UC Irvine

Labor unions are a controversial and relatively little understood species of organization. While empirical research on the effects of unions on various labor market outcomes proliferated in the 1970s and 80s, theoretical understanding of the behavior of unions remains at its early stages. My dissertation aims to advance such understanding.

The first chapter delves into the rationale behind strikes, the ultimate source of a union's bargaining power and therefore the basis of its entire existence. The very occurrence of strikes has been challenging for standard economic theory to explain: given their manifest inefficiency, a mutually beneficial bargain should always be available such that strikes only serve as a weapon of non-use. Existing models of strikes tend to overcome this problem by assuming some form of private information. Borrowing insights developed in the field of conflict economics, I show that strikes can occur even in the absence of uncertainty (about the adversary's strength, costs, or reservation payoffs) once we consider the effect of the outcome of a strike on the relative strategic positions of the union and the firm in future interactions. I also show that shifts in the institutional environment in favor of one side can trigger strikes motivated by a desire to disrupt the existing balance of power, and use this framework to analyze strike waves observed in the United States since the late 19th Century.

The second chapter focuses on a special aspect of the internal politics of unions, namely the tension between the union leadership and the rank and file that has been regularly noted by labor movement scholars and practitioners. I develop a theoretical model demonstrating that when union leaders maximize the size of their organization rather than the welfare of the workers they represent, they can seek a settlement with the employer that is poorer than what the workers can expect to obtain through a strike. This result agrees well with the historical pattern of union officials often acting as a conservative break against more militant rank-and-file demands, and formalizes some insights present in Michels (2015)'s famed Iron Law of Oligarchy. In an extension of the baseline model, I show that union security agreements can amplify incentives for the union leadership to under-represent the interests of the workers.

Right-to-work (RTW) laws prohibit union security agreements in many U.S. states, and are frequently portrayed as a mortal threat to unionism in the country. It is claimed that by encouraging free-riding on union services, these laws undermine the bargaining power and viability of unions. The third chapter tests the substance of such claims by evaluating the impacts of RTW laws passed in six U.S. states in the 21st Century. Uing a mix of empirical methods including difference-in-differences, event studies, and synthetic controls, I find evidence that in the private sector, RTW laws decrease union coverage by more than 10 percent, all else equal. Surprisingly, I find RTW laws to have only a small and insignificant effect on free-riding behavior as measured by the share of unionized workers who are nonmembers. Moreover, union formation through NLRB-administered elections do not appear to be adversely affected by RTW. I also find evidence that RTW legislation increases union wage differentials by up to five percentage points, which is suggestive of a change in union bargaining behavior. Separately evaluating the effects of the 2018 U.S. Supreme Court decision in Janus v. AFSCME, which effectively made the entire U.S. public sector right-to-work, I find union coverage in the affected states to have changed little, and union wage differentials to have increased by more than five percentage points. Some of these findings are quite novel, and challenge conventional assumptions about how RTW laws impact unions.