Both during lectures and problem sets, we have been discussing the classical \"return to education\" question multiple times. The starting point is almost always a regression of this form: log(wagei) = [30 + 31 educi + 32 agei + ui where wage is the monthly wage in dollars and educ is the number of years of education. (a) Some researchers suggests using distance to nearest school (dschool) and distance to nearest college (dcollege) at 16 years of age as instruments. Discuss whether these are likely to be good instruments. (b) How would you test for the relevance of these two instruments? Write down the exact hypothesis test you would be using (c) Some other researchers instead suggest running the following regression: log(wagei) = 30 + 31 educi + 32 agei + a1 dschooli+ a2 dcollegei + ul- And then plan to test for whether a1 = a2 = 0 arguing that this could test the Instrument exogeneity assumption. Do you agree with this claim