(Causal Inference with Machine Learning: an application to tax evasion)
This is a Lato Sensu specialization capstone project where I try to first, replicate the results from Fellner et al. (2013) and second, extend it by using novel machine learning techniques to infer heterogenous treatment effects, (Chernozhukov et al., 2018 and Athey, Tibshirani, Wager et al., 2019).
Codes in R and Python are provided.
ATHEY, Susan; TIBSHIRANI, Julie; WAGER, Stefan et al. Generalized random forests. The Annals of Statistics, Institute of Mathematical Statistics, v. 47, n. 2, p. 1148–1178, 2019.
CHERNOZHUKOV, Victor et al. Double/debiased machine learning for treatment and structural parameters. The Econometrics Journal, v. 21, n. 1, p. c1–c68, jan. 2018. DOI: 10.1111/ectj.12097.
FELLNER, Gerlinde; SAUSGRUBER, Rupert; TRAXLER, Christian. Testing enforcement strategies in the field: Threat, moral appeal and social information. Journal of the European Economic Association, Oxford University Press, v. 11, n. 3, p. 634–660, 2013.