Temperature, Worker Productivity, and Adaptation: Evidence from Survey Data Production

Abstract

This paper estimates the impact of daily weather on worker productivity by using household survey data to study interviewers. Using data from over 9,000 Demographic and Health Survey interviewers in 46 countries, I find that interviewers complete 14 percent fewer interviews per hour on the hottest and most humid days. Workers maintain the same total output by starting earlier in the day and spending more time on each interview at the expense of spending more hours in the field with the same total pay. In addition, interviewers become differentially less productive on tasks that are less easily monitored. These findings are important not only because temperatures are predicted to rise significantly over the next century, but also because they shed light on how adaptation by workers to adverse conditions can be costly in ways not captured by aggregate productivity measures such as GDP.