Answer:
There are several factors that can cause ordinary least square (OLS) estimates of a simple regression model, y = β0 + β1x + u, to be biased:
1. Non-random sampling: If the sample used in the regression analysis is not random, then the OLS estimates may be biased. For example, if the sample only includes people who live in a specific area or only those who are a certain age, then the estimates will not be representative of the population as a whole.
2. Endogeneity: If one of the independent variables in the model is correlated with the error term, then the OLS estimates will be biased. This occurs when there is a omitted variable that is related to both the dependent variable and one of the independent variables.
3. Multicollinearity: When two or more independent variables in the model are highly correlated with each other, it can cause the OLS estimates to be biased. This occurs because the estimates become unstable when the variables are highly correlated.
4. Non-constant variance: If the variance of the errors in the model is not constant, then the OLS estimates will be biased. This occurs when the errors are not homoscedastic.
5. Specification error: If the wrong variables are included or excluded from the model, then the OLS estimates will be biased. This occurs when the model is misspecified.