differences among the girls, and between the girls and their brothers. While one of Uri’s girls was much more competitive than her sister, all of our girls always preferred dolls to their brothers’ trucks and baseballs. We began to ask ourselves the same question most parents of daughters ask: In a man’s world, what sort of chances would they have? Would they be able to play, and succeed in a culture in which the opportunities are still unequal, despite all the strides women have made?3

  The sad fact is that while women are doing better than men in some areas, such as higher education, there is no reason to celebrate an upending of the millennia-old, male-dominated order of things just yet. In the United States and around the world, men still occupy the highest ranks of society. The proportion of women in the market workforce has risen from 48 percent in 1970 to 64 percent in 2011,4 but only one in five senior management positions are held by women, and fewer than 4 percent of CEO positions in Fortune 500 companies. Some consider these facts as an achievement, since they are the highest in US history. Yet, women are still paid less than men for equivalent jobs. Even in public positions women still have not achieved parity. In Congress, for example, they still hold fewer than 17 percent of seats.

  Scholars have theorized for decades about the reasons why women can’t seem to make faster progress in breaking through the glass ceiling. Personally, we think that much of it boils down to this: men and women have different preferences for competitiveness, and they respond differently to incentives. Our research shows that many women tend to avoid competitive settings and jobs in which salary is determined by relative rankings.

  To illustrate, consider the following large-scale field experiment that we conducted on Craigslist.5 In this experiment, we wanted to directly discover the factors that drive people to apply for entry-level jobs. How would men and women respond to different compensation scenarios? Would women go after jobs that required some competitiveness and risk-taking if the salary offered were higher?

  To get some answers, we placed two ads on Internet job boards in sixteen cities for an administrative assistant, one of the most common jobs in the United States. An example of one of our job postings in Seattle read as follows:

  POSTING CATEGORY: admin/office jobs

  TITLE: Seeking Sports News Assistant

  The Becker Center is seeking a Seattle-area administrative assistant to help gather information on sports stories in the Seattle region. While the Becker Center is based in Chicago, we have a satellite project in Seattle. The assistant will provide us with up-to-date information on local news and views on basketball, football, baseball, soccer, Nascar, golf, tennis, hockey and other sports. Responsibilities for the position include reading local sports-related news coverage (pro, semi-pro, and college) and preparing short reports. The successful candidate will also be comfortable with typical administrative duties—light correspondence, proofreading, filing, email and phone communication, etc.

  COMPENSATION: Hourly

  Our second job description looked almost identical to this one, but it didn’t mention sports. Instead, the description noted, “The assistant will provide us with up-to-date information on community events, arts and culture, business, entertainment, policy issues, crime, and other stories. Responsibilities for the position include seeking out, reading, and summarizing local news stories and preparing short reports.”

  Over a period of four months, nearly 7,000 interested job seekers applied for these jobs listed in the various cities.6 Upon responding to our ad, some were told that they would be paid on an hourly basis, while others were told that their pay would depend on how they performed compared to a coworker.

  Our goal was to see if the competitive aspect influenced one sex more than the other.7 What do you think we found after placing several months of ads on Craigslist? Which sex continued to show interest in our job after being told of the wage structure?

  Not surprisingly, guys were more interested in the sports-oriented ad and women responded to the non-sports-oriented ad: whereas 53.8 percent of the job seekers for the sports-related job were women, 80.5 percent of those responding to the alternate job were female.

  But the real differences showed up when we described the compensation schemes. According to one scheme, the job would pay a
flat $15 per hour. Not too shabby for an entry-level office administration gig. The competitive pay scheme, on the other hand, rewarded the workers based on how they performed in relation to a coworker. Applicants were told that they would be paid $12 per hour but would be compared to another worker, and whichever of the two performed better would receive a bonus of $6 per hour in addition to the $12-per-hour base pay. In this way, each of the two compensation schemes paid an average of $15 per worker, but one was highly incentivized, whereas the other was not.

  You might be surprised (and saddened) by the actual gender breakdown of who applied for each kind of job. In general, women didn’t like the competitive option; in fact, they were 70 percent less likely than men to go after the competitive job. Further, the women who did apply for the highly incentivized job tended to have more impressive resumes than the men who applied for those same jobs. These findings seemed to underscore the fact that, when it comes to competition, men aren’t nearly as shy as women.8

  A successful career as a CEO demands a high level of engagement and responsiveness to competitive situations. No wonder, then, that so few women are at the top. Just Google the phrase “every man has his price,” and you’ll get lots of nice quotes about how every man can be bribed to do almost everything. But if you Google “every woman has her price”—well, that has a very different meaning.