Mainstream economics is famous for its assumption that people are rational maximizers, acting in their own interest in predictable ways. And indeed for some purposes, this assumption is helpfully predictive—for instance, in estimating how much Americans would cut back on driving if Congress were to add a $1 carbon tax to a gallon of gas. But in other, more complicated situations it presents a misleading and inaccurate model of human behavior. In recent years the subdiscipline of behavioral economics has begun addressing this shortcoming. Small-scale experiments are conducted just about anywhere—in college cafeterias, suburban malls, or farm villages in India and Africa—carefully tracking the actions of participants in situations both hypothetical and real.
Last year’s Scarcity: Why Having So Little Means So Much (Times Books), co-authored by Harvard economist Sendhil Mullainathan and Princeton psychologist Eldar Shafir, summarizes a large body of findings assessing how the experience of scarcity reduces cognitive effectiveness. In the process the book lays out strong—and surprising—similarities between the poor, who lack income, and the busy, who lack time. Applying psychological insights to economic life, the authors illuminate the difficulties facing both groups.
The first step toward grasping this book’s implications is to recognize how powerfully our frame of reference affects our decisions. Experiments demonstrate that while many people will drive an extra half-hour roundtrip to save $40 on the purchase of a $100 DVD player, few will take the same detour to save $40 on a $1,000 computer. If we were perfectly rational, we would recognize that we’re saving the same amount of money in either case; but we are affected by the fact that $40 doesn’t seem like much when we’re making a big purchase. Would you negotiate to save a final $40 while buying a car? Probably not—though a $40 difference in your bill at the grocery store would be a big deal. The frame of reference, in other words, has powerful effects. In the context of scarcity, every last dollar looms large.
Mullainathan and Shafir show how scarcity captures the mind and leads us to “tunnel”—to focus obsessively on conserving what we’re short of. In one experiment, participants were invited to play an invented video game, Angry Blueberries. Modeled on the smart-phone app Angry Birds, the game rewarded players for using digital blueberries to hit waffles flying overhead. Some players were given six blueberries per round and others three. Those with six scored more points, of course, but those with only three took more time before shooting and were more accurate. Scarcity led them to “tunnel,” focusing on saving blueberries in order to compensate for the shortage. Similarly, we all know the experience of becoming more productive as a deadline approaches. Scarcity focuses the mind.
But while tunneling can help us address whatever problem we’re focusing on, the downside is that we’re less able to attend to all the other things that may be going on while we focus on the most pressing. This “tunneling tax” seems to be the best explanation for why a leading cause of death among firefighters is traffic accidents. The authors recount the death of a firefighter who forgot to buckle his seatbelt, causing him to fall out of the fire truck when the door opened during a turn. Firefighters are trained to be careful, but an intense focus on getting to the fire leaves less psychic attention left for safety en route. Similarly, poor farmers in India generally don’t buy crop insurance for the next season —even when it’s 90 percent subsidized —because they are so focused on paying the bills today.
The authors use the notion of “bandwidth” to capture this phenomenon. We’ve all experienced the frustration of our computer slowing down when it is trying to do lots of things simultaneously. So too with each of us. Technically, bandwidth in humans entails both cognitive capacity (solving problems, retaining information, reasoning logically) and executive control (planning, paying attention, impulse control). Operationally it determines how well (or poorly) we process our tasks. Scarcity, whether of time or money, taxes our bandwidth. You may have a pressing project at work but must attend an unrelated meeting—where you end up being not much help, because your attention keeps wandering back to the project. You’re simply not thinking as well as usual.
This effect has surprising consequences. In one experiment, the authors had subjects engage in a two-stage process. Participants were given a hypothetical problem: your car needs $150 of repairs, and you don’t know how long it can go before breaking down. They were asked to discuss how they felt about this and how they’d decide what to do; then they took a test measuring their cognitive capacity. Participants had provided personal information, allowing researchers to divide them into upper and lower income groups. It turned out that the “rich” and the “poor” groups performed equally well on the cognitive test. But when the scenario was altered to a $1,500 repair bill, the poor group did substantially worse, scoring an equivalent of thirteen to fourteen IQ points lower than the rich group.
The conclusion is that the stress of scarcity reduced their ability. To understand this finding, it’s important to keep in mind that—according to a 2011 study—fully half of all Americans say they could not come up with $2,000 in thirty days even if they needed it. For half the people in this country, in other words, a $1,500 bill for car repair is a calamity. The bandwidth tax generated by such scarcity lowers cognitive capacity. Basically, if you have to think nonstop about money, you can’t think about much of anything else—at least, not very well.
Eager to know whether this principle holds true in a very different culture, Mullainathan and Shafir turned to rural India. Many Indian farmers are paid just once a year, at harvest time. Thus they tend to be “rich” for a few months, and then very poor in the months preceding the next harvest. Testing farmers in these “rich” and “poor” periods yielded the same result: poverty reduces cognitive capacity. Thus what might look like an irrational refusal to buy very cheap crop insurance may be attributable to the way capability drops when the mind is captured by scarcity. Similarly, a goodly number of corporate mistakes in the United States are caused when a very busy manager, pressured by time and facing an issue of secondary concern, makes an errant decision.
An important insight here concerns “slack,” or the room to maneuver that gives you a sense of abundance—as when, for instance, we switch from a small suitcase to a larger one because we want to pack something that won’t fit. This room to maneuver—this cushion—makes for a fundamental difference between the poor and the relatively affluent. Almost every time I shop for groceries, I pass an individual or a couple standing in the canned-goods aisle intently comparing the coupons they’re holding with the prices displayed on the shelf. They’re counting pennies. While I don’t buy lobster at the grocery, I don’t have to be so careful—and can even splurge sometimes. I have slack, and they don’t.
Slack highlights the interesting comparison between time and money. Most of us who are very busy—who experience a scarcity of time—nonetheless typically have something we’re working on that we can put off. We have a lot of balls in the air, and that can be daunting. But we also know that if we let one hit the floor, the damage likely won’t be serious. For the poor, the stakes are much higher and the downside of failure more serious. The utility bill is unpaid for two months and the current bill threatens to cut off the electricity. The car needs repairs, the preschooler needs new shoes, and the rent is past due. There is no slack for the poor, and this makes their scarcity so much more debilitating than ours. It takes its toll on cognitive capacity—which in turn leads to bad decisions, decisions which often exacerbate poverty. Poverty is relentless.
ONE LESSON TO TAKE FROM the work of Mullainathan and Shafir is that we ought to be more sympathetic toward the mistakes made by the poor—whether the poor across town or on the other side of the planet. If we recognized that in their situation we ourselves would be less cognitively capable than we are in our affluent lives—less able to cope with daily challenges—we might realize how big the payoff can be for, say, subsidized daycare for the poor. Relieving time-scarcity improves other decisions by freeing up bandwidth.
The authors point to other policy implications. They argue, for example, that the current sixty-month lifetime limit on welfare support under the government’s TANF program is founded on a fundamental misunderstanding of poverty. Given the tunneling effect of scarcity, the poor don’t much notice long-term deadlines. Far better would be a rule limiting support to X months every Y years. How big X and Y are will depend on the generosity of the law, but nearer and more frequent deadlines seem likely to work better.
As for the Indian farmers, providing direct aid may be less important than setting up a financial instrument that would transform once-a-year payments into twelve monthly ones, effectively reducing the bandwidth tax in the months prior to harvest. And since the unpaid portion over the year can be invested to pay a return, each payment could be slightly larger than one-twelfth of the original value of the harvest and still generate enough to pay the cost of the financial services required. Currently the poor lack access to trustworthy financial institutions and won’t risk signing on to such a scheme. But international aid organizations could provide such services—or, even better, could promote them locally and guarantee their fiscal credibility until they gain the farmers’ confidence.
Mainstream economics has been the target of many criticisms in recent years, some of them well founded. The rise of behavioral economics, as exemplified by Mullainathan and Shafir in their fascinating and provocative book, holds the promise of an economic science that is both more realistic and more helpful.
