Q2. What are the future works in this paper?
Generally, however, future research needs to more carefully control for a comprehensive set of regulatory variables and to account for grassroots activities, such as those by Mothers Against Drunk Driving, if it is to provide useful information on the independent impact of the different types of DUI legislation. Stricter anti-DUI laws, unless draconian in nature, are unlikely to yield a significant further decline in traffic fatalities. For instance, the 'preferred ' econometric estimates suggest that restoring the real 1988 beer tax to the level prevailing 13 years earlier ( a 78 % increase ) would have resulted in a 7 to 8 % reduction in highway fatalities, saving 3300 to 3700 lives annually. Without a full benefit—cost analysis and a further investigation, it is not possible to say whether such a tax increase is desirable.
Q3. What are the main factors that affect crash deaths?
Crash deaths are also likely to be affected by cyclical changes in the type and amount of alcohol consumed and possibly in driving patterns.
Q4. What is the main cause of death for people under 40?
Traffic fatalities, a major source of accident deaths at all ages and the leading cause of mortality for persons under 40, often involve alcohol.
Q5. What does the research by Manning et al. (1989) and Kenkel (19?
research by Manning et al. (1989) and Kenkel (1993b) suggests that liquor taxes only partially cover the external costs of drinking and that alcohol problems can more efficiently be reduced by raising taxes than by increasing the legal drinking age.
Q6. What could be correlated with the alcohol regulations in Table 2?
The alcohol regulations controlled for in Table 2 could be correlated with related policies (e.g. anti-plea-bargaining statutes, mandatory fines, or sobriety checkpoints), resulting in possible omitted variables bias.
Q7. How does the reduction in the per capita death rate from raising the minimum drinking age decrease?
For instance, the predicted reduction in the per capita vehicle death rates from raising the minimum drinking age from 18 to 21 declines by 70%, and becomes statistically insignificant, with the addition of covariates for per capita incomes, unemployment rates, and five DUI statutes.
Q8. How do they control for the heterogeneity?
Other researchers (e.g. Saffer and Grossman, 1987a,b; Chaloupka et al., 1993) attempt to control for the heterogeneity by including an unusually wide set of explanatory variables.
Q9. What is the reason for the decline in traffic deaths?
the percentage of traffic deaths involving drinking did not fall significantly over the time period, suggesting that driving may have become less risky for reasons unrelated to the prevalence of alcohol (e.g. the establishment of mandatory seat belt laws and the increased availability of vehicle safety features such as anti-lock brakes and air bags).
Q10. What is the sensitivity of the tax coefficient to the choice of fixed effects?
These results (which are not displayed) confirm that the tax coefficient is frequently positive and highly sensitive to the choice of regressors when fixed effects are omitted hut significantly negative and relatively robust to these changes in the FE estimates.
Q11. What are the only variables that have a statistically significant negative impact on traffic mortality?
In the most fully specified model, dram shop and administrative per se laws are the only regulatory variables ever observed to have a statistically significant negative impact on traffic mortality.
Q12. Why have you not used this strategy?
I have not used this strategy because of the difficulty in identifying the model by plausible exclusion restrictions, rather than relying on possibly arbitrary functional form assumptions.
Q13. What is the relationship between unemployment and traffic deaths?
Unemployment is expected to be negatively correlated with traffic fatalities, since total alcohol consumption and the proportion of drinking occurring in bars and restaurants is likely to fall during downturns.