marginsplot after xthtaylor command

[Rattiya Lippe]

Dear all,

I am currently using Hausman-Taylor model to test the the relationship of deforestation (independent variable) and informal employment (dependent variable). In the model, I also add the squared term of deforestation rate which is calculated manually as deforestation * deforestation. When it comes to the marginsplot, I seems to obtain the linear graph although the regression estimates show that the deforestation and squared deforestation have positive and negative effect on informal employment, repsectively.

But if I used the random effect model by using the ## function as follow: xtreg informal deforestation c.deforestation##c.deforestation, re

The marginsplot commands seems to produce the right graph.

Could anyone suggest how to generate the predictive margin plot after the xthtaylor command?

Thank you very much,
Rattiya

[Andrew Musau]

margins does not know that \(deforestation\) and \(deforestation^2\) are related as xthtaylor does not support factor variables. FernandoRios wrote a command f_able which you can download from SSC that will make this relationship apparent.

[Rattiya Lippe]

margins does not know that \(deforestation\) and \(deforestation^2\) are related as xthtaylor does not support factor variables. FernandoRios wrote a command f_able which you can download from SSC that will make this relationship apparent.

Thanks a lot !

[Andrew Musau]

If interested, see https://www.statalist.org/forums/for...tic-regression