Job Market Paper
I show theoretically and empirically that an important reason for firms to pay workers through profit sharing is to weaken their unions. I do that by first documenting a series of stylized facts that support this hypothesis with a new data set on French firms. Both profit sharing and the presence of unions increase with firm size, firms with unions are more likely to resort to profit sharing, while strike incidence decreases with its usage. Second, I develop a model to study the effects of profit sharing on union behavior which introduces two novel mechanisms. The first one revisits an intuition Hicks had about strikes as the “weapon” that unions use to build reputation for being strong. The second is the effect that profit sharing has on unions. By making employee compensation depend on output, unions internalize the cost of their strikes and are less inclined to organize collective actions. This in turn damages the credibility of their strike threats. Over time unions lose reputation and bargaining power and as a result wages grow more slowly. Third, I test the model which predicts that firms increase their usage of profit sharing when unions are more likely to organize strikes. For that I use arguably exogenous dates of elections of union representatives, which give incentives for unions to organize collective actions in a competition for voters. I show that employers anticipate the effect of elections by increasing their usage of profit sharing, which payment leads to a reduction in strike length the same year, and to a drop in wage growth by about 12 percent the year after. The effect is heterogeneous and concentrated on lower skilled employees for whom wage growth is almost halved. It is driven by a reduction in the bargaining power of unions which are less likely to conclude wage agreements with employers.
We provide empirical evidence on the discrimination in terms of wages, working conditions, and career opportunities of union representatives by employers in France. On average union representatives are paid 4 percent less than non unionized workers and are one third less likely to be promoted, but this hides a dichotomy situation. Representatives are divided into a group of strongly penalized members, and another, who are advantaged in all aspects. This fits a theory where employers strategically discriminate against union representatives to affect the split of the surplus: they would either bribe representatives willing to accommodate with them, or penalize those that adopt a defying stance. Many aspects of the data fit the theory. First, representatives are more penalized when they bargain often, and when they disagree with the employer on the terms of bargaining. Second, advantaged representatives do not participate in strikes when they happen, and they belong to unions known to be softer, while the opposite is true for those that are penalized. Eventually we show that the strategy of employers is beneficial for them. The larger the penalty of union members, the less employees unionize and the more they are likely to declare they are afraid for the consequences unionization may have on their careers.
We develop a new framework of spatial firm-competition based on a demand system that allows a wider degree of substitutability or complementarity between goods than those in the existent literature. This allows us to characterize a variety of geographic configurations between captive and competitive areas based on the locations of firms that cannot be done using traditional models, which assume unitary demands. In particular, we show that entry of competitors can result in higher prices for customers. When the new competitor is located far enough and caters only to a few customers at the border of the geographic market of an incumbent monopolist, the latter gives up these customers who have become less profitable. It prefers to focus on a narrower geographic area of customers who are still captive because the transportation costs to shop at the competitor are too high. The customers’ loyalty allows the monopolist to raise its prices, which were lower to reach out to the fringe that is left over to the new competitor after entry. Consumers who are only served by the monopolist are strictly worse off. We use data on prices and locations of gasoline stations in California to provide supportive evidence for the theory.
Work in Progress
The Responsiveness of Profit Sharing to Tax Incentives
Over the last decades politicians of many developed economies have set up fiscal incentives to stimulate the usage of profit sharing plans by employers. To what extent have they served their purpose? To answer this question I use a reform of the social security payroll tax in France that lead to a wide exogenous variation in the taxes paid on wages across firms between 2003 and 2005. Since profit sharing is exempt from payroll taxes the relative tax rate between profit sharing and wages either increased or decreased depending on whether firms had complied with legislation on working hours prior to the reform. I find that firms adjust their usage of profit sharing in response to variations in its relative tax rate but the adjustment is asymmetric. The adjustment is stronger in response to a reduction of its relative taxation than to an increase of the same amount.
Disability and Optimal Retirement Age Policy
Social security retirement benefits are based on career income but vary only marginally along other dimensions of jobs such as hardship or mental stress. On top of that, workers do not have a complete choice over their retirement age which is a key adjustment variable to hedge against late life illness. To the extent that income does not fully compensate for health risks, social security systems are a contributor to the differences in lifetime expectancy across occupations that can go past ten years at the legal retirement age. To assess to what extent these inequalities are due to the rigidity of current systems I build a structural model of career and retirement decisions to conduct a set of a policy simulations. First, I allow workers to freely choose their retirement age, then I introduce retirement benefits that compensate workers for the hardship of their occupations. This exercise gives rise to counterfactual distributions of life expectancy, which I use to quantify the reduction in the gap between occupations that more flexibility would allow for.