Which of the following can cause OLS estimators to be biased

Which of the following can cause OLS estimators to be biased? Heteroskedasticity. Omitting an important variable. A sample correlation coefficient of .95 between two independent variables both included in the model.

Solution

The only circumstance that will cause the OLS point estimates to be biased is b, omission of a relevant variable. Heteroskedasticity biases the standard errors, but not the point estimates. High (but not unitary) correlations among regressors do not cause any sort of bias.