Consider data set available at http://www.econometrics.com/comdata/wooldridge/WAGE1.shdwith variable descriptions available at http://www.econometrics.com/comdata/wooldridge/wage1D.text1.

We are interested whether there is marriage premium in wages for individuals with same number of years of education.

Construct new categorial variablemarrstatfor mariage status such thatmarrstat= 1 for single female, 2 for single male, 3 for married female, and 4 for married male.Assign also the above value labels to the values.

Make a scatter plot with education as x-axis andlwage(logarithmic wage) as y-axis such that individuals in each of the above marriage status groups are separated with their own colors.Impose regression lines in each of the groups into the scatter plot.

Construct new indicator variablesmmaleto indicate married man,mfemaleto indicate married woman, andsfemaleto indicate single female, such that mmale = 1 for married male and 0 otherwise, mfemale = 1 for married female and 0 otherwise, and sfemale = 1 for single female and 0 otherwise. Estimate regressionlwage = a + a1*sfemale + a2*mfemale + a3*mmale + b1*educ + b2*exper + b3*tenure + u, whereuis the error term. On the basis of the estimated result, is there marriage premium for men (that is do married men earn more than single men)?. Note that here each a-coefficient indicate wage difference to single males (the dummy variable left out from the regression).