Conference Paper Making Linked Employer-Employee Data Policy-Relevant Workshop
Workers, Workplaces and Working Hours
16 Sep 2005
Under plausible assumptions about firm preferences over the working time of their employees, the number of hours worked is likely to depend on firm, as well as individual, characteristics. I use the WERS98 employer-employee matched dataset to analyse the role played by differences between firms on the one hand, and between individuals on the other, in the observed variation in hours of work. I analyse whether and how hours vary within firms according to individual characteristics, and evaluate the degree to which individuals are sorted into different firms based on their human capital characteristics and working time preferences. Overall, I find substantial roles for both firm-level differences in technology and individual characteristics. A large share of hours variation is also due to the sorting of individuals into firms based on human capital characteristics. By contrast, there is less evidence of sorting on labour supply preference characteristics like marital status and parenthood, despite differing working hours across firms. Within firms, after controlling for the effects of human capital on hours, preference characteristics have a statistically and economically significant effect on hours of work.