sbordt / PublicGoodsHeterogeneity

Replication files for "Estimating Grouped Patterns of Heterogneity in Repeated Public Goods Experiments"

We used Python 2.7, Stata 14 and Matlab R2015b.

Replication Guide

Fischbacher and Gächter (2010) data setup

• Obtain the dataset from Fischbacher and Gächter (2010) and unzip it into data/new data zip file.
• Set the variable project_path in stata/paths.do to the project root directory.
• Run fg2010_data_setup.do

deOliveira et al. (2015) data setup

• Obtain the dataset from deOliveira et al. (2015) and unzip it into data/deOliveira2015.
• Set the variable project_path in stata/paths.do to the project root directory.
• Run deOliveira_data_setup.do

Grouped-Fixed-Effects in Python

• Run FG2010.ipynb and deOliveira2015.ipynb.
• The bootstrapped standard errors are computed inBootstrapped_Standard_Errors.ipynb.
• Run csv_to_dta.to to convert the Python output to Stata *.dta files.

C-Lasso in Matlab

• Setup the replication files by Zhentao Shi. Our code requires generic_functions/SSP_PLS_est.m and app_saving_PLS/hat_IC.m.
• Run FG2010.m and Oliveira2015.m.
• For the information criterion, see lambda.m.

Tables/Figures in Stata

• Run csv_to_dta.do to convert the Python/Matlab output to Stata *.dta files.
• Run regressions.do to obtain Tables 1 and 2.
• Run compare_classifications.do to obtain Table 3 (Panel B requires data by Fallucci et al. (2018) ).