Union Leaders: Experimental Evidence from Myanmar

Abstract : Economic theory suggests that leaders may play a key roles in enabling social movements to overcome collective action problems through a variety of distinct mechanisms. Empirical tests of these theories outside the lab are scarce due to both measurement and identification challenges. We conduct multiple field experiments to test theories of leadership in the context of Myanmar’s burgeoning labor union movement. We collaborate with a confederation of labor unions as it mobilizes garment workers in the run-up to a national minimum wage negotiation. We present three sets of results. First, we document that union leaders differ from union members and non-members along several traits that psychologists and organizational sociologists have associated with ability to influence collective outcomes. Second, we randomly embed leaders in group discussions on workers’ preferred and expected minimum wage levels. A leader's presence in the group improves group engagement and increases workers’ consensus around the unions’ preferred minimum wage levels. Third, we conduct a mobilization experiment in which workers are invited to participate in an unannounced activity that features strategic complementarity in turnout. Leaders influence participation through both coordination and social pressure mechanisms rather than by simply motivating workers.

---

Labor Rationing

Abstract : This paper measures excess labor supply in equilibrium. We induce hiring shocks-which employ 24% of the labor force in external month-long jobs-in Indian local labor markets. In peak months, wages increase instantaneously and local aggregate employment declines. In lean months, consistent with severe labor rationing, wages and aggregate employment are unchanged, with positive employment spillovers on remaining workers-indicating that over a quarter of labor supply is rationed. At least 24% of lean self-employment among casual workers occurs because they cannot find jobs. Consequently, traditional survey approaches mismeasure labor market slack. Rationing has broad implications for labor market analysis.

---

Financial Incentives in Multi-layered Organizations: an Experiment in the Public Sector

Abstract : A classic problem faced by organizations is to decide how to distribute incentives among their different layers. By means of a field experiment with a large public-health organization in Sierra Leone and a structural model, we show that financial incentives maximize output when they are equally shared between a front-line worker and her supervisor. Compared to incentive schemes that target exclusively the worker or the supervisor, this intervention significantly raises the number of completed health visits by 61 percent. Also, the shared incentive uniquely improves overall health-service provision and disease incidence. We use these experimental results to structurally estimate a model of service provision and find that shared incentives are effective because (i) worker and supervisor effort are highly complementary, and (ii) contracting frictions discourage transfers between agents. These features have several implications for optimal incentive design. For example, the strong effort complementarity increases the relative effectiveness of schemes that incentivize joint output compared to schemes that directly reward agents for their effort.

---