Three Essays on Bargaining, Gender and Economic Policy

Shuwen Li

Advisor: Daniel E Houser, PhD, Department of Economics

Committee Members: Cesar A. Martinelli, Thomas Stratmann

Vernon Smith Hall (formerly Metropolitan Building), #5075
June 19, 2018, 02:00 PM to 04:00 PM


My dissertation makes both methodological and substantive contributions to non-cooperative bargaining and social dilemmas. Methodologically, I design two novel experiments to inform bargaining in a stochastic environment and effect of price-based interventions under a public goods environment with collective risk. Substantively, my findings of gender differences in bargaining over losses but not gains may help to explain some well-known gender results in the labor market.

The first chapter, “Stochastic bargaining over gains and losses: Evidence from the lab”, presents a novel experimental design to investigate predictions of the Merlo and Wilson (1995) stochastic bargaining theory. I construct four Merlo-Wilson games where the timing and efficiency of the equilibrium predictions differ, and study them in both gain and loss domains. I find little equilibrium play, but rather that it is most common to agree to equal splits in efficient stages, i.e. stages with maximum aggregate surplus. When there is an equilibrium in an efficient stage, however, proposers are significantly less likely to offer an equal split, and instead make a proposal that is unequal in the direction of the equilibrium outcome. In addition, I find that males and females are equally likely to propose an equal split over gains, while male proposers are significantly more selfish over losses. This gender effect cannot be explained by differences in risk attitudes. My results may shed light on some well-known gender differences in bargaining behavior.

The second chapter, “Revisiting gender differences in ultimatum bargaining: Experimental evidence from the US and China”, reports results from a replication of Solnick (2001), which finds using an ultimatum game that, in relation to males, more is demanded from female proposers and less is offered to female responders. I conduct Solnick’s (2001) game using participants from a large US university and a large Chinese university. I find little evidence of gender differences across proposer and responder decisions in both locations, suggesting that gender differences in bargaining can be context-based.

The third chapter, “Climate policies under collective risk: A laboratory investigation on the provision of local irrigation systems”, is motivated by the need for governments to partner with local farmers so that modern irrigation channels can be constructed in order to avoid losses from drought. I model the irrigation system provision problem as a threshold public goods game of loss avoidance and focus on testing whether government subsidies, as a type of price-based intervention, increase rate of successful provision. I design an experiment with scenarios involving either high or low risk of loss, while the local government provides either no subsidy, a one-to-one matching subsidy or a lump-sum subsidy. Results show that with no subsidy, the rate of provision of the public good is below the social optimum, and this is especially the case under high risk. Subsidies substantially increase the likelihood the irrigation project is undertaken, though groups remain under-insured in relation to the social optimum. A one-to-one matching subsidy appears to be at least as effective as lump-sum grants, although the improvement in success rates is not significant.