Discrimination and public policies[Discrimination et politiques publiques]

In all three chapters of this dissertation, I try to see whether discrimination and own-group preferences exist, in different contexts, and what kind of public policies could mitigate or balance the negative effect of these preferences. In my first chapter, I rely on a famous ruling that changed the monopsony power of firms to see whether, as predicted by Becker, market failures have an impact on wage discrimination. I find that as monopsony power decreases, firms are no longer able to act on their prejudice, and wage discrimination disappears. This result shows that labour market context is essential in evaluating public policies, and that prejudice need not necessarily translate into wage discrimination. In the second chapter, I analyse the effect of another public policy, a 2015 reform that imposed gender quotas in academic recruitment committees. The reasoning of the policymakers was that increasing the share of women evaluators would improve the outcomes for women. I find the opposite instead; women are ranked worse by hiring committees after the reform. However, this result does not show that women discriminate against women. There is some evidence that this result is caused by the reaction of male jurors to the reform, since the negative effect of the reform is found only in committees that are helmed by male jury presidents. This chapter shows that it is necessary to evaluate public policies, lest reforms that are well-meaning in intention turn out to cause more harm than good. In the third and final chapter, I show two stylised facts: When cities decline, they tend to become more black, and black residents are disproportionately located in cities that pay low wages. One explanation for this could be that living in cities with a larger share of black residents is a positive amenity for black workers. I try to see whether workers have preferences for living in cities that have a larger share of co-ethnic residents, when controlling for wages, rents, transfers and network amenities. I find that these preferences are significant, and then try to see what share of the wage gap these preferences, and the imperfect sorting they imply, could explain.