QUESTION 7

1. This exercise refers to the drunk driving panel data regression, summarised in Table 10.1 on pages 31 and 32 of the Lecture 8 slides.

   Suppose you are interpreting the estimated coefficient on Real Income Per Capita (Income) as the causal effect on Traffic Fatality Rate. Choose the correct statement:

   	a.	According to column (3), the true model must not include Income in the analysis.
   	b.	According to column (4), when Income increases by $1, the number of fatalities would increase by 1.82 deaths per 10,000 but the prediction is not statistically significant.
   	c.	According to column (5), when Income increases by $1, the number of fatalities would increase by 0.09 deaths per 10,000 but the prediction is not statistically significant.
   	d.	According to column (6), when Income increases by 1 percent, the number of fatalities would increase by 0.0179 deaths per 10,000 and the prediction is statistically significant.
   	e.	According to column (7), when Income increases by 1 percent, the number of fatalities would increase by 0.01 deaths per 10,000 and the prediction is statistically significant.

TABLE 10.1 Regression Analysis of the Effect of Drunk Driving Laws on Traffic Deaths Dependent variable: traffic fatality rate (deaths per 10,000). Regressor Beer tax. Drinking age 18 Drinking age 19 Drinking age 20 Drinking age Mandatory jail or community service? Average vehicle miles per driver Unemployment rate Real income per capita (logarithm) Years State effects? Time effects? Clustered standard errors? Drinking age coefficients=0 (1) 0.36** Unemployment rate, income per capita = 0 R (0.05) no 1982-88 1982-88 1982-88 no no (2) (3) -0.66* -0.64* (0.29) (0.36) 0.091 yes no yes yes yes yes 0.889 (4) -0.45 (0.30) 0.891 F-Statistics and p-Values Testing Exclusion of Groups of Variables Time effects=0 0.028 -0.010 (0.070) (0.083) -0.018 -0.076 (0.050) (0.068) 0.038 (0.103) 0.032 -0.100+ (0.051) (0.056) 0.008 (0.007) -0.063** (0.013) 1.82** yes yes (5) yes 10.12 4.22 (0.002) (