Date of Graduation

Spring 5-19-2023

Document Type


Degree Name

Master of Science in International and Development Economics (MSIDEC)


College of Arts and Sciences



First Advisor

Bruce Wydick


The goal of this paper is to examine how increased temperatures affect human behavior in terms of cooperation. Currently, there is an abundance of literature regarding increased aggression in humans as temperatures rise. However, there is ambiguity concerning whether a similar effect exists in altering levels of cooperation in human interactions. Additionally, people of different characteristics may be influenced by heat at varying levels. First, we run a baseline model of prosociality on different heat measurements, such as standard and wet-bulb temperatures. Included in this model are experimental specification variables, a “trigger”, and characteristics. The trigger is defined as losing a competition under heat stress before prosociality is measured to see if this alters the effect of heat on prosociality. Next, In order to find out how heat alters cooperation among individuals of varying backgrounds, we apply three methods of heterogeneous treatment effect estimation: OLS estimation, causal forests, and “generic” machine learning with LASSO. These three methods are employed for comparison purposes and allow us to examine possible heterogeneity in different ways. Our results show the heat treatment to have overall slight heterogeneous treatment effects on human cooperation across OLS and Causal Forests; specifically, we find heterogeneity in treatment responsiveness when examining the location of Davis, gender, and in stress response measurements derived from various Likert scale survey questions. This heterogeneity is present in those with “Tend” or “Befriend” responses to stress.