METU Department of Economics 2019 Fall Econ301 Instructor: Erdal Özmen REVIEW EXERCISES 2 Q 1.Suppose you would like to estimate an earnings equation for T persons. Your independent variable is the earnings of the individuals (Y). As independent variables, you must include education (E) (measured in number of years) and ability (A). However, there is no data on ability and a proper proxy for it. Therefore, you ignore it in the regression model. i) Formulate the correct (Model 1) and misspecified (Model 2) regression equations. State what kind of specification error you make. ii) If you ignore ability variable, it is claimed that a) The least squares estimator of the slope term for education (b1) will always be biased. b) Since the bias will always be negative, it will be easy to adjust b1. c) There will be a gain in efficiency, since the variance of the estimator of b1will become smaller. d) The estimators of the intercept terms obtained from both models will be the same. e) The estimators of the education coefficient obtained from both models will be the same. f) The coefficient of determination for Model 2 will be higher that that of Model 1. State whether you agree or disagree in each case and give you econometric proof.