Stephen Hansen, Michael McMahon, and Sorawoot Srisuma, "Estimating Bayesian Decision Problems with Heterogeneous Expertise", Journal of Applied Econometrics, Vol. 31, No. 4, 2016, pp. 762-771. The file "simulations.R" produces output for the Monte Carlo simulations. The following files allow one to generate estimates of justice characteristics from the Supreme Court voting data in Iaryczower and Shum (2012): 1. getitright.csv. Contains the voting data from Iaryczower and Shum (2012). 2. issue_business_MM.csv. Contains the estimates of first-stage parameters reported in Iaryczower and Shum (2012) for business cases. These are used as starting values for the first-stage regression on business cases in "Estimating Bayesian Decision Problems with Heterogeneous Expertise". 3. issue_rights_MM.csv. Contains the estimates of first-stage parameters reported in Iaryczower and Shum (2012) for rights cases. These are used as starting values for the first-stage regression on rights cases in "Estimating Bayesian Decision Problems with Heterogeneous Expertise". 4. regression.R. Code for estimating first- and second-stage parameters for Supreme Court justices. For the second stage, there is a separate preference and expertise estimate for each unique combination of case covariates. Running this will take awhile. 5. aggregateIS.R. Code for taking output of regression.R and aggregating across cases to form single preference and expertise parameter for each justice. The above files are all ASCII files in DOS format. They are zipped in the file hms-files.zip. Unix/Linux users should use "unzip -a".