interaction instrumental variable analysis for reghdfe

[Roxanna Sh]

Hi all,
I have a question about reghdfe.
Does reghdfe support interaction for instrumental variable analysis? I have the following model

reghdfe aa bb (dd#c.ee= c.ff #c.ee), absorb (id state)
or
reghdfe aa bb (i.dd#c.ee= i.ff #c.ee), absorb (id state)

< ff here is the instrumental variable for dd, and ee is my other variable that is not endogenous>


and I got the following error
_tss_0b_dd invalid name

I should mention that the model runs with ivreg2 ( I deleted absorb for ivreg2)

Thanks

[Nick Cox]

reghdfe is understandably a very popular download, but its author Sergio Correia hasn't posted here in a while.

If you get a good answer here from him or a user of that command, then great. Otherwise he suggests on his website that you make use of https://github.com/sergiocorreia/reghdfe/issues

[Andrew Musau]

I have a question about reghdfe.
Does reghdfe support interaction for instrumental variable analysis? I have the following model

reghdfe aa bb (dd#c.ee= c.ff #c.ee), absorb (id state)

reghdfe does not support that kind of syntax and is not designed for IV estimation. You may be confusing it with ivreghdfe from https://github.com/sergiocorreia/ivreghdfe, which is a separate command. In any case, if your endogenous variable and instrument are interactions, what is the problem with creating them beforehand instead of complicating the estimation?