Should I run a pooled OLS with panel data or just use one of the years?

[Solomon Lin]

I have unbalanced panel data for two time periods. I am planning on running a fixed effects model. However, I would also like to run a cross-sectional OLS. Should I use a pooled OLS or simply run a cross-sectional regression for just one of the years? Is there a test to determine whether it is appropriate to run a pooled OLS on panel data (given the potential biases).

[Carlo Lazzaro]

Solomon:
the F-test at the foot of the outcome table afterr -xtreg, fe- wil show you whether or not fixed effect panel data regression should be preferred vs POLS (please, see .Example 2: Fixed-effects model in -xtreg, fe- entry, Stata .pdf manual).