Has Work Sharing Worked in Germany?

Starting in 1985, (West) German unions began to reduce standard hours on an industry-by-industry basis in an attempt to lower unemployment. Whether ‘work-sharing’ works – whether employment rises when hours per worker are reduced – is theoretically ambiguous. I test this using both individual data from the German Socio-Economic Panel and industry data to exploit the cross-section and time-series hours variation. For the 1984–9 period, I find that, in response to a one-hour fall in standard hours, employment rose by 0.3–0.7%, but that total hours worked fell by 2–3%, implying possible output losses. As a group, however, workers were better off as the wage bill rose. The employment growth implied by the mean standard hours decline, at most 1.1%, was not enough to bring German employment growth close to the US rate. Results for the 1990–94 period were more pessimistic.