10.1 This exercise refers to the drunk driving panel data regression summarized in Table 10.1.

a. New Jersey has a population of 8.85 million people. Suppose New Jersey increased the tax on a case of beer by $2 (in 1988 dollars). Use the results in column (5) to predict the number of lives that would be saved over the next year. Construct a 99% confidence interval for your answer.

b. The drinking age in New Jersey is 21. Suppose that New Jersey lowered its drinking age to 19. Use the results in column (5) to predict the change in the number of traffic fatalities in the next year. Construct a 95% confidence interval for your answer.

c. Suppose real income per capita in New Jersey increases by 3% in the next year. Use the results in column (6) to predict the change in the number of traffic fatalities in the next year. Construct a 95% confidence interval for your answer.

d. How should standard errors be clustered in the regressions in columns

(2) through (7)?

e. How should minimum drinking age be included in the regressions?

Should it enter as a continuous variable or as a series of indicator variables?

Be specific about the information you use to assess this question.