Seminar in Law , Economics , and Organization

The extent to which firm ownership is concentrated or dispersed varies considerably from country to country. Explanations based on differences in the protection of minority shareholders leave a significant part of the variance unexplained. This paper offers a novel explanation for the variation in ownership concentration across countries based on differences in the quality of labor relations. We show empirically that countries in which labor relations are hostile tend to have more concentrated ownership. These results continue to hold when we instrument labor relations using religion, which has been argued is a key factor in the history of European labor organizations. We find similar results using historical data for Canada, documenting a surprisingly strong correlation between strike activity and ownership concentration over the past 50 years. Theoretically, we show that these findings are consistent with a simple model of repeated wage bargaining under asymmetric information in which labor relations and ownership concentration are jointly determined. JEL codes: D2, G3, J3, J5 PRELIMINARY AND INCOMPLETE ∗We thank Randall Morck, Michael Percy, Gloria Tian and Bernard Yeung for providing us with their data on the history of corporate ownership in Canada. We thank Pierre Fortin for providing us with data on strikes and union membership in Canada. †Stern School of Business, New York University.