Hausman Test (significant) -> Fixed Effects Model -> half of the Variables omitted because of collinearity

[Eva Lan]

Hi everyone,

I just used the Hausman Test to get to know, if I need the Fixed or Random Effects Model. The test is significant. That's why I used the Fixed Effects Model.
My data is panel data with several people with individual IDs, Variables about salary, working hours, working times and so on (changes within and between the individuals). The dependent variable is vacation days. Moreover, I've got information about age, sex, ... (those values don't change within the individuals, but partly between).
Latter get omitted by the Fixed Effects Model (omitted because of collinearity).

It makes sense that those variables are omitted - for sure. But how can I consider the effects of the latter variables of my panel data on vacation days? I am not used to stata and really don't get it. Especially because the Hausman Test told me to use the Fixed Effects Model.

Thank you very much!!

Bests
Eva

[Carlo Lazzaro]

Eva:
welcome to this forum.
For the future, please post what you typed and what Stata gave you back via CODE delimiters (see the FAQ on this and other posting-related topics). Thanks.
As per its machinery, -hausman- compares the coefficients that are common to both -fe-and -re- specification.
If your variables are omitted due to collinearity under -fe- there's nothing you can do, but change your model specification.
As an aside, see https://blog.stata.com/2015/10/29/fi...dlak-approach/

[Eva Lan]

Hallo Carlo,

okay, thank you very much!