Department of Social Science and Policy Studies Working Papers
Using data from a repeated public good game, I conduct a Granger causality test and find that contributions and beliefs about the contributions of others decline together,with neither variable leading the other. As a result, I model contributions and beliefs using a system of simultaneous equations. Estimating the system provides evidence on the magnitude of the projection bias. Since contributions and beliefs move together, indicating that current and/or past values of one series are not useful for predicting future values of the other, I develop and test the hypothesis that contribution heterogeneity predicts changes in average contributions. I find support for my hypothesis using data from a variety of public good game experiments (with and without belief elicitation; fixed and random matching).
(2014). On the Dynamics of Contributions and Beliefs in Repeated Public Good Games. Department of Social Science and Policy Studies Working Papers, No. 2014-002.
Retrieved from: http://digitalcommons.wpi.edu/socialsciencepolicystudies-pubs/2