Desired hours worked over the business cycle: stylised facts for European countries

August 4, 2020
TEP Working Paper No. 1320
Attachment Size
Download PDF 3 MB
External Link

This paper documents stylised facts on desired hours per employed worker in European countries and identifies the effect of recessions on desired hours. Actual hours worked are usually used to estimate preferences on the labour market. However, actual hours are constrained by labour demand and therefore measure hours worked in the general equilibrium. Descriptive statistics from EU Labour Force Survey show that desired hours are countercyclical and that the underemployment gap increases due to higher desired hours worked of employed individuals. I identify the effect of recessions on desired hours using variation in regional unemployment rates from 2000 to 2017. I find that a 1 percentage point higher unemployment rate increases desired hours, on average, by 2 - 8 hours on a yearly level (3 - 5 minutes in the reference week). The results offer a lower bound estimate for the whole sample period of booms and busts. To narrow the sample period, I use a panel of individuals from the French LFS (EEC) and find even bigger effects. In France, from 2007q4 to 2009q1, an increase in regional unemployment rate by 1 percentage point increases desired hours by 1.6 hours in the reference week. Bottom decile of the income distribution significantly increases desired hours in all countries, suggesting an income effect labour supply response in recessions.