Scarcity: Why Having Too Little Means So Much (7 page)

The
even-numbered words were the same for both groups. The odd-numbered words were neutral words for the first group but tempting ones for the second:
STREET
became
CAKE
,
PICTURE
became
DONUT
, and so on. We then looked at how long it took participants to find the same words, those they had in common, the even-numbered neutral ones.

For most subjects, changing the odd-numbered words had no effect. Not so for dieters. Dieters took 30 percent longer to find
CLOUD
after they had just searched for
DONUT
. Dieters were not slow overall—they found
CLOUD
just as quickly as nondieters when it was preceded by
PICTURE
.
The
DONUT
was the problem.
What is happening here is clear. It is a version of what psychologists call
proactive interference
. The mention of a donut brings it top of mind. The nondieter searches for it, finds it, and moves on. The dieter, in contrast, finds it hard to move on. Even while searching for the next word, for
CLOUD
, that donut, every bit as disruptive as a passing train, is still there, drawing attention. And it is hard to find
CLOUD
when your mind is elsewhere.

Surely you’ve experienced something similar. If not with food, then perhaps with time. You are against a tight project deadline but must attend an unrelated meeting. How much of this meeting will you process? Sitting at the meeting you try to focus, but despite your best efforts, your mind keeps wandering back to that deadline. Your body is at the meeting, but your mind is elsewhere. Like the word
DONUT
for the dieter, the deadline keeps pulling you away.

Imagine that you are surfing the web on your laptop. On a reasonably fast computer, you easily go from page to page. But imagine now that there are many other programs open in the background. You have some music playing, files downloading, and a bunch of browser windows open. Suddenly, you are crawling, not surfing, the web. These background programs are eating up processor cycles. Your browser is limping along because it has less computing power to work with.

Scarcity does something similar to our mental processor. By constantly loading the mind with other processes, it leaves less “mind”
for the task at hand. This leads us to the central hypothesis of this chapter:
scarcity directly reduces bandwidth
—not a person’s inherent capacity but how much of that capacity is currently available for use.

To test this hypothesis, we need to refine our definition of
bandwidth
. We are using the term as a placeholder for several more nuanced and carefully researched psychological constructs. In effect, we are walking a fine line. As psychologists, we care about the distinctions, functional and otherwise, between the various constructs and their corresponding brain function. And
bandwidth
is a generic term that obscures those distinctions. But as social scientists interested in the effects of scarcity, we are willing to leave the fine distinctions alone, much as one might refer to
democracy
or
subatomic particles
while avoiding the many finer distinctions that these afford. By way of compromise, we will continue to use the blanket term
bandwidth
to refer to two broad and related components of mental function, which we will now explain in greater depth.

The first might be broadly referred to as
cognitive capacity
, the psychological mechanisms that underlie our ability to solve problems, retain information, engage in logical reasoning, and so on. Perhaps the most prominent in this category is fluid intelligence, the ability to think and reason abstractly and solve problems independent of any specific learning or experience. The second is
executive control,
which underlies our ability to manage our cognitive activities, including planning, attention, initiating and inhibiting actions, and controlling impulses.
Much like a central processor
, executive control is essential to our ability to function well. It determines our ability to focus, to shift attention, to retain things in memory, to multitask, to self-monitor. Cognitive capacity and executive control are multifaceted and rich in nuance. And scarcity affects both.

A central feature of cognitive capacity is fluid intelligence. To test for the impact of scarcity on people’s cognitive capacity, we use the most
prominent and universally accepted measure of fluid intelligence, the
Raven’s Progressive Matrices test
, named after the British psychologist John Raven, who developed the test in the 1930s. For an example, look at the following, which is similar to a typical Raven’s test item, and ask yourself which of options 1–8 fits in the missing space:

You may recognize this test from your school days.
It is a common component of IQ tests
. While IQ tests are complex and variegated, most agree that the Raven’s Progressive Matrices test is one of the most important and reliable components. Raven’s requires no knowledge of world events and little formal study. It is the most common way that psychologists, educators, the military, and others measure what is called fluid intelligence, the capacity to think logically, analyze and solve novel problems, independent of background knowledge. A mechanic reasoning about why an engine won’t start uses both background automotive knowledge and reasoning skills. The same mechanic looking at a Raven’s Matrix is applying his reasoning skills in a context in which he has no expertise—he’s on par
with a farmer in India. This has made Raven’s particularly useful as a measure of general intelligence, one that supposedly transcends specific culture. Still, there are skeptics.
Those who have familiarity with tests and test taking
will surely perform better. Those who have taken geometry might do better. In fact, it is known that there are benefits to schooling—children with more years of school do better than those of equal age with fewer years. The debates about what IQ really measures persist even for fluid intelligence. Fortunately, these debates do not matter for our purposes. We will not be comparing fluid intelligence between one person and another or from one culture to the next. We are interested in how scarcity affects the same person’s cognitive capacity. It may strike you as odd that a person’s “capacity” can be so easily affected, but that is precisely the point—we are used to thinking of cognitive capacity as fixed, when in fact it might change with circumstances.

To see the effect of scarcity on fluid intelligence, we ran some studies with our graduate student, Jiaying Zhao, in which we gave
people in a New Jersey mall
the Raven’s Progressive Matrices test. First, half the subjects were presented with simple hypothetical scenarios, such as this one:

Imagine that your car has some trouble, which requires a $300 service. Your auto insurance will cover half the cost. You need to decide whether to go ahead and get the car fixed, or take a chance and hope that it lasts for a while longer. How would you go about making such a decision? Financially, would it be an easy or a difficult decision for you to make?

We then followed this question with a series of Raven’s Matrices problems. Using self-reported household income, we divided subjects, by median split, into rich and poor. In this setup we found no statistically significant difference between the rich and poor mallgoers. Of course, there may have been some difference, but it was not big enough for us to detect in this sample. The rich and the poor looked equally smart.

For the remaining subjects, we ran the same study but with a
slight twist. They were given this question instead (with the change shown in bold):

Imagine that your car has some trouble, which requires an
expensive $3,000 service
. Your auto insurance will cover half the cost. You need to decide whether to go ahead and get the car fixed, or take a chance and hope that it lasts for a while longer. How would you go about making such a decision? Financially, would it be an easy or a difficult decision for you to make?

All we have done here is replace the $300 with $3,000. Remarkably, this change affected the two groups differently. Coming up with half of $300 or $3,000 was easy for those who were well off. They could just pay out of savings or put it on a credit card. For the less well off, finding $150 for an important need was not too hard either. Not enough to make them think too much about scarcity and their own finances.

Not so for the $3,000 car expense: finding $1,500 was going to be hard for those with low incomes. A 2011 study found that close to half of all Americans reported that they would be
unable to come up with $2,000 in thirty days
even if they really needed it. Of course the question we gave the mall respondents was hypothetical. But it was realistic, and it likely got them thinking about their own money concerns. They may not have a broken car, but experiencing money scarcity would mean they had monetary issues close to top of mind. Once we tickled that part of the brain, the all-too-real nonhypothetical thinking about scarcity would come spilling out.
Coming up with $1,500 would be hard. My credit cards are maxed out. Already the minimum payment due is so large. How will I make the minimum payment this month? Can I afford to miss another payment? Should I take a payday loan this time instead?
A little tickle could raise a racket in the brain.

And this racket affected performance. The well-off subjects, with no racket, did just as well here as if they had seen the easy scenario.
The poorer subjects, on the other hand, did significantly worse. A small tickle of scarcity and all of a sudden they looked significantly less intelligent. Preoccupied by scarcity, they had lower fluid intelligence scores.

We have run these studies numerous times, always with the same results. This is not merely an artifact of the $3,000 being
mathematically
more challenging. When we ran nonfinancial problems, we found absolutely no effect of giving similarly small versus large numbers. The effect is specific to hard problems that are financial in nature (for those who are short on money). It is also not the result of a lack of motivation. In one replication of the study, we paid people for every correct answer on the Raven’s test. Presumably the lowincome participants have a
greater
incentive to do better: after all, the money matters to them more. But they did not do any better; in fact, they did just a tiny bit worse than before. Low-income participants who presumably could have used the extra pay left the mall with less money after having contemplated the harder scenarios, an effect that was absent for those financially more comfortable.

In all the replications,
the effects were equally big
. To understand how big these effects are, here is
a benchmark from a study on sleep
. In this study, one group of subjects was put in bed at a normal time. Another group was forced to stay awake all night. Pulling an allnighter like this is terribly debilitating. Imagine yourself after one night without any sleep. The next morning, the sleeping group was awakened, and both groups were given a Raven’s test. Not surprisingly, the sleep deprived did much worse.

In comparison, how big was our effect at the mall? It was even bigger. How smart do you feel after a night of no sleep? How sharp would you be the next morning? Our study revealed that simply raising monetary concerns for the poor erodes cognitive performance even more than being seriously sleep deprived.

There is another way to understand the size of our findings. Because the Raven’s test is used to measure fluid intelligence, it has a direct analogue with IQ. Typical studies of IQ assume a normal distribution of IQ scores, with a mean of 100 and a standard deviation
of 15. (Standard deviation is a measure of the dispersion of scores around their mean. In a normal distribution, almost 70 percent of scores fall within one standard deviation of the mean.) One can calibrate the impact of an intervention by looking at how its effect compares to the standard deviation. For example, if an intervention has an effect equivalent to one-third of a standard deviation, then that effect corresponds to
about five IQ points
.

By that measure our effects correspond to between 13 and 14 IQ points. By most commonly used descriptive classifications of IQ, 13 points can move you from the category of “average” to one labeled “superior” intelligence. Or, if you move in the other direction, losing 13 points can take you from “average” to a category labeled “borderline deficient.” Remember: these differences are not between poor people and rich people. Rather, we are comparing how the same person performs under different circumstances. The same person has fewer IQ points when she is preoccupied by scarcity than when she is not. This is key to our story. The poor responded just like the rich when the car cost little to fix, when scarcity had not been rendered salient. Clearly, this is not about inherent cognitive capacity. Just like the processor that is slowed down by too many applications, the poor here
appear
worse because some of their bandwidth is being used elsewhere.