Researchers: Matias Iaryczower (Princeton), Matthew Shum (Caltech)
Some of the most important decisions in societies are made in committees voting bodies with a small number of members). In the US, the Supreme Court, the House of Representatives, and municipal school boards are examples of committees who make decisions with far-reaching consequences for the public. The overarching goal of this proposal is to develop an empirical framework for understanding how committees "work": how the disparate preferences and know-how of its members are aggregated into a collective outcome in different institutional environments.
To do this, we develop a new statistical and econometric model for analyzing decision-making in committees, in which members' actions depend not only on their preferences, but also on their information. This model is based on the theoretical literature of strategic voting with incomplete information, which has become the preferred approach to modeling committee voting in frontier research in political economy. Using data on the voting records of committee members in real-world committees, we estimate the model parameters, which correspond to the preferences (biases) as well as information of each committee member. Disentangling bias and quality of information in turn allows us to measure the value of information in the committee, to measure the effectiveness of committee decision-making, and to compute counterfactual simulations to assess how differently committee decisions would have been under alternative voting rules or committee compositions.
This proposal contains four projects which illustrate the power of our analytical framework in several important real-world institutions. The first project focuses on decisions in the United States Supreme Court. In this context, we consider whether case-specific information have enough power to overturn the prior biases and ideological considerations of the justices. The second and third projects focus on decision-making in state supreme courts, and address questions regarding differences in bias and quality of information of appointed and elected justices (bureaucrats and politicians), and also the effects of campaign financing on elected judges' voting behavior. In the fourth project, we analyze voting in the US congress during the founding fathers period (Congresses 1 to 17).
Our work has important implications for the analysis of policy and effective design of voting institutions. Our empirical framework allows us to quantify (with minimal data requirements) the effectiveness of committee decision-making in different institutional settings. This in turn allows both an evaluation of real-world committees and, via counterfactual simulations, to assess whether alternative institutional arrangements (such as majority vs. unanimity rules, or limits on campaign contributions) could lead to better decisions. The computer code and data files generated by our research can be applied readily by other researchers to analyze committee voting in myriads of other settings.
