Dear all,
I have a theoretical question regarding fixed effects in cross-sectional data.
Omitting region fixed effects will only bias the results of a linear regression if some unobserved, region-specific characteristics are correlated with both x and y, and will not bias the results, if only correlated with y, right?
Best,
Marco
I have a theoretical question regarding fixed effects in cross-sectional data.
Omitting region fixed effects will only bias the results of a linear regression if some unobserved, region-specific characteristics are correlated with both x and y, and will not bias the results, if only correlated with y, right?
Best,
Marco
Comment