Three-way fixed effects estimation

[Charlotte Rogers]

Hi all,

I want to estimate the following model:

A_ijs=X_ijs+a_i+b_s+c_ij+d_sj+eps_ijs

where A_ijs is achievement of individual i in subject j and school s, a is a student fixed effect, b is a school fixed effect, X is characteristic of individual i in school j and subject j and d is a school by subject effect.

I have panel data that contains two values of X and A depending on the subject. This allows the estimation of a student fixed effect model.

Any ideas on how I can do this?

[FernandoRios]

Hi Charlotte
Look Into reghdfe command which is a user written command in Stata.
Best