Quantile treatment effect estimator by Abadie, Angrist, and Imbens (2002)

[Peter Irabor]

Dear Statalist,

Can anyone help with how to calculate percentage effects after estimating a quantile treatment effect estimator by Abadie, Angrist, and Imbens (2002) using ivqte Stata command. Abadie et al. (2002) in a footnote explain that ''Percentage effects were computed as the coefficient on training, divided by fitted values with the training dummy set to zero and other covariates set to means for the treated'', but not sure how this can be done using Stata's margins command. The reference to the paper is as follows: Abadie, A., J. Angrist, and Imbens, G. (2002). Instrumental variables estimates of the effect of subsidized training on the quantiles of trainee earnings. Econometrica 70: 91–117.

Thank you

[Blaise Melly]

Dear Peter,
You have to follow their instructions step by step. ivqte does not allow using margins or predict. You do the following: 1) You estimate your model with ivqte with the option aai, 2) You calculate the fitted value at the mean of the regressor by multiplying all the coefficients with the average values of the regressors (but keep the treatment variable to zero). 3) You divide the coefficient on the treatment variable by the fitted value at the mean.
Best regards,
Blaise Melly