1. What determines the wages earned by men? Some obvious candidate variables are education levels and experience. But there could be other factors as well. For example, there could be unions that bargain hard for the employees. Is there any racial discrimination on the part of the employers? It turns out (some of you may be surprised to know!) that marriage increases the wages commanded by men in the job market (does it?). The dataset is a sample of 545 males in full-time employment drawn from the National Longitudinal Survey (Youth Sample) conducted in the USA.1 Each of these 545 men in the sample worked in every year from 1980 through 1987. Some variables in the data set change over time: experience, marital status, and union status are the three important ones. Other variables such as race and education do not change. Use three methods to estimate a wage equation for men: pooled OLS, random effects, and fixed effects with log(wage) as the dependent variable to answer the following questions: 
2. 
   
3. (a) Do you see evidence of racial discrimination? How much is the difference in wages between white and Hispanic workers? 
4. 
   
5. (b) Which variables get dropped in the fixed effect analysis and why? 
6. 
   
7. (c) Observe the difference in coefficients for the variables married and union in the pooled OLS and FE models. Can you explain why the coefficients are different? 
8. 
   
9. (d) Which model will you prefer and why?