Answer to Question #269777 in Economics for Siya

Question

Briefly explain multicollinearity and its impact on the standard error of an OLS estimator

Answer 1

Exact multicollinearity usually results from incorrect model specifications. In our case, you just need to exclude one of the variables from regression equations. However, multicollinearity usually means inaccurate multicollinearity, but a situation where there is no exact linear relationship between the variables, but they still change in a similar way. Formally, in this case, you need

talk about quasi-multicollinearity, but usually, it is also called simply multicollinearity.

The estimates are efficient and have the least variance among the unbiased estimates. Moreover, the estimates are also consistent. This is not mentioned in the Gauss-Markov theorem, but it is so. This means that if the sample size tends to infinity, then the estimates will tend to the true values ​​of the parameters that we appreciate.