(Requires Appendix material) When the fifth assumption in the Fixed Effects regression $\left( \operatorname { cov } \left( u _ { i t } , u _ { i s } \mid X _ { i i } , X _ { i s } \right) = 0 \text { for } t \neq s \right)$ is violated, then
A) using heteroskedastic-robust standard errors is not sufficient for correct statistical inference when using OLS.
B) the OLS estimator does not exist.
C) you can use the simple homoskedasticity-only standard errors calculated in your regression package.
D) you cannot use fixed time effects in your estimation.

Question

Quizplus · Accepted Answer

When the assumption of no serial correlation in the error terms is violated, heteroskedastic-robust standard errors are not sufficient, and one should use standard errors that are robust to both heteroskedasticity and serial correlation for correct inference.

(Requires Appendix Material) When the Fifth Assumption in the Fixed (cov⁡(uit,uis∣Xii,Xis)=0 for t≠s)\left( \operatorname { cov } \left( u _ { i t } , u _ { i s } \mid X _ { i i } , X _ { i s } \right) = 0 \text { for } t \neq s \right)(cov(uit​,uis​∣Xii​,Xis​)=0 for t=s)

(Requires Appendix Material) When the Fifth Assumption in the Fixed $\left( \operatorname { cov } \left( u _ { i t } , u _ { i s } \mid X _ { i i } , X _ { i s } \right) = 0 \text { for } t \neq s \right)$