Persons affected by conditions as diverse as Attention Deficit Hyperactivity Disorder, compulsive gambling behavior, nicotine addiction, and alcoholism share a common trait: abnormally high impulsivity. In particular, they tend to prefer immediate to delayed rewards. This decrease of subjective value over time is referred to as temporal discounting. Steep temporal discounting has been shown to be associated with negative outcomes, and interventions that reduce discounting have shown positive welfare effects. Surprisingly, however, very little evidence exists of how temporal preferences arise and which environmental forces shape them. This proposal tests a simple hypothesis of how short-sighted temporal preferences could arise endogenously. Specifically, we hypothesize that poverty leads to stress, stress leads to short-sighted temporal preferences, and short-sighted preferences in turn reinforce poverty. Thus, our proposal seeks to identify a potential micro-mechanism for behavioral change. To establish the generality of this mechanism, we will conduct experiments in both developing and developed countries, both in the laboratory and in the field;
we aim to establish causality in all studies, and employ novel measures of welfare (cortisol and cytokines). First, in a laboratory experiment, we ask whether stress has a causal effect on temporal discounting;we induce stress in participants and subsequently measure its effect on temporal discounting. This experiment will reveal whether stress leads to short-sighted decision-making. In the second study, we ask whether poverty causes stress and short-sighted choices. We develop a novel laboratory paradigm that mimics some features of poverty, and then ask a) whether this manipulation increases levels of the stress markers cortisol and alpha-amylase, and b) whether "poorer" people show higher temporal discounting than others. This result would suggest that poverty may cause stress and short-sighted choices. The third study takes place in the field, and in a developing country, and asks for the causal effect of a poverty increase on behavior. We take advantage of a unique natural experiment among the Maasai of Southern Kenya: using rainfall data as an instrumental variable, we can identify the causal effect of an exogenous increase in poverty on stress and temporal discounting. Finally, the fourth study tests the converse question, namely the effects of an exogenous decrease in poverty decreases stress and leads to more far-sighted decision-making. This study will consist of a randomized controlled trial in Kenya, in which randomly selected poor households receive an unconditional cash transfer, free health insurance, or no intervention. Through these exogenous interventions, we can identify the causal effect of poverty alleviation on behavior. Together, these studies will answer whether the candidate mechanism for short-sighted decision-making outlined above withstands causal scrutiny, and thus identify a potential micro-level mechanism for behavioral change. (29 lines)

Public Health Relevance

Poor health-related behaviors are strongly associated with short-sighted decision-making. Short-sighted decision-making, in turn, is associated with poverty and stress, but it is unclear whether these variables are causally related. This proposal tests a potential mechanism through which short-sighted decisions might come about: if poverty increases stress, and stress leads to short-sighted choices, this mechanism might provide a starting point for interventions that improve health-related behavior through reducing short-sighted decision-making.

National Institute of Health (NIH)
National Institute on Aging (NIA)
Research Project (R01)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-BBBP-R (50))
Program Officer
King, Jonathan W
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Zurich
Zip Code
Haushofer, Johannes; Fehr, Ernst (2014) On the psychology of poverty. Science 344:862-7
Haushofer, Johannes; Cornelisse, Sandra; Seinstra, Maayke et al. (2013) No effects of psychosocial stress on intertemporal choice. PLoS One 8:e78597