5. The model below explains salaries of 353 major league baseball players (salary) expressed in thousands of pounds as a function of years in major leagues (years), runs scored per year (runsyr), and games per year in league (gamesyr). The ethnicity of players is indicated by a binary variable (hisp) which identifies whether the player Hispanic or not. The variable takes value 1 if the player is Hispanic and zero otherwise. The regression based on a sample of 353 players is as follows: = 267.50 + 67.60 years + 43.30 runsyr - 5.20 gamesyr - 22.34 hisp (16.35) (31.40) (5.20) (3.15) (15.60) R-squared = 0.55 a) Interpret the intercept of this regression. [5 marks] b) The variable years is a measure of experience of the player in the major leagues. How would you augment the estimated model above to allow the returns to experience to differ by ethnicity? [5 marks] c) Test the null hypothesis that Hispanic players earn less than non- Hispanic players, after controlling for experience (years) and performance in the league (runsyr and gamesyr). [3 marks] d) Suppose we want to test whether effects of years, runsyr and gamesyr on the salary of players differ by ethnicity of the player. Explain how you would implement this test and be clear about any assumptions you make. [8 marks] e) Suppose we have another variable other which is equal to 1 if the player is not Hispanic and zero when the player is Hispanic. If we add this variable to the model above, can we estimate the ceteris paribus effect of other on salary? Explain your answer. [4 marks]