xtdpdgmm with fixed effects

[Katherine Oleas]

Hello everyone,

I am making a dynamic panel for which I am using xtdpdgmm. Here is the regression:
xtdpdgmm Domestic_Health L.Domestic_Health GPE_subindex GDP_PPP Urbanization y*, gmm(Domestic_Health,c) gmm( GDP_PPP Urbanization,c lag(1 3)) gmm(GPE_subindex,c lag(2 3)) two vce(robust)

I would like to know how I can run this regression with fixed effects.

Can someone help me thanks.

[Sebastian Kripfganz]

What kind of fixed effects do you have in mind? Time-fixed effects can be included with the teffects option. Country-fixed effects are not explicitly included. The idea of the GMM approach is to use instruments which are orthogonal to (or at least uncorrelated with) the omitted fixed effects. Typically, this is done by transforming the model into first differences, which can be done with the model(diff) option. In your case, all of the specified instruments currently refer to the model in levels, which implies a random-effects assumption. Also, you have not specified any instruments for y* (time dummies?). Such exogenous dummy variables still need to be instrumented (typically by themselves, i.e. with option iv(y*)).

Please have a look at my 2019 London Stata Conference presentation:

• Kripfganz, S. (2019). Generalized method of moments estimation of linear dynamic panel data models. Proceedings of the 2019 London Stata Conference.

Also, the latest version of the xtdpdgmm package comes with the new xtdpdgmmfe command, which allows you to simply specify exogenous and endogenous variables, and the command will choose the right instruments for you; see https://www.statalist.org/forums/for...84#post1675484