The labor supply of married women has been a subject of a great deal of economic research. The data file is \(m r o z\), and the variable definitions are in the file mroz.def. The data file contains information on women who have worked in the previous year and those who have not. The variable indicating whether a woman worked \(L F P\), labor force participation, takes the value 1 if a woman worked and 0 if she did not.

a. Calculate the summary statistics for the variables: wife's age, the number of less than 6-year-old children, and the income from other sources than from the wife's employment, NWIFEINC, for the women who worked \((L F P=1)\) and those who did not \((L F P=0)\). Define NWIFEINC \(=\) FAMINC - WAGE \(\times\) HOURS. Comment on any differences you observe.

b. Consider the following supply equation specification:

What signs do you expect each of the coefficients to have, and why? What does NWIFEINC measure?

c. Estimate the supply equation in (b) using OLS regression on only the women who worked \((L F P=1)\). Did things come out as expected? If not, why not?

d. Estimate the reduced-form equation by OLS for the women who worked, using work experience, EXPER, as an additional exogenous variable.

Based on the estimated reduced form, what is the effect upon wage of an additional year of education?

e. Check the identification of the supply equation, considering the availability of instrument EXPER.

f. Estimate the supply equation by two-stage least squares, using software designed for this purpose. Discuss the signs and significance of the estimated coefficients.

Transcribed Image Text:

HOURS = B +2 In(WAGE) + 3EDUC + BAGE +5KIDSL6+ KIDS618+ BNWIFEINC + e