"To overcome the problems inherent in cross sectional data, in the present paper we use panel data for 179 firms, from 1972-76 and from 1981-85, allowing for adjustment to the 1976 law. Both periods contain severe recessions, as well as more normal years, so the sample is still, if anything, biased towards exaggerating the costs of co-determination, though obviously less so than the single recession year cross sections. The empirical results suggest a significant, though small, positive influence on productivity from the 1976 strengthening of co- determination law."