Association for Public Policy Analysis & Management

This paper attempts to replicate the controversial findings of Hamermesh et al. (2017) who use the American Time Use Survey to show that there are racial differences in work effort that could bias measurement of wage inequality. Our paper examines the effects of non-response rates on estimates of hours worked and the share of the work day spend not working, a proxy for laziness. By using the American Time Use Survey 2003-15, this paper seeks to identify the rate at which workers spend time at work performing non-work related activities. We identify population level biases in the survey, discuss ways that these biases may be wrongly attributed to race or ethnicity, and connect ways these attributions may prevent workers from earning wages based on their effort. The paper addresses core questions posed in measuring racial disparities in work effort.