Enforcement of the Zero Tolerance Law in California: A Missed Opportunity?

Abstract

This paper reports on the enforcement of a zero tolerance (ZT) law implemented on January 1, 1994, in California. Some provisions in that legislation merely intended to make enforcement easy; for example, using handheld, portable, breath-testing devices at the road-side. The research employs a model of the expected official actions, target group reactions, and underage fatal crashes to trace the impact of the law on underage drinking and driving. The results indicate little increase in the intensity of enforcement and no change in the perceived risk of arrest among the target group; however, results do indicate a possible reduction in drinking drivers involved in fatal crashes in that group.

ENFORCEMENT OF ZERO TOLERANCE LAWS

Between 1982 and 1996, the alcohol-related fatality rate for youth was cut in half, from 22 per 100,000 to 10 per 100,000 (NHTSA, 1998). The National Highway Traffic Safety Administration (NHTSA) credits much of this decline to the passage of minimum legal drinking age (MLDA) laws by all the states. NHTSA estimates over 16,000 lives have been saved for the nation as a whole by the MLDA laws (NHTSA, 1998). The universal adoption of this MLDA law and its apparent success in reducing underage alcohol-related highway deaths encouraged the enactment of additional federal legislation with-holding highway funds from states that did not pass ZT laws that make it an offense for a person under age 21 to operate a vehicle with any measurable blood alcohol concentration (BAC). In July 1998, South Carolina became the 50^th state to pass ZT legislation.

RATIONALE

Political logic held that if youth consumption of alcohol is illegal, than driving with positive BACs should also be illegal. There is considerable scientific support for this position. Using data from the Fatal Analysis Reporting System (FARS), Zador (1991) found that, for BACs in the .05 to .09 range, 16- to 20-year-old drivers were 21 times more likely to be in a single vehicle fatal crash than if sober and 12 times more likely to be in such a crash than drivers age 25 and over with the same BAC range. Mayhew and colleagues (1986), using Canadian crash data, found similar higher risk levels for underage drivers. Hingson and his coworkers (1991, 1994) published a series of studies on ZT laws in states that were among the first to adopt this legislation. Their results indicated that lowering the BAC limit for underage drivers reduced their nighttime single vehicle crashes by 10%. Consequently, they estimated 300 lives could be saved each year if all states enacted ZT laws.

FEDERAL REQUIREMENTS

While the NHTSA requirements for meeting the congressional mandate for ZT laws required states to specify a BAC limit no higher than .02, there was no specific model for this legislation. Each state could implement the limit within the context of its standard driving-under-the-influence (DUI) laws. As a result, the ZT laws are enforced and offenders prosecuted in different ways from state to state. Since the standard DUI enforcement process involves a sometimes complex set of laws providing for implied consent, administrative license suspension, and varying BAC limits, the grafting of ZT laws onto these existing statutes resulted in some practical problems for several states.

ENFORCEMENT ISSUES

Several potential problems with the resulting legislation can be noted. NHTSA required that the ZT laws not be secondary laws; that is, detection must not be dependent upon another offense. However, NHTSA specifications were silent on what observations were required of the officer to make the ZT charge. Detection of drinkers at low BACs where few, if any, of the traditional signs of intoxication are present is a problem. The law clearly intended to make a youth driving after drinking any alcohol an offense, regardless of whether impairment can be demonstrated. However, most states enacted legislation that did not clarify whether the officer must obtain evidence of impairment before issuing a ZT citation.

Another problem is that most of the ZT laws are per se laws, meaning it is necessary to have an evidentiary BAC test. Officers, however, derive their authority to require tests from implied consent laws. Therefore, these laws must be modified to apply to the .02 BAC level before the officer will have the authority to require a test from those for whom there is no probable cause to believe they have illegal BACs of .08 or .10.

Besides the problems of conflicts within the legislation, enforcing the ZT law may be limited by the typical operational policies of police departments and by attitudes of police officers. Voas and Williams (1986) compared the DUI arrest rates of drivers under age 21 to the frequency of their alcohol-related crashes. They found that the relative frequency of arrests for this age group was lower than for drivers over age 21. Preusser, Ulmer, and Preusser (1994) confirmed and extended this result, identifying several obstacles to enforcement of DUI laws among underage drivers. Among these was the deployment of patrols to downtown bar neighborhoods where adult impaired drivers, rather than young impaired drivers, were likely to be found. They noted that since young drinkers were less likely to be served in bars, they were more likely to be drinking in parks, on beaches, or in parking lots away from the normal police patrol areas.

The Preusser et al. study noted several other obstacles to enforcement of DUI laws with drivers under age 21. These included the time when most young people drink (which tends to be limited to weekend evenings), the different driving cues coming from young drivers with low BACs; the ability of young drivers to perform better on sobriety tests than older adults; and, perhaps the most important obstacle, the extra time that may be required to process a juvenile offender. Young offenders cannot be placed in an adult jail and usually must be held until released to their parents. Consequently, officers frequently have to “baby-sit” them for some time before returning to patrol duty.

THE CALIFORNIA ZERO TOLERANCE LAW

On January 1, 1994, California implemented a unique ZT law to overcome some of the obstacles to enforcing drinking-and-driving laws with underage offenders (Martin & Andreasson, 1996). This law enabled officers who had reason to believe that a driver under age 21 had been drinking, to require suspects to provide a roadside BAC measure by blowing into a handheld preliminary alcohol sensor (PAS). If the driver refused the test or if the test produced a BAC of more than .01, the officer was authorized to seize the driving permit (which was forwarded to the Department of Motor Vehicles [DMV]) and issue a citation that allowed the individual to drive for up to 30 days pending the imposition of a one-year suspension of the driver’s license. The action was purely administrative. While the offender had the opportunity to request a hearing with a DMV hearing officer, no charge was made in the criminal court; no trial was held; and no fines, jail, or other criminal penalties were imposed.

An important feature of this California ZT law is that it does not require officers to place suspects under arrest and transport them to the police station. In principal, the vehicles of cited drivers can be towed, and the driver can be released in any safe location, thus avoiding the necessity to detain or jail those with BACs less than the .08 DUI limit. In practice, most departments hold minors under age 18 until they are picked up by their parents. Because it is administrative, this law did not criminalize the underage offender, thus avoiding the costs of a court trial. Simultaneously, it provided for a significant penalty. The loss of the driver’s license for one year is a severe penalty for a young driver.

RESEARCH OBJECTIVE

This paper reports on the first 3 years of enforcement in California of the ZT legislation. Figure 1 provides the model that informs this research. The objective of the California ZT law is to reduce highway deaths and injuries among drivers under age 21. To achieve this objective, youth in that age group must be aware of the law and believe it is being enforced. This, in turn, should result in reduced impaired driving and fewer alcohol-related crashes. This study attempts to answer the questions implied by the model in Figure 1: “Were young drivers aware of the law?”; “Did the police enforce it?”; “Did young drivers believe it was being enforced?”; “Did they report reduced driving after drinking?”; and “Was the number of alcohol-related crashes reduced?”

Fig. 1.

Model of one expected impact of the California zero tolerance law

METHODS

DATA SETS

Four data sets were used to detect effects of the law and its implementation: (1) driver’s records from the California DMV, (2) random-digit dial telephone surveys of drivers ages 18 through 20, (3) interviews with underage drivers in weekend roadside surveys, and (4) fatal crash records involving underage drivers in California from NHTSA’s FARS file.

DRIVER’S RECORD DATA

Driver’s records of suspensions for alcohol-related offenses in California were obtained from its DMV. These records contain the reason for the suspension action and the BAC when arrested as reported by the arresting officer. Some demographic information is also available including the driver’s age.

The data set drawn for this research included drivers of all ages who had been suspended by the DMV between January 1, 1991, and December 31, 1996. Of the 1,347,927 violations during that time, 112,088 (or 8.3%) involved a driver under age 21. As would be expected, some drivers had more than one suspension. Each record is a violation for which an administrative license action was processed by the DMV. Thus, the file contains both suspensions for the adult .08 DUI offense and suspensions for the .01 ZT offense. These suspensions resulted from either an over-the-limit test result or a refusal to take the test, so they included every arrest for DUI, citation for ZT, or implied consent violation. The records also included cases canceled after a hearing (less than 1%). These apparently resulted from “invalid” arrests and were excluded.

TELEPHONE SURVEY DATA

A second source of data was three cross-sectional, statewide, random-digit dial (RDD) telephone surveys of 16- to 20-year-old licensed drivers in California. These surveys, conducted in January 1994, 1995, and 1996, covered drinking and drinking-and-driving behavior in the previous month (December 1993 December 1994, and 1995). Interviews were approximately 25 minutes and were given in either English or Spanish at the respondent’s request. A sample of randomly generated telephone numbers was obtained for each working prefix in California. Each number was called up to eight times at various times and days of the week.

Once contact was made, interviewers determined whether the number belonged to a business or a residential household. Nonresidential and nonworking numbers were eliminated from the sample. When a residence was reached, household eligibility was ascertained: at least one licensed driver, age 16- to 20, had to live in a household to be included in the study. If more than one 16- to 20-year-old licensed driver lived in the household, a random-selection procedure was used to choose one participant for the survey. Young people who lived in institutional settings or who were not well enough to complete the interview were excluded from the study. For respondents under age 18, the study was first described to a parent whose permission was obtained for the interview.

Interviewees were told that the survey was anonymous and participation was voluntary. They were assured that they could refuse to participate, refuse to answer specific questions, or decide at any time to end their participation in the survey. Each year, each survey attempted to contact 500 respondents equally divided between males and females. The number interviewed is shown in Table 1. Results reported were weighted to compensate for the number of phone lines in the home and the age of the respondent. The estimated response rate was 69%. (For a fuller description of the survey, see Grube & Voas, 1996.)

Table 1.

Number of drivers under age 21 in telephone surveys in California for 1993, 1994, and 1995

Gender	1993	1994	1995	Total
Female	243	246	252	741
Male	260	268	278	806
Total	503	514	530	1,547

ROADSIDE SURVEY DATA

As part of the NIAAA “Community Trials” project (Holder et al., 1997), roadside voluntary breath-test surveys were conducted on alternate weekends in two California cities—Oceanside and Salinas—over 4 years from July 1992 to June 1996. Surveys were randomly rotated through a dozen preselected locations, with police randomly stopping noncommercial motorists between 9:00 PM and 2:00 AM. Once safely stopped, drivers were approached and asked to participate in a brief survey and to provide a breath test. Of those stopped, 90% agreed to participate in the survey. One of the questions in the survey was “What is the legal BAC limit in California?” Responses to that question from those under age 21 were used in this study to determine when young drivers became aware of the California ZT law (For a fuller description of the roadside survey procedure see Voas, Lange & Treno, 1997).

FATAL CRASH DATA

FARS has records of all fatal crashes on public roadways in the United States. These data are collected from state record systems by NHTSA and entered into the FARS database. The database contains the time and location of the crash and the demographics of drivers, passengers, and nonoccupants involved in the crash. The BACs of drivers are recorded when these data are available from the states. Klein (1986) has developed an imputation system that permits assigning a probability that drivers fall into one of three BAC intervals when no measured results are available. These imputed BAC values were used in this research. Nine years of FARS data (January 1, 1988, to December 31, 1996) were obtained from NHTSA’s National Center for Statistics and Analysis. Records of all drivers involved in fatal crashes from California and four comparison states (Colorado, Nevada, Texas, and Wyoming) were drawn from the FARS file for that 9-year period. In 1993 when the study was designed, six states were selected for comparison with California; however, two states (Washington and Idaho) passed ZT laws 6 months after California and had to be dropped from the comparison group. (The remaining four states were only two thirds the size of California but otherwise similar with approximately 25% Hispanic populations.)

Driver’s records were divided into two age groups: youth under age 21 who were the target of the ZT law and a comparison group of adults, ages 21 to 34. Within the two age groups, three monthly time series were constructed: the experimental under age 21 drivers in California and two comparison groups—21- to 34-year-old drivers in California and under age 21 drivers in the four comparison states. The value for the monthly data points was the proportion of all crash-involved drivers with a BAC greater than zero during that month. Using a percentage, rather than the raw number of drivers with positive BACs, minimizes the effect of factors that influence the total number of fatal crashes but are unrelated to drinking and driving such as vehicle safety features, road safety characteristics, vehicle mileage, and the economy. Drivers with any amount of alcohol were included because the objective of the ZT law is to deter driving even at low BACs. Box and Tiao’s (1975) time-series analysis using ARIMA models, provided with the SPSS package, was used to analyze each of the three time series (which were first transformed using the natural log, so that the resulting exponentiated coefficients could be interpretable in terms of the relative percentage change).

RESULTS

KNOWLEDGE OF THE LAW

Figure 2 illustrates the responses of drivers under age 21 on their knowledge of the .01 BAC limit in California at roadside surveys in Salinas and Oceanside between 1992 and 1994. Before July 1993, between 10% and 20% of young nighttime drivers believed that the limit for those under age 21 was .01 or less. This increased to 50% to 60% in the latter half of 1993, before the California ZT law was implemented on January 1, 1994. The Figure 2 data, though from only two cities, are supported by results from the statewide telephone survey of underage drivers’ knowledge of the ZT law in December 1993. In that survey, 48.5% responded that the BAC limit was zero and 53.9% answered “none” to the question, “How many drinks can you have and drive legally?” Two years later, in the December 1995 telephone survey, 52.5% said the BAC limit was zero and 55.5% said “none” in response to the same question. Thus, there was no evidence that the law’s provisions were better understood by the underage drivers after the California ZT law had been in effect for 2 years.

Fig. 2.

Percentage of underage drivers who responded that the legal BAC was .01 or less in roadside surveys in Salinas and Oceanside

ENFORCEMENT

Since the number of licensed drivers in the target age group (under 21) varied during this period, it is important to consider population size in assessing the level of enforcement. The California DMV sampled 10% of its license records to estimate the number of drivers in various age groups as of midyear for each year covered in this study. In general, California had been experiencing a decline in under age 21 drivers (from a high of 1,229,660 in 1991 to a low of 1,084,380 in 1995), then rising a bit to 1,086,540 in 1996. The end of the downward trend mirrors the census data indicating that the baby-boomers’ children are beginning to reach driving age. Figure 3 shows the suspensions from either a DUI arrest or a ZT citation per 1,000 licensed drivers under age 21. This number was declining before the California ZT law was implemented and has risen slightly since the law became effective.

Fig. 3.

Suspensions resulting from DUI arrests or ZT citations per 1,000 drivers under age 21

While there is little overall change in the number per thousand receiving alcohol-related suspensions, there is a striking change in the offense charged. After the law was implemented, more than half of all the actions against underage drivers were based on the new ZT law rather than the regular DUI law with the .08 BAC limit. This might appear to suggest that the police redirected much of their effort from youth with high BACs to youth with low BACs. However, Figure 4 demonstrates that while the BACs below .08 increased by a factor of 5, from 4% to 20%, a majority of the underage drivers suspended for a drinking driving offense continued to have BACs over .08. Of those cited under the California ZT law, 57.4% had BACs above .08.

Fig. 4.

Percentage of suspension above and below .08 for drivers under age 21 before and after the .01 ZT law

PERCEIVED RISK OF ARREST

The apprehension rate for alcohol-impaired drivers is generally low, estimated to be about 1 in 300 drunk trips (Beitel et al., 1975; Voas & Hause, 1987). Thus, while the police need to improve the apprehension rate of impaired drivers, the most important feature of an enforcement program is how effectively it deters potential offenders from drinking and driving. To achieve this goal, the enforcement effort must raise the potential offenders’ perception of the risk of arrest. Figure 5 gives the responses from approximately 500 drivers under age 21 who were surveyed in the month before the ZT law and at 1 and 2 years after its effective date. The figure shows answers to three questions: (1) How likely do you believe it is that you will be stopped by the police? (2) If stopped, how likely is it that you will be breath tested? (3) If breath tested, how likely is it that you will have your license suspended? Only about one in four believed it was very likely that they would be stopped by the police. This may reflect, in part, that only 30% of the respondents reported driving after drinking in the previous month (see Figure 6). However, the majority of all respondents did believe that, if stopped, they would be breath tested and would probably lose their licenses. Most significant is the fact that their perception of the risk of apprehension did not increase between December 1993, the month before the law, and December 1995, 2 years after the law. In fact, the percentage believing it was very likely that they would be stopped declined.

Fig. 5.

Perceived likelihood of being stopped and, if stopped, breath tested and having license suspended

Fig. 6.

Percentage of respondents under age 21 in three cross-sectional statewide surveys who drank in the previous month and drove after drinking at least once

REPORTED CHANGE IN DRIVING BEHAVIOR

Important to the evaluation of the effectiveness of a DUI enforcement program is evidence of change in the targeted drinking-and-driving behaviors. Figure 6 shows the percentage of California youth in the RDD survey who responded that they drove after drinking in December 1993, before the law, and in December 1994 and 1995, 1 and 2 years after the law. The trend is in the right direction—downward; however, the differences between years are not significant (Chi Square ‘93 vs. ‘95=1.815, p=.263). Similarly, from December 1993 to December 1995, those who said that they had on at least one occasion decided not to drive because they had been drinking (Figure 7) moved in the right direction—upward. The differences, however, were not statistically significant (Chi Square ‘93 vs. ‘95=.325, p=.589).

Fig. 7.

Percentage of respondents under age 21 in three cross-sectional statewide surveys who consumed a drink in the previous month and decided on at least one occasion not to drive after drinking

CRASH INVOLVEMENT

Figure 8, shows a time series from 1988 to 1996 of monthly proportions of California drivers under age 21 involved in fatal crashes who were imputed (Klein, 1986) to have a positive BAC when the crash occurred. The dashed line shows the raw data while the solid line shows the ARIMA-predicted value. Clearly, the proportion of crash-involved youth who had been drinking was lower after October 1993. This was confirmed by a time-series analysis which detected a 20.9% reduction in the proportion of youthful drinking drivers in crashes. This difference is represented by a b coefficient of −0.2342 and standard error of b=0.0987, yielding a t=2.37, with a p=.02.

Fig. 8.

Percentage of alcohol-positive drivers involved in fatal crashes

In Table 2, this 20.9% reduction in the proportion of youthful crash-involved drinking drivers is compared with adult (ages 21 to 34) crash-involved drivers in California and under age-21 drivers from the four comparison states—Texas, Colorado, Nevada, and Wyoming— selected in 1994 for their similarity to California (Martin & Andreasson, 1996). The proportion of drinking drivers in fatal crashes among California adults, ages 21 to 34, decreased by 4.1%, while the proportion of crash-involved drinking drivers in the comparison states decreased 11.6%.

Table 2.

Results of time-series analysis of 3 monthly series of the proportion of drivers in fatal crashes with BACs greater than zero

	b (coef)	raw effect	se (b)	diff (b)	effect	se (diff)	tval	2-tail prob	df
CAL<21	−.2342	−20.9	0987	−.1918	−17.5	.1115	−1.72	.088	100
CAL 21–34	−.0423	−4.1	.0518	−.1111	−10.5	.1253	−.89	.377	99
comp <21	−.1231	−11.6	.0771

DISCUSSION

Given these results, what can we conclude regarding the impact of the California ZT law? Figure 1 provides a framework for the analysis of the law’s impact. The research questions posed for this study follow.

“WERE YOUNG DRIVERS AWARE OF THE LAW?”

Over half of those under age 21 became aware of the ZT law 3 to 6 months before it was implemented when there was some news coverage but no specific public information programs devoted to the legislation (Figure 2). This demonstrates the alertness of this age group to legislation that affects them. On the other hand, there was little or no growth during the first 2 years of the legislation in the number who could cite the correct legal BAC limit or who were aware that they could not drink at all and drive. This suggests that neither the enforcement effort nor the state-sponsored information program, which included PSAs developed by young people themselves, were successful in educating those in the target group who did not pick up the information on their own from news reports or friends. This result is similar to that of Hingson et al. (1991), who reported that 3 years after Maine implemented the first ZT law, only one third of the teenagers queried in a statewide telephone survey were aware of the law and its ramifications (i.e., they could lose their licenses for a year). Thus, it may be typical of this age group that a third to a half may not be able to correctly describe the law. This does not, of course, mean that they are unaware that driving after heavy drinking is illegal.

“DID THE POLICE ENFORCE THE ZT LAW?”

The intent of the California ZT law was to make enforcement easier for the officer by providing for a handheld breath test at the roadside and by making the offense administrative. The police seem to have taken advantage of this easier processing by using the new ZT procedure instead of making a traditional arrest for DUI. This supports the assumption that citing a youth for the ZT offense is an easier procedure; however, it reduced the number of under age 21 drivers charged with DUI following implementation of the ZT law. When both DUI and ZT actions against underage drivers are considered (Figure 3), the downward trend during the 3 years before the law was reversed and replaced by a slight upward trend following implementation of the ZT law. Thus, the law appeared to have produced a small increase in the number of young drinking drivers receiving the license sanction.

While the number of youthful drivers cited at BACs below the adult .08 limit significantly increased, considering the large number of at-risk youths in the 16- to 20-age group** newly covered by this legislation, the increase in low BAC citations was small.

The failure to prosecute young drivers with BACs over .08 removed the possibility of their receiving treatment if they demonstrated a drinking problem. On the other hand, the ZT law suspended licenses for 1 year compared to only 4 months suspension for the .08 DUI offense. Initially, therefore, those charged under the new ZT law received a license sanction three times longer than those charged with DUI. However, this inequity was corrected by the DMV which imposed the same 1-year suspension on DUI offenders who were under 21. This, however, resulted in a period of license suspension for the DUI offense for those under age 21 that was three times as long as for adults over that age.

“DID YOUNG DRIVERS BELIEVE THE LAW WAS BEING ENFORCED?”

It is noteworthy that the target group of drivers believed that, if stopped, they would be breath tested and would probably to lose their licenses (Figure 5). However, only one in four felt that they were likely to be stopped by the police; this percentage appeared to decrease over the next 2 years, suggesting that there was no increase in perceived risk during this period. Since young people became aware of the law during the 6 months before the law became effective (Figure 2), it is possible that there was a rise in perceived risk during that period before the December 1993 telephone survey and January 1994 implementation of the law.

While perceived risk of apprehension is believed to be a necessary element of deterrence, it is obviously not the only mechanism for reducing impaired driving. Grube and Voas (1996) demonstrated a model for prediction of impaired driving among underage drivers that considered four factors: perceived risk from DUI enforcement; perceived risk from crash injury; peer norms; and available alternatives to impaired driving. Of these four, only the perceived risk of enforcement had no relationship to the probability of driving after drinking. It is to be expected that peer attitudes could have an important effect on the California ZT law’s impact on impaired driving.

“DID YOUNG PEOPLE REPORT LESS DRINKING AND DRIVING AFTER THE LAW?”

Responses to the RDD surveys show a decrease in the number of times a youth drove after drinking in the previous month (Figures 6 & 7) and in the 2 years after the ZT law. While these reductions were not statistically significant, they were in the expected direction and are not inconsistent with the hypothesis that the California ZT law has had an impact on underage drinking and driving. It is important to keep in mind that these are self-report measures and could be influenced by a change in the norms regarding the acceptability of drinking and driving. If the ZT law led to the perception that driving after drinking was less acceptable, then respondents might be less willing to admit this behavior.

“DID THE CALIFORNIA ZT LAW REDUCE THE NUMBER OF YOUTHFUL DRIVERS IN FATAL CRASHES WHO HAD BEEN DRINKING?”

A conservative approach to this issue was the time-series analysis of changes in the percentage of all drivers in fatal crashes who had positive BACs. Both the trend shown in Figure 8 and the results of the time-series analysis in Table 2 provide a convincing indication that the percentage of underage drinking drivers was reduced following the law. This supports the position that the law had its intended effect and would translate to 34 fewer drinking drivers in fatal crashes in 1994, 19 fewer in 1995, and 24 fewer in 1996. However, the validity of this conclusion is threatened because we cannot fully reject the alternative hypothesis that there was a general reduction for all age groups of drivers in California, since the percentage of drinking drivers in the 21- to 34-age category also declined. When this 4.1% decline for adult crash-involved drivers was compared to the reduction for underage drivers (Table 2), the p value is .089, which does not quite reach significance. Further, we cannot reject the competing hypothesis that the percentage of under age 21 crash-involved drivers who had been drinking reflects a national trend. The under age 21 drivers in the four comparison states also declined. When this 11.6% reduction is tested against the California under age 21 trend, the p value (.377) does not approach significance (Table 2).

SUMMARY

The data presented are consistent with the hypothesis that the California ZT law has had a small marginal effect on the number of youthful drinking drivers in fatal crashes. The easier citation procedures provided by the ZT law have converted approximately half of the underage DUI arrests to ZT citations and produced only a relatively small increase in overall drinking-driving actions against underage drivers. California youth reported no increase in their perception of risk of arrest or in their knowledge of the law during the 2 years following its implementation. Nevertheless, a reduction in driving after drinking and alcohol-related crashes was indicated. A question remains as to whether a stronger enforcement would have produced more substantial results. Was this a missed opportunity?