Data and R Replication Code for Duch and Solaz We conduct tax compliance experiments in which subjects earn real money, are subject to a tax, and can lie about their earnings. Performing better on incentivized real effort tasks results in more cheating. Cheating rises as the tax rate increases; but the higher levels of cheating by those who excelled at the tasks persists in high and low tax treatments. This correlation persists when earnings are associated with luck; the correlation persists in experimental treatments with more redistributive taxation. The correlation persists for choices made in other games by the same subjects : High performance types give less in a conventional Dictator Game and also cheat more in a classic die game in which they privately report the results of tossing a die.

This replication packages includes: Experimental ZTree code: To reproduce the experimental sessions

• Oxford 2012 MT - Baseline
• Oxford 2013 MT - Redistribution
• Oxford 2013 MT - Status
• Oxford 2015 MT - Shock
• Oxford 2016 MT - Non-Fixed

Data:

• "MasterfileOxfordChile_20160506.dta", full dataset with all experimental sessions.
• "MasterfileOxfordChile_20160506_b.dta", full dataset with recoding used for figures.
• "base.dta", which includes 'baseline' and 'status' treatment sessions.
• "shock.dta", includes 'shock' treatment sessions.
• "redistribution.dta", includes 'redistribution' treatment sessions.
• "non_fixed_Oct2017.csv", includes 'Non-Fixed' treatment sessions with incentivized and fixed die payments.

Code:

• "Bootstrap_replication.do", to reproduce the boostrap data used in figure 2.
• "Duch_Solaz_Laroze_figures.R", replication code for other figures and summary statistics in the text.
• "Dach_Solaz_Multivariate_Laroze.R", replication code for table 4, figure 4, and appendix material.

Instructions:

The instructions for the different treatments are included in each of the folders. Within each treatment there were different sessions that varied with regards to the level of taxation (from 10-40%) and audit rate (0-100%). The instructions for each of these variations only change with regards to the level of taxation and audit that was, correctly, informed. All the rest stayed the same.

Manuscript:

• PDF of the manuscript
• Original .tex of the manuscript.