participants to the 3 rd ECB Labour Market Workshop on " How are wages determined in Europe " for their comments on previous versions of this paper. The views expressed in the paper are personal and do not engage the organisations to which the authors are affiliated. Abstract The paper explores the link between wage premia and the determinants of product market rents. We first estimate 2-digit industry premia from 1996 wage earnings data by category of worker (age, sex, education and type of contract) in 10 European countries, the US and Canada. Using industry-specific regulation data, we then look at the effects of restrictions to competition and public ownership on wage premia in non-manufacturing industries, where regulatory conditions vary the most and are better documented. We find that, given workers' bargaining power, anticompetitive regulations significantly increase wage premia, reflecting the presence of rents. However, premia decline in industries dominated by legal public monopolies, suggesting a hump-shaped relationship between regulation and premia. We show that the hump-shape is consistent with a model of non-pecuniary rent-sharing between workers and a populist public monopolist. If the labour market was perfectly competitive, wage differentials would reflect only the characteristics of hazards, etc.). This is far from being the case, however, and the term of "wage premium" refers to the difference between the wage actually earned by a worker, and the wage that he would have been expected to earn, based on his observable characteristics (and working conditions). In this working paper, we study to what extent these wage premia reflect interindustry differences in competitive pressures and employee bargaining power. The intensity of competition influences the magnitude of rents firms are able to extract from product markets. If workers and firms bargain over wages, part of these rents is likely to be appropriated by workers, to an extent depending on their bargaining power. There is abundant evidence of a positive relationship between product market rents (or measures of market power) and wage premia (or workers' bargaining power). One problem with this evidence is that it is often affected by potential measurement and endogeneity problems: proxies for product market competition are difficult to construct, and most available measures (such as profit per worker, markups or concentration rates) are likely to be determined jointly with the wage outcomes. Moreover, there is no univocal relationship between many of these empirical measures and the actual degree of product …