Hausman Test

[Pepito Grillo]

Hello,
This is my first post so forgive me if I explain anything wrong.
I am currently running a panel data and want to know whether to use Fixed Effects regression or a Random Effects regression, in order to do so I believe the Hausman test is the best possible way of doing this. From what I understand if the value of Prob>chi2 is above 0.05 then Random Effects should be used. However I get a value of 0.9375 which seems incredibly high, is there any explanation for this or am I doing something wrong?
Here is the code:

** Fixed Effects**
xtreg gincdif1 Immigrationfromforeigncountri RegionalDebtGDP1 UnemploymentRate lnRGDP, fe
estimates store fixed
** Now we run a Random Effects**
xtreg gincdif1 Immigrationfromforeigncountri RegionalDebtGDP1 UnemploymentRate lnRGDP, re
estimates store random

hausman fixed random

Many thanks,
Pepito

[Marcos Almeida]

Welcome to the Stata Forum / Statalist,

You may wish to take a look at threads on a similar subject, such as this as well as this one.

Hopefully that helps.

[Pepito Grillo]

Hello Marcos,
Many thanks for your help, by reading the other posts as mentioned I understand that in this case the random effects model is then the one to use, and that there is not too high of a value for the test?. The random effects provides me significant results whereas the fixed effects does not, so is it wise to just disregard the fixed effects results?
Thanks a lot.

[Marcos Almeida]

In general terms, yes, that's the explanation.

With regards to "disregard" - or not - one model, you may also keep an eye on the specific literature and the very purpose of the analysis.

[Pepito Grillo]

Perfect!
Many thanks for your help.