A Panel Data Analysis of the Effects of Wages, Standard Hours and Unionisation on Paid Overtime Work in Britain

Abstract

This study examines the effects of the basic wage rate, standard working hours and unionisation on paid overtime work in Britain using individual-level data from the New Earnings Survey over the period 1975-2001. For this purpose we estimate a panel data model. We show that to obtain consistent estimates it is important to allow for both the censoring of paid overtime hours at zero and for correlations between the explanatory variables and the unobserved individual specific effects. The main empirical results are that a reduction in standard hours increases both overtime incidence and overtime hours, while an increase in the wage rate decreases the incidence of overtime but brings a small increase in overtime hours for those working overtime. For men the effects are stronger than for women. Union coverage is of minor empirical importance. The occupation and industry structure of employment has shifted from high to low overtime jobs. Taken together these economic variables can explain almost half of the changing incidence of overtime for men, and most of the change in overtime hours worked by women, but are less successful in explaining the changes in overtime hours worked by men or the incidence of overtime for women.

Keywords: panel data, censoring, overtime work

JEL Classification: C23, C24, J2