How to estimate the marginal effect after truncated regression?

[Faruque Sunny]

Dear all,
I am trying to estimate the marginal effect after truncated regression, but when using the command "margins, dydx(*)" after truncated regression, the result that STATA shows is identical to the truncated regression coefficients. I will be grateful if anyone kindly explains the procedure to estimate marginal effect after truncated regression.

[John Mullahy]

The margins help file entry for truncated regression is here

help truncreg_postestimation##margins

The default is that margins will compute the marginal effects for the linear predictor, which is why you are obtaining the result you report. As you can see in the help file there are other parameters for which margins can produce marginal effects; hopefully one of these is what you are seeking.

[Faruque Sunny]

Dear John Mullahy, thank you very much.