iii. Run the regression yet again while substituting the fixed effect dummy vari- ables with the expression " i. year". What does Stata do now? [Hint: all three regressions should yield an identical estimate off 1; if you are using an older version of Stata you may need to preface the regression with "ri: " (see the xtreg help file). ] (f) How do you interpret the coefficient on8 1 in part 3e? What variation in the fatality rate are the time fixed effects controlling for, and what variation is being captured by3 1? (g) Regress log( fatalityrateit) # o $ 1speed70it +3 2 log(incomeit) + 0i + uit, where Gi is a set of state fixed effects. Do so by including the regressor "i. fips" in the regression. How do you interpret the coefficient on 1 now? What variation in the fatality rate are the state fixed effects controlling for, and what variation is being captured by8 1? (h) Run the same regression again, but replace the robust option with the cluster(fips) option. Have the coefficients changed? Have the standard errors changed? What does this regression account for which the previous one did not? (i) Tell Stata that the data are panel data, in which the fips variable defines entities and the year variable defines time periods by entering the command tsset fips year. Look up the help file for this command. (j) Run the previous regression again, this time using the atreg command, with the fe and cluster(fips) options. Verify that none of the results have changed. What has Stata done differently than in part 3h? Hint: where have the dummy variable coefficients gone?] (k) Regress log( fatalityrateit) 3 0+8 1speed70it 18 2 log(incomeit) + ot + 0it uit by using the xtreg command as in part 3j and including i. year as a regressor. How do you interpret the coefficient on8 1 now? What variation in the fatality rate are the state fixed effects controlling for, and what variation is being captured by B 1? Every state in the U.S. - as well as every county, and many other entities - have numerical identifiers known as FIPS codes. FIPS stands for Federal Information Processing Standards. California, for example, has the FIPS code 6