Question 2 Assume that the following annual earnings (Y) equations were estimated separately for union (subscript U) and nonunion (subscript N) workers. ED denoted years of education (averaging 12 for union workers and 14 for nonunion workers). SEN denotes years of seniority with the company (averaging 10 for union workers and 8 for non-union workers). Yu =22,000 + 1000 EDu + 1000 SENT YN = 8000 + 2000 EDN + 500 SENN a) Comparing the estimated coefficients for each variable (including the intercept) for the two regression results for U and N workers, what do these coefficients imply about union wage-setting practices? b) Calculate the expected earnings for the average union and nonunion worker. [note/hint: The expected earnings for the average union and nonunion worker can be calculated by substituting in the average value for their characteristics.] These numbers also give you the average earnings of union and non-union workers. c) What is the total or "raw" union-nonunion earnings differential i.e. not adjusted for differences in their wage-determining characteristics of education and seniority? [note/hint: remember that the estimated regression line Y=a + bX passes through the means of the data (Y, X) ] d) What are the expected earnings that a union worker would receive if paid accordingto the nonunion pay structure? e) Use the Oaxaca decomposition technique to decompose the average union-nonunion earnings difference into its component parts: one component attributable to the average differences in their endowment of wage-determining characteristics; the other due to pay differences for the same characteristics