This is an audit experiment. To avoid contamination, no information is provided at this point. A detailed description is provided in a separate data analysis plan that will be made public upon completion of the audit experiment.
Experimental design
This is an audit experiment. To avoid contamination, no information is provided at this point. A detailed description is provided in a separate data analysis plan that will be made public upon completion of the audit experiment.
Design details
Randomization method
Corruption instrument
-- Brazil's public audit lottery
Audit experiment
-- randomization using Stata
Randomization unit
Corruption instrument
-- Unit of randomization: municipalities
Audit experiment
-- Unit of randomization: municipalities
-- Block randomized with respect to state and audit instrument
Clusters
5320 municipalities
Observations
20,994 interactions with 6998 offices in 5320 municipalities
Arms
Treatment
-- 7,560 interactions with 2,520 offices in 1,867 municipalities
Control
-- 13,434 interactions with 4,478 offices in 3,453 municipalities
Primary outcomes
This is an audit experiment. To avoid contamination, no information is provided at this point. A detailed description is provided in a separate data analysis plan that will be made public upon completion of the audit experiment.
Cited by
No citation data available for this registration yet.