–To include people with zero incomes. This assumes including people with zero income is actually useful and reasonable in this context…5. Why did I not include a dummy for people who have not completed high-school?–When regressing across a set of complementary and mutually exclusive category dummies, I need to keep one group as my reference group (the mean of which is absorbed in the constant term for the model).6. Why do you think I included these variables? Can you think of other variables that would be important for explaining variation in income across individuals?–These seem to be the ‘usual suspects’ in terms of explaining variation in income across individuals: we expect people with higher education to have higher incomes, we know that income tends to increase with age (due to experience/tenure), and definitely with hours worked. It might also vary depending on whether you are partnered or not, and have children, though potentially differently for men and women.Week 6Pre-Lab ExerciseQuestions:7. What is the relationship between education and income? How does it differ between males and females?–Income does improve with education, though more strongly for men than for women. This appears to show that the financial returns of education is higher for males! To find out of this is true we’d have to also check for occupation. –Nonetheless, it seems that the difference between the bottom and top education groups is almost twice as high for males as for females. Why is this? Jobs dominated by women earn less than jobs dominated by men. Glass ceilings. The jury is still out…–Specifically, the difference for males can be worked out as follows:ln???= 9.44 + 0.597𝑃?–So when PG=0: ln???= 9.44; which means that Income = e9.44= $12,581–When PG=1 ln???= 9.44 + 0.597 = 10.037; which means that Income = e10.037= $22,857–Thus, the predicted “returns” to a postgrad degree, compared to no qualification, is $10,276, when we compare two males who are the same age, civil status (including kids), and work the same hours per week.