I just need parts (e) (f) and (g) done please.

For part (e) can you provide the summary table of the variables for the year asked. Can you also provide the codes you used to get this table also? And can you provide a small description of how you calculated this data and what the data means.

Can you follow the same requirements for parts (f) and (g) asked above also. Please provide a good bit of detail on what the code on each parts means please.

Assignment

(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 pe- riod. 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 in the same format as Table1below.

(c) Explain why the fixed effects estimates are less likely to be biased than the OLS estimates in Table1. Explain whether the most reliable estimate(s) in Table1 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 Table1change compared to those reported inCohen and Einav(2003, Table 2, 3).

(e) Use the estimates from columns four and eight from Table1 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 Table1. Report the regression results you used to compute your predic-tion. Explain how you computed your prediction from the estimates and be sure to provide a measure of accu- racy 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.

Table 1: This table reports regression estimates of seatbelt usage on fatalities per 100m VMT. The outcome variable in the first four columns is occupant fatalities per 100m VMT and the outcome variable in the last fout columns is non-occupant fatalities per 100m VMT. The estimated standard errors in parentheses are either the usual OLS standard errors (if ‘Clustered SE' is N) or clustered by state (if ‘Clustered SE' is Y). If ‘Additional X' is Y, the same additional set of controls as inCohen and Einav(2003, Table 2, 3) but without the log transforms are included in the regression. If ‘State FE' is Y, a set of dummy variables for each state is included in the regression to control for state fixed effects. If ‘Year FE' is Y, a set of dummy variables for each year is included in the regression to control for time fixed effects. In this case the ‘Joint year F ' reports the F -statistic for the joint significance of the time fixed effects with p-value in square brackets.

Attachment:- Effects Mandatory Seat Belt Laws.rar