What is the impact of multicollinearity on a linear model?

Answer 1

It doesn't have much impact on the reliability of the model, but adds to the noise with unnecessary overfitting. Multicollinearity impacts on your assessment of which factors are really influential. Factors that are redundant should be dropped in good model design. For example, you could come up with a fairly good linear model predicting fuel economy that includes engine capacity and engine weight. But since capacity and weight are correlated one is redundant and should be dropped.