Drunk Driving The First Review: Eisenberg's Evaluation of policies' effectiveness in relation to drunk driving is the first policy review piece chosen for this article. In the research Eisenberg (2003) presented novel findings on the effectiveness of public policies on the state-level linked to drunk driving, effectiveness in this context refers to reduction in the number of fatal crashes. The author argues that endogenous policies lead to biased conventional estimates of policy effects. With respect to the date of adoption, the concern was addressed by the writer by examining the policy effects' time pattern. According to results, the 0.08BAC law has an upward bias, but the policy is still of use. This type of analysis also evaluated MADD (Mothers Against Drunk Driving) organization and graduated licensing programs for drivers of a younger age bracket. There is no effect implied according to the MADD estimates, this could be due to crude nature of the variable used. The policy has an effect as far as young drivers are concerned, but the time model estimates are overstated in conventional analyses for graduated licensing (Eisenberg, 2003).

The Second Review: The research paper, chosen as the second policy analysis, is by Carpenter and his assessment of the whether or not there is any effectiveness in the Zero Tolerance Drunk Driving Law. The first comprehensive analysis was provided in his paper on ZT Laws effects on an individual level based on; involvement in drinking that was self-reported, heavy periodic consumption of alcohol, previous month's consumption of drinks in number, and young adults who are involved in drunk driving. Particularly, Behavioural Risk Factor Surveillance System (BRFSS) was used by the research which is a vast source of micro data, from the year 1984 to 2001 and includes data on the drunk driving and alcohol usage behaviour for young individuals over the age of 18.

In the United States the figures on drunk driving are staggering. Over 42,000 lives have been claimed due to motor vehicle accidents in 2002 alone. The factor for 40% of these cases was alcohol. The statistics for the youth are even more alarming. Crashes that involved alcohol for below the age of 21 were nearly double than that of older individuals (Carpenter, 2004)

Educational programs, stringent penalties and laws, and college campus outreach linked to young individuals who drink and drive are the range of ways proposed by political analysts in response to the problem of drunk driving and underage driving. According to Zero Tolerance Laws, it is illegal for drivers to carry detectable levels of alcohol in the bloodstreams. The violation results in prompt suspension of license and monetary fines. The enacting of these laws began after 1983, and since that time, as a response to the federal compulsion and monetary incentives, all the fifty states and Columbia have introduced ZT jurisdictions. The research conducted regarding the effectiveness of such laws has increasingly showed their effectiveness. That is, Zero Tolerance Laws are correlated with reductions in highway accidents involving alcohol, for the concerned age bracket (Carpenter, 2004).

The foundation of Zero Tolerance Laws is the theory of broken windows. It is a criminology theory regarding the establishment of norms and indicating effect or riots and theft on additional crime and misbehaviours. According to the theory, keeping an eye on urban environments and maintaining them in a peaceful condition might impede additional vandalism and an increase into worse anti-social behaviour. Since the theory was introduced by George L. Kelling and James Q. Wilson, renowned social scientists, in a 1982 article, there has been great debate on it both in academic and public circles. Furthermore, it has also been used for many reforms in criminal laws as an incentive (Carpenter, 2004).

Although, ZT laws are generally believed to be effective, there is still debate going on what are the mechanisms through which they affect alcohol related highway fatalities. Although minorsmay adopt an attitude of restricting their alcohol usage, making it more moderate and/or driving less often at times when they are under the influence of alcohol, these laws may even increase the alcohol usage by underage individuals if they choose to substitute from driving to drinking alcohol indoors or utilize public transportation. Furthermore, all these variables are policy relevant: research has increasingly indicated that alcohol consumption might have a causal relationship with several undesirable outcomes for the young population. So, information about how much and what are the effects of alcohol driving policies on youth might be a source of information of several health laws (Carpenter, 2004).

The standard difference-in-difference type method innate in the two-way fixed effects was supplemented by the identification of a controlled group of people who were otherwise like treated people but should have experienced the effects of Zero Tolerance laws. In short, a group of 22- to 24-year-olds in a difference-in-difference-in-differences (DDD) type framework were used for the purpose of comparing the group for alcohol induced treatment, 18- to 20-year-olds (Carpenter, 2004).
It was found that the overriding effect of such laws among the 18- to 20-year-old population was the reduction in extreme episodic drinking by around 13% which was accompanied by similar escalation in the chances of becoming a less-frequent drinker of alcohol and a decline in the overall drinks taken by the male youth. The highlighting results for male population are largely in affect by the addition of further repressors, which include state-specific linear time trends. Furthermore, in a DDD framework, the reduction in extreme occasional drinking are evident and consistent even when differences in age are sharp, that is, 20-year-old versus 22-year-olds. It was also found that the timing patterns of the decrease in extreme alcohol drinking were in line with the responses calculated using total overall fatality data (Carpenter, 2004).

The drinking behaviour of women falling in the age group of seniors is not controlled by the models furnished by the DD models. These models define the impact of ZT laws and intense occasional drinking as a major decrease is witnessed because of these laws. No decrement in the ratio of drinking or driving after being drunk was witnessed by the researcher, with respect to the policies designed to make it low (Carpenter, 2004).

According to the researcher, the disregarded cross state heterogeneity is specified by the models gauged by the researcher in order to determine the stage and tendency with respect to drinking. The models of highway traffic fatalities and usage of alcohol among young people have demonstrated biasness because of not taking the effects of year and unrestricted state into consideration, although this framework is being applied in the text of appraising policies and economics. Moreover, the controlling of month effects, year effects, state effects, and state trends have also demonstrated significant variation because of the policy (Carpenter, 2004).

With respect to the advances of this research as compared to the previous one, the sample size was 500% larger. It was 49000 young people. Because of having a sample size of significant number, the evaluation of ZT laws and its impacts was easier. It is necessary to evaluate these laws from time to time because it was implemented in most of the states between the years of 1995 to 2000. The observation till 2001 is included in BRFSS along with all the related information. Moreover, the related information regarding all the 50 states and District of Columbia is included in the current data. On the contrary, the previous researches only consist of pre and post comparisons using individual level data (Carpenter, 2004).

The subjects were used intimately by the researcher in age to the treatments as the age group of the samples in BRFSS was 18 plus, but the laws should not affect the subjects falling in the age group of 22 and 24, in order to net out disregarded shocks of year or state. Moreover, this research also included the demographic information, which was used in considering the subgroups affected by ZT laws, and in controlling the biasness in the estimator which was difference-in-difference in nature and regression adjusted. Besides, the body weight is also included in this research which will help in determining the level of ZT treatment (Carpenter, 2004).

Dummies are incorporated for study month in order to control seasonality which is very crucial empirically in the usage of alcohol. The inclusion of real date of interview in the research enabled the researcher replacing the usage of the common technique of allotting partial policy values for mid-year law adoption with the correspondence of a particular ZT policy in the result of every examination (as cited in Carpenter, 2004).

The ultimate objective after relating the minimum purchase age (MPA) laws and minimum legal drinking age (MLDA) laws with the typical ZT laws is whether it is applicable on the individual level data in order to observe the procedure behind the observed reduced form…