Fixed Effect - Constant Term in Regression

Dear all,

I have a question relating to fixed effect. I have encounter it many times in quantitative macro paper. For example, like in the following picture from a paper by Ottonello & Winberry.

As you can see, in the regression equation, they put two constant terms \alpha_j and \alpha_{st}, saying that which are firm and sector fixed effects respectivily.

So I am a bit confused, are they indeed dummy variables, for example \alpha_j will be 1 for firm j and 0 otherwise? I mean the dummy variables which are often used to kill fixed effects?

Is there anyone having an idea? Really appreciate!

Thank you,


Is this question related to Dynare?

I believe the answer to your question is positive.


1 Like

This is clearly not related to Dynare. But the fixed effects are indeed dummy variables. \alpha_j=1 will be one for firm j at all points in time and 0 otherwise. \alpha_{st} is a set of dummies depending on time t and sector s. If there are S sectors and T time periods, then you have S\times T dummies that are 1 if a firm j belongs to sector s and the time period is t and 0 otherwise

1 Like