The Psychological Lives of the Poor †

American Economic Review: Papers & Proceedings 2016, 106(5): 435–440 http://dx.doi.org/10.1257/aer.p20161101 The Psychological Lives of the Poor † By Frank Schilbach, Heather Schofield, and Sendhil Mullainathan* There is growing interest in understanding Psychologists often study this underlying the psychology of the poor—biases that may resource by imposing “cognitive load” to tax affect decision-making are of particular interest. bandwidth and measure the impact on judg- The sheer diversity of potential biases—hyper- ments and decisions. The many ways to induce bolic discounting, probabilistic, and judgmental load produce similar results on various band- errors just to name a few—poses a key chal- width measures and consequences from reduced lenge. These psychological biases cannot easily System 2 thinking. This insight is particularly be put into a common unit such as money spent. useful because it implies that bandwidth is both However, two insights from psychology make malleable and measurable. It also suggests a this problem more tractable. unified approach of studying the psychology First, a large body of work points toward a of poverty. We can understand factors in the two-system model of the brain.1 System 1 thinks lives of the poor, such as malnutrition, alco- fast: it is intuitive, automatic, and effortless, and hol consumption, or sleep deprivation, by how as a result, prone to biases and errors. System they affect bandwidth. And we can understand 2 is slow, effortful, deliberate, and costly, but important decisions made by the poor, such as typically produces more unbiased and accurate technology adoption or savings, through the lens results. of how they are affected by bandwidth. Clearly, Second, when mentally taxed, people are bandwidth is not the only important aspect of the less likely to engage their System 2 processes. psychological lives of the poor; no single metric Put simply, one might think of having a men- can take on this role. However, it provides a way tal reserve or capacity for the kind of effortful( to at least partly understand a great many of the thought) required to use System 2. When bur- thought processes that drive decision-making by dened, there is less of this resource available for the poor. use in other judgments and decisions. Though there is no commonly accepted name for this I. Bandwidth capacity, we will refer to it in this article as “bandwidth” Mullainathan and Shafir 2013 . Much like human capital is an abstraction of ( ) a diverse set of skills with common elements, bandwidth comprises a diverse set of psychological constructs with common elements. At an * Schilbach: Department of Economics, Massachusetts intuitive level, bandwidth captures the brain’s Institute of Technology, 50 Memorial Drive, Building ability to perform the basic functions that under- E52 Room 560, Cambridge, MA 02139 e-mail: fschilb@ lie higher-order behavior and decision-making. mit.edu ; Schofield: Perelman School (of Medicine and the Wharton) School, University of Pennsylvania, 423 Underlying this broad construct are two core Guardian Dr., Blockley Hall 1102, Philadelphia, PA 19104 components, measures of which are typically e-mail: [email protected] ; Mullainathan: Department used to capture bandwidth. ( ) of Economics, Harvard, 1805 Cambridge Street, Littauer The first component is cognitive capacity, Center, M-18, Cambridge, MA 02138 e-mail: mullain@fas. the psychological mechanisms that underlie our harvard.edu . We thank our helpful discussants.( † Go to http://dx.doi.org/10.1257/aer.p20161101) to visit ability to solve problems, retain information, the article page for additional materials and author disclo- engage in logical reasoning, and so on. The sure statement s . ( ) second is executive control, which underlies 1 This two-system model has direct relevance to many of the ability to manage our cognitive activities. the heuristics and biases familiar to economists. Kahneman and Frederick 2002 and more recently Kahneman 2011 Executive control oversees planning, attention provide reviews.( Fudenberg) and Levine 2006 develop( a) allocation, initiating and inhibiting actions, and model with two systems in the context of time( discounting.) impulse control. It determines our ability to 435 436 AEA PAPERS AND PROCEEDINGS MAY 2016 focus, to shift attention, to work with informa- by limited time to respond, the correlations tion in our memory, and to self-monitor. between judgments of risks and benefits were These components are rich in nuance, but they significantly more pronounced than when given share the common feature that both are scarce more time to ponder a response. The same affec- resources, the taxing of which causes negative tive evaluation apparently serves as a heuristic spillovers to other aspects of cognitive function- attribute for assessments of both benefits and ing. In this sense, while the detailed distinctions risks when resources are limited. between different brain capacities are central to Economists have applied these ideas to more any psychological investigation, they are less standard economic tasks, such as small-stakes central to those interested in the underlying risk aversion or monetary discounting, typically determinants or downstream consequences of finding an impact of diminished bandwidth these capacities. Deck and Jahedi 2015 . Similar results have One important feature of bandwidth is that it been( found in many other) decisions that rely on can be readily measured, both in the lab and in cognitive capacity and executive control, such field settings. One example you may be familiar as food choice. Shiv and Fedorikhin 1999 with, the Raven’s matrices test, measures indi- is a canonical example in which participants( ) viduals’ capacity to think logically and solve chose between slices of cake and fruit salad problems in novel situations, independent of under varied levels of load, manipulated through acquired knowledge. This task is a nearly uni- digit rehearsal. Those whose minds were busy versally accepted measure of fluid intelligence rehearsing a seven-digit number chose the cake, and a common component of IQ tests Raven the impulsive choice, 50 percent more often than 1936 . The online Appendix describes( three those who were rehearsing a two-digit number. other) such measures of bandwidth and three Not all replications have produced the same common features of these tasks: i ease of results, and the magnitudes of the original effects administration; ii broad applicability;( ) and iii appear likely to be an aberration. However, the ease of instruction.( ) ( ) idea that occupying mental bandwidth dimin- The basic premise of the tests used to study ishes capacity for self-discipline seems to be bandwidth is that it is possible to “load up” cog- more generally supported by the data. nitive resources, and to use this additional load to examine how bandwidth, behaviors, and choices Productivity.—In contrast to the rich body change. These cognitive load studies have been of evidence on the link between bandwidth and conducted for over 70 years and are in the canon decision-making, evidence on the relationship of experimental psychology, reliably replicating between bandwidth and productivity is much in many contexts. As a result, by studying the more limited. There is good reason to believe effects of cognitive load, we have experimental that this link exists: impaired cognitive func- evidence of the impact of diminished bandwidth tion, judgment, and decision-making likely have on a wide variety of aspects of mental function. consequences for one’s performance in the labor market, especially in work that relies heavily Decision-Making.—Prospective memory, or on cognitive capacities such as attention, perse- the ability to remember to execute tasks in verance, or memory. For instance, a rag picker the future, and executive control are particu- trying to find valuable items among mountains larly affected by cognitive load Marsh and of garbage may be particularly affected by Hicks 1998 . For instance, dieters (exhibit less reductions in bandwidth. Although these argu- self-control )in the eating arena and people dis- ments are intuitive, it would be presumptuous count delayed rewards at significantly higher to believe that these effects must exist, and the rates when under load Ward and Mann 2000; magnitude of effects may vary widely with con- Hinson, Jameson, and Whitney( 2003 . text. This is an area of research ripe for further These shifts in underlying cognition) mani- investigation. fest in myriad contexts and for wide-ranging outcomes. For example, Finucane et al. 2000 Utility.—All economists would agree that pov- asked respondents to judge the risks and( bene)- erty lowers utility by decreasing consumption. fits of various products and technologies e.g., However, it may lower utility through an addi- nuclear power . When bandwidth was taxed( tional channel: individuals with low bandwidth ) VOL. 106 NO. 5 THE PSYCHOLOGICAL LIVES OF THE POOR 437 say, due to physical pain may find consump- II. Poverty and Bandwidth tion( of other goods less enjoyable.) That is, the utility from a given basket of goods may be While it may seem odd that a person’s fun- reduced by low bandwidth. There is suggestive damental “capacity” can be easily affected in evidence of such a link. In a study to determine many basic dimensions, that oddity is precisely how best to rehabilitate prisoners of war from the point. We have traditionally viewed cogni- malnourishment, 32 volunteers semi-starved tive capacity as fixed, but in fact it can change themselves for six months and then followed with circumstances. More specifically, we will varied rehabilitation diets Keys et al. 1950 . now discuss how bandwidth can be influenced The changes to the participants’( physiological,) by poverty. Our discussion includes some of the physical, cognitive, and psychological func- factors which have already been shown to influ- tions were closely tracked. Unsurprisingly, ence bandwidth, and others for which evidence interest in food increased as the starvation is limited but suggestive, warranting additional period progressed.

Load more