4. A researcher is interested in estimating the returns to education for male workers in their forties. Using a random sample of data on 524 individuals from 2005, the following estimated equation is obtained In(wage;) = 0.206 + 0.112educi + ui where i indexes the individuals, wage; is the individuals hourly wage, educi is the years of education the individual has obtained, and ui is an error term. (a) Interpret the estimated coefficient on education. Comment on whether you think the researcher can make ceteris paribus claims of the effect of education on potential wages. [8 MARKS] (b) It is suggested to the researcher that they can use observations on the individuals' parental education as an instrumental variable. Discuss the conditions required using parental income as an instrument for education. Be precise and include a brief discussion on which of these conditions the researcher can test. [8 MARKS] (C) A second researcher has obtained a panel dataset containing observations on the same group of individuals between 2005 and 2010. Using the first-difference (FD) estimator the researcher hopes to estimate the returns to education. What issues may the researcher encounter with such an estimation? [4 MARKS] 4. A researcher is interested in estimating the returns to education for male workers in their forties. Using a random sample of data on 524 individuals from 2005, the following estimated equation is obtained In(wage;) = 0.206 + 0.112educi + ui where i indexes the individuals, wage; is the individuals hourly wage, educi is the years of education the individual has obtained, and ui is an error term. (a) Interpret the estimated coefficient on education. Comment on whether you think the researcher can make ceteris paribus claims of the effect of education on potential wages. [8 MARKS] (b) It is suggested to the researcher that they can use observations on the individuals' parental education as an instrumental variable. Discuss the conditions required using parental income as an instrument for education. Be precise and include a brief discussion on which of these conditions the researcher can test. [8 MARKS] (C) A second researcher has obtained a panel dataset containing observations on the same group of individuals between 2005 and 2010. Using the first-difference (FD) estimator the researcher hopes to estimate the returns to education. What issues may the researcher encounter with such an estimation? [4 MARKS]
