Cohen and Einav: Econometric and Forecasting- Report Writing Assignment

Internal Code: MAS342

Report Writing Assignment:

Question: (a)Cohen and Einav (2003, Figures 1, 2) plot the aggregate time series of traffic fatalities per vehicle mile traveled (VMT) and seatbelt usage over the sample period. Plot and compare each of the three aggregate time series with the time series for your assigned state. Comment on whether there is any visible effect of the passage of the seatbelt law(s) in your assigned state in the plots. (b)Report regression results of fatality on seatbelt usage in one table. (c)Explain why the fixed effects estimates are less likely to be biased than the OLS estimates in Table. Explain whether the most reliable estimate(s) in Table  provide evidence in support of the Peltzman effect. (d)Cohen and Einav (2003) report heteroskedasticity robust standard errors for regression estimates. Explain the advantages of using clustered standard errors instead. Explain how the statistical significance of the estimates in Table  change compared to those reported in Cohen and Einav (2003). (e)Use the estimates from columns four and eight from Table 1 for this part. Predict how many occupant and non-occupant lives could be saved if seatbelt usage in your assigned state were 0.9 instead of the actual value in 1997. Explain how you computed your prediction from the estimates and be sure to provide a measure of accuracy of your prediction. (f) Predict how many occupant and non-occupant lives could be saved in 1997 if your assigned state had a primary enforcement seatbelt law in 1997 instead of a secondary enforcement law. Base your prediction by running two way fixed effects regressions of fatalities on ds (a dummy variable for secondary enforcement), dp (a dummy variable for primary enforcement), and the same set of additional controls X as used in Table. Report the regression results you used to compute your prediction. Explain how you computed your prediction from the estimates and be sure to provide a measure of accuracy of your prediction. (g) Someone argues that the effect of seatbelt usage on fatalities may depend on whether or not the speed limit is 70mph or above. Explain what regression(s) you would run to test this hypothesis. Explain whether the regression results provide evidence in support of this hy- pothesis or not.