reghdfe and constant term reporting

[Jennifer Xie]

Hi there,

I use the code below, but no constant term shown in the results. Can someone help me with the code to report the constant term?

reghdfe y x1 x2 x3, absorb(industry year) vce(cluster stock)

Regards,
Jennifer

[FernandoRios]

Make sure you have the latest version available from ssc
earlier versions did not report a constant
Also, the constant, in this type of models, may be of little of no interest
best regards
Fernando

[Phil Bromiley]

To expand on Frenando's comment, with fixed effects, you're allowing different intercepts for each of the absorb() conditions. Like any inclusion of dummy variables, you must either omit the intercept or omit one of the conditions so the dummies don't add up to 1. Thus, the reported intercept (if any) depends on which of the conditions is omitted. And almost always, we have no good reason for one of the panels to be the omitted panel rather than another. That reghdfe doesn't report an intercept means it uses the first option which is to include all the panel dummies but omit the intercept. [This is what it does in essence, but the calculation is done differently.]