Based on the Current Population Survey (CPS) of March 1995, Paul Rudd extracted a sample of 1289 workers, aged 18 to 65, and obtained the following information on each worker:
Wage = hourly wage in $
Age = age in years
Female = 1 if female worker
Nonwhite = 1 if a nonwhite worker
Union = 1 if a union member
Education = years of schooling
Experience = potential labor market experience in years.
The full data set can be found as Table 6-16 on the textbook's Web site.
a. Based on these data, estimate the following model, obtaining the usual regression statistics.
In Wagei = B1 + B2 Age + B3 Female + B4 Nonwhite + B5 Union + B6 Education
+ B7 Experience + ui
where In Wage = (natural logarithm of Wage)
b. How do you interpret each regression coefficient?
c. Which of these coefficients are statistically significant at the 5% level? Also obtain the p value of each estimated t value.
d. Do union workers, on average, earn a higher hourly wage?
e. Do female workers, on average, earn less than their male counterparts?
f. Is the average hourly wage of female nonwhite workers lower than the average hourly wage of female white workers? How do you know? (Hint: interaction dummy.)
g. Is the average hourly wage of female union workers higher than the average hourly wage of female non-union workers? How do you know?
h. Using the data, develop alternative specifications of the wage function, taking into account possible interactions between dummy variables and between dummy variables and quantitative variables.