Working hour arrangements and working hours: A microeconometric analysis based on German time diary data

The labour market providing individual resources and economic well-being is still a topic in the economic and social policy discussion. In the course of time the traditional full-time work is diminishing, new labour arrangements are discussed (keyword: flexible labour markets). This study will contribute to the discussion of working hour arrangements by quantifying patterns of explanation of who is working when within a workday. In particular we want to disentangle certain working hour patterns and the final hours of work according to those different patterns allowing for market and non-market influences. The daily working hour patterns are analysed by two dimensions: the fragmentation of a working day (by the number of working episodes) and the timing of work time by location of those episodes within the day's period. Deducting such patterns allows not only to describe possible workday interruptions and workday behaviour in general, but to give hints for which groups of the society non-traditional working time is important. Once quantified, labour market policy has a sound base for a targeted policy. Our model is based on a microeconomic labour supply approach, however extended by two dimensions: first, by daily working time arrangements with focus on core and non-core working time crossed by number of episodes and, second, by labour supply factors in a market and non-market context. Our microeconometric estimates use a multinomial logit (MNL) model to explain the working hour arrangement probability and a MNL selectivity bias corrected hours estimation for arrangement specific working hours with correct asymptotic covariances. Our study is the first German study of this kind which could analyse the actual available German Time Use Survey 1991/92 from the Federal Statistical Office with ca. 32.000 time diaries.