Acknowledgements

SEARCH eRiposte!

ALCOHOL AND DRUNKEN DRIVING
An eRiposte Review

In the following, we review some key facts and trends surrounding drunken driving and the pros and cons of possible policy initiatives that could  help reduce this problem - including prohibition, alcohol prices/taxes, reduced media advertising, counter-advertising, anti-drunken-driving laws by state, safety of cars, trucks, SUVs, etc. The data shown in the charts below were obtained from the list of references shown at the bottom of this page. Unless otherwise specified, the charts was plotted by us by pulling together data from the appropriate references.

I. ALCOHOL-RELATED TRAFFIC FATALITIES

We will first review quickly the trends associated with alcohol-related traffic fatalities. (Note that we show only fatalities here and do not show the large number of non-fatal injuries also resulting from drunken driving.) 

Figure 1 simply shows the recent trend in traffic fatalities (total and alcohol-related) - which has been somewhat flat over the past several years. 

Although Figure 1 may suggest we are not doing much to curb traffic fatalities, the fatalities per year must be normalized to an appropriate measure of how much driving is occurring in that year - in this case, the total number of miles driven (shown by the red line in Figure 1). This normalized trend is shown in Figure 2, as the traffic fatalities (total and alcohol-related) in the U.S. per 100 million miles driven. This chart clearly shows a gradual decline in normalized traffic fatalities up until the year 1997, beyond which the curve flattens and even shows a slight uptick in 2000, which suggests that we should review this issue and establish whether additional measures are need to keep the trend moving downwards.

II. KEY STATISTICS RELATED TO ALCOHOL-RELATED ACCIDENTS/FATALITIES

Figure 3 shows a distribution of blood alcohol content (BAC) in drivers who were drinking before a fatal crash. Note that large percentage of people (>75%) who had a large BAC of >= 0.10 g/dl. 

What is interesting, as shown in Figure 4, is that drivers killed in alcohol-related traffic accidents tend to have a disproportionately higher rate of previous DUI/DWI convictions than those who are killed in non-alcohol-related traffic accidents. They also have higher rates of previous speeding convictions and suspensions or revocations. 

Figure 5 shows that alcohol-related traffic accidents have a greater tendency to occur on Fridays and weekends, than during the rest of the week. Figure 6 shows that such fatalities are much more likely to occur at night. Both of these facts are perhaps not surprising.

Figure 7 shows the distribution, by age group, of drivers involved in fatal alcohol-related crashes. Note that as a percentage of total fatalities, drivers in the age group of 21-44 tend to dominate. However, this belies more revealing statistics that examine the representation of age groups as a function of the number of drivers of that age group in the general population. That, is examined in Figures 8 and 9 below.

Figure 8 shows that drivers in the age group 16-39 tend to be disproportionately more likely to be hauled in for DUI compared to the number of members of their age group who are licensed to drive. 

Figure 9 shows a statistic that applies to ALL traffic accidents, but is also qualitatively representative of what is observed in alcohol-related traffic accidents. What the chart shows is that young people and old people tend to have a much higher death rate in motor vehicle accidents than one might expect based on their share of the population. A combination of Figures 8 and 9 indicate that in alcohol-related crashes, young people (especially males) tend to be over-represented in the deaths due to drunken driving. (The highest risk age group for alcohol-related drunken driving appears to be ~16-24+.) 

III. WHAT CAN BE DONE TO REDUCE ALCOHOL-RELATED TRAFFIC ACCIDENTS/FATALITIES?

A. LINK TO ALCOHOL CONSUMPTION

In the following, we first examine how direct the link is between alcohol consumption in the U.S., and alcohol-related traffic fatalities. Figure 10 shows the reported per capita alcohol consumption in the U.S. over the past several decades. (We will return to the caption marked "Drop in 1991" subsequently). Figure 11 shows the break-up of alcohol consumption by type of alcohol - beer, wine and spirits. Note that beer dominates the overall alcohol market. The link between alcohol consumption and alcohol-related traffic fatalities is examined in figures 12 and 13.

Figure 12 shows (not surprisingly) that alcohol-related traffic fatalities per 100 million miles drive, i.e., normalized alcohol-related traffic fatalities, are directly proportional to per capita alcohol consumption. Figure 13, which uses a subset of the year range of Figure 13, uses updated fatality numbers, but shows a similar trend.

Clearly, if alcohol consumption is directly tied to alcohol-related traffic fatalities, as one might expect, then one of the ways in which we can try to reduce the fatalities is by reducing alcohol consumption. We will look into that possibility first.

B. ALCOHOL PROHIBITION?

To see if prohibition might be effective, we can refer back to the United States' own experience in the early part of the 20th century. A survey of the impact of prohibition is available for example in Ref. 16, and we highlight some of their findings below - in Figs. 14-16. 

Figure 14 shows that in the first year of prohibition (1921), alcohol consumption did indeed drop significantly. However, it picked up rapidly in a year or two, thanks largely to illegal alcohol sales. As stated in Ref. 16, (with bold text being our emphasis) "...consumption of alcohol actually rose steadily after an initial drop. Annual per capita consumption had been declining since 1910, reached an all-time low during the depression of 1921, and then began to increase in 1922. Consumption would probably have surpassed pre-Prohibition levels even if Prohibition had not been repealed in 1933.^[6] Illicit production and distribution continued to expand throughout Prohibition despite ever-increasing resources devoted to enforcement.^[7] That pattern of consumption, shown in Figure 1, is to be expected after an entire industry is banned: new entrepreneurs in the underground economy improve techniques and expand output, while consumers begin to realize the folly of the ban. Third, the resources devoted to enforcement of Prohibition increased along with consumption. Heightened enforcement did not curtail consumption. The annual budget of the Bureau of Prohibition went from $4.4 million to $13.4 milion during the 1920s, while Coast Guard spending on Prohibition averaged over $13 million per year.^[8] To those amounts should be added the expenditures of state and local governments..." 

Another significantly negative impact of prohibition emerged. Again, quoting Ref. 16 (bold text is our emphasis): "...The most notable of those consequences has been labeled the "Iron Law of Prohibition" by Richard Cowan.^[9] That law states that the more intense the law enforcement, the more potent the prohibited substance becomes. When drugs or alcoholic beverages are prohibited, they will become more potent, will have greater variability in potency, will be adulterated with unknown or dangerous substances, and will not be produced and consumed under normal market constraints. ^[10] The Iron Law undermines the prohibitionist case and reduces or outweighs the benefits ascribed to a decrease in consumption. Statistics indicate that for a long time Americans spent a falling share of income on alcoholic beverages. They also purchased higher quality brands and weaker types of alcoholic beverages. Before Prohibition, Americans spent roughly equal amounts on beer and spirits.^[11] However, during Prohibition virtually all production, and therefore consumption, was of distilled spirits and fortified wines. Beer became relatively more expensive because of its bulk, and it might have disappeared altogether except for homemade beer and near beer, which could be converted into real beer.^[12] Figure 2 (Figure 15 in this eRiposte webpage - eds.) shows that the underground economy swiftly moved from the production of beer to the production of the more potent form of alcohol, spirits.^[13] Prohibition made it more difficult to supply weaker, bulkier products, such as beer, than stronger, compact products, such as whiskey, because the largest cost of selling an illegal product is avoiding detection.^[14] Therefore, while all alcohol prices rose, the price of whiskey rose more slowly than that of beer.

Fisher used retail alcohol prices to demonstrate that Prohibition was working by raising the price and decreasing the quantity produced. However, his price quotations also revealed that the Iron Law of Prohibition was at work. The price of beer increased by more than 700 percent, and that of brandies increased by 433 percent, but spirit prices in creased by only 270 percent, which led to an absolute increase in the consumption of spirits over pre-Prohibition levels.^[15] A number of observers of Prohibition noted that the potency of alcoholic products rose. Not only did producers and consumers switch to stronger alcoholic beverages (from beer to whiskey), but producers supplied stronger forms of particular beverages, such as fortified wine. The typical beer, wine, or whiskey contained a higher percentage of alcohol by volume during Prohibition than it did before or after..."

Moreover, as Ref. 16 states, "...According to Thomas Coffey, "the death rate from poisoned liquor was appallingly high throughout the country. In 1925 the national toll was 4,154 as compared to 1,064 in 1920. And the increasing number of deaths created a public relations problem for . . . the drys because they weren't exactly accidental."^[18] Will Rogers remarked that "governments used to murder by the bullet only. Now it's by the quart." Patterns of consumption changed during Prohibition. It could be argued that Prohibition increased the demand for alcohol among three groups. It heightened the attractiveness of alcohol to the young by making it a glamour product associated with excitement and intrigue. The high prices and profits during Prohibition enticed sellers to try to market their products to nondrinkers--undoubtedly, with some success. Finally, many old-stock Americans and recent immigrants were unwilling to be told that they could not drink. According to Lee, "Men were drinking defiantly, with a sense of high purpose, a kind of dedicated drinking that you don't see much of today."^[19]...An examination of death rates does reveal a dramatic drop in deaths due to alcoholism and cirrhosis, but the drop occurred during World War I, before enforcement of Prohibition.^[28] The death rate from alcoholism bottomed out just before the enforcement of Prohibition and then returned to pre-World War I levels.^[29]...^"

Among the numerous negative impacts of prohibition, was a significant increase in crime rates. Again, quoting Ref. 16, "...America had experienced a gradual decline in the rate of serious crimes over much of the 19th and early 20th centuries. That trend was unintentionally reversed by the efforts of the Prohibition movement. The homicide rate in large cities increased from 5.6 per 100,000 population during the first decade of the century to 8.4 during the second decade when the Harrison Narcotics Act, a wave of state alcohol prohibitions, and World War I alcohol restrictions were enacted. The homicide rate increased to 10 per 100,000 population during the 1920s, a 78 percent increase over the pre-Prohibition period...According to a study of 30 major U.S. cities, the number of crimes increased 24 percent between 1920 and 1921. The study revealed that during that period more money was spent on police (11.4+ percent) and more people were arrested for violating Prohibition laws (102+ percent). But increased law enforcement efforts did not appear to reduce drinking: arrests for drunkenness and disorderly conduct increased 41 percent, and arrests of drunken drivers increased 81 percent. Among crimes with victims, thefts and burglaries increased 9 percent, while homicides and incidents of assault and battery increased 13 percent.^[42] More crimes were committed because prohibition destroys legal jobs, creates black-market violence, diverts resources from enforcement of other laws, and greatly increases the prices people have to pay for the prohibited goods. Instead of emptying the prisons as its supporters had hoped it would, Prohibition quickly filled the prisons to capacity..."

An example of the crime situation during the prohibition years is shown in Fig. 16 (taken from Ref. 16). "...homicide rate increased from 6 per 100,000 population in the pre-Prohibition period to nearly 10 per 100,000 in 1933. That rising trend was reversed by the repeal of Prohibition in 1933, and the rate continued to decline throughout the 1930s and early 1940s.^[48]_..." 

In essence, as explained in more detail in Ref. 16, Prohibition was remarkably ineffective in achieving its goals. It made a lot of things worse than it made better. The past data suggests it is a blunt, ineffective tool to reduce alcohol consumption.

B. PRICE OF ALCOHOL?

As Ref. 17 points out, a lot of the existing data on the link between alcohol price and consumption shows that youth in particular are influenced by higher alcohol prices - in the direction of reducing their consumption. What is particularly interesting about the data is that young people who drink more and drink frequently are more responsive to price increases. This is good because young people who drink heavily are more likely to be involved in alcohol-related traffic accidents as shown above. An example of a direct link between price increases and consumption is seen in comparing Figures 17 and 10. As a significant increase in alcohol taxes was enacted in 1991, the per capita alcohol consumption that year dropped noticeably.

On the whole, high alcohol taxes or appropriate means to raise alcohol (esp. beer) prices may be a good tool to control and reduce excessive alcohol consumption, while not eliminating the ability to drink alcohol in moderation.

C. IMPACT OF ADVERTISING?

There is a fair amount of work that has focused on the relationship between alcohol advertising and alcohol consumption. Ref. 24 provides a recent survey of this work and tries to provide some direction on this issue. We quote Ref. 24 here (bold text is our emphasis): "...The methods used in this review include a theoretical framework for evaluating the effects of advertising. This theory suggests that the marginal effect of advertising diminishes at high levels of advertising. Many prior empirical studies measured the effect of advertising at high levels of advertising and found no effect. Those studies that measure advertising at lower, more disaggregated levels have found an effect on consumption. Results: The results of this review suggest that advertising does increase consumption. However, advertising cannot be reduced with limited bans, which are likely to result in substitution to other available media. Comprehensive bans on all forms of advertising and promotion can eliminate options for substitution and be potentially more effective in reducing consumption. In addition, there is an increasing body of literature that suggests that alcohol counteradvertising is effective in reducing the alcohol consumption of teenagers and young adults. Conclusions: These findings indicate that increased counteradvertising, rather than new advertising bans, appears to be the better choice for public policy. It is doubtful that the comprehensive advertising bans required to reduce advertising would ever receive much public support. New limited bans on alcohol advertising might also result in less alcohol counteradvertising. An important topic for future research is to identify the counteradvertising themes that are most effective with youth..."  

That said, we also provide below some limited data we were able to find comparing advertising and consumption. Figure 18 shows the "measured media" alcohol advertising dollars spent over several years. This does not include a lot of unconventional and non-measured media advertising - and thus represents a lower limit of alcohol advertising dollars. 

Figure 19 shows that over the time period of 1975-1999, there was not a very clear relationship between measured media alcohol advertising dollars and per capita alcohol consumption in the U.S. However, it is interesting that, as seen in Fig. 20, a direct one-to-one (but not necessarily CAUSAL) relationship that is evident between these two if data from 1985-1999 is examined. It must be noted that a lot of negative publicity against drunken driving (especially thanks to MADD) and associated law changes occurred starting in the early 80s. The impact of the negative publicity and changes in driving laws could surely had some impact on both the tendency to advertise alcohol and to consume it. This makes it difficult to infer that drops in advertising result in  lower alcohol consumption. 

D. IMPACT OF LAWS AGAINST DRUNK DRIVING?

There does not seem to be any doubt that anti-drunk-driving laws played a significant role in lowering drunk driving in the United States. Figure 21 shows for example the estimated lives saved due to minimum drinking age laws over a period of 18 years. (Organizations like MADD have played a pivotal role in driving progress in laws as they relate to drunken driving and it is hard to quantify all the benefits in an easy way.) In the following we will briefly examine how individual states are faring on this issue.

Clearly, there are differences in drunken driving laws state-to-state. For example, figure 22 shows which states have a BAC 0.08 law and which have a BAC 0.10 law (as of 2001).

Just as we need to look at miles traveled per year to normalize traffic fatalities nationwide, it is important to consider miles traveled within each state to assess how the state is doing in terms of controlling traffic fatalities. Consider the example shown in Figure 23 which shows the total (not just alcohol-related) traffic fatalities by state in the year 2000 on the Y-axis and the total traveled highway miles by state in the X-axis. Note the close-to-linear behavior. We highlighted two data points of note - CA which far and away has the largest highway miles traveled has disproportionately lower traffic fatalities. Texas, which had the second highest miles traveled in 2000, has a disproportionately higher number of traffic fatalities (above the dotted line). The spread of points around the "predicted" trend represents how it is possible for some states to do better or worse than the trend. We will now look at alcohol-related traffic fatalities in the states in Figure 24. 

We made an attempt to find as many years of data possible wherein the following information was available by state (a) per capita alcohol consumption, (b) total miles traveled, AND (c) alcohol-related traffic fatalities. Unfortunately, after a fair amount of searching we were only able to find one year for which all this data was available - 1997 - and the data is shown in Figure 24 below. 

There is one very interesting feature of Figure 24, which becomes apparent when it is compared to Figure 12. 
- In Figure 12, the fatalities and alcohol consumption are plotted for the entire United States, for a period of several years, with each data point representing one year. Figure 24 is plotted for each state of the Union, for a fixed year 1997, with each data point representing one state. 
- Figure 12 shows a linear relationship between alcohol-related traffic fatalities per million miles traveled and the per capita alcohol consumption per year. Figure 24 does not (or if it is - at least it is not very clear). The question is why the apparent discrepancy?
We speculate here that there is no real discrepancy. In Figure 12, we are looking at the behavior of the whole country over time, which averages over the behavior of the states each year. Differences between the states get submerged and from year to year, any changes in the behavior of one (or many) states may get dwarfed by the behavior of the others. Thus, over a period of time, the country as a whole may be exhibiting an average behavior which trends well with alcohol consumption as a whole. At a state level though, it is clear that as a function of local populations, vehicles used, social/cultural mores, and laws, in a given year each state may react differently than others. Thus, it is entirely possible in this scenario that for nearly the same per capita alcohol consumption, one state has substantially higher or lower alcohol-related traffic fatalities than another. Indeed, that is evident in the example we have highlighted below, comparing Texas and California. California had an-above average per capita alcohol consumption in 1997, but a below average fatality rate per million miles traveled. Texas has nearly the same per capita alcohol consumption, but a substantially above-average fatality rate. 

What the above chart depicts is a snapshot of the impact of various factors on how individual states compare on controlling alcohol-related traffic accidents/fatalities. MADD puts together yearly rankings of state laws and enforcement as they relate to drunken driving. (It turns out California, a known liberal-leaning state is often the top of the pack, while Texas has often ranked near the bottom.) It may be worthwhile for states that do worse than average on fatalities and accidents to look at what more successful states are doing to learn from them.

E. WHAT ABOUT THE VEHICLES USED BY THE DRIVERS?

As much as one might like to simply focus on alcohol, one must not neglect the medium which facilitates fatal accidents due to drunk drivers - the vehicles they drive. In this context it is worthwhile to examine the relative safety of various classes of vehicles.

Figure 25 is a quick comparison of crash deaths per million vehicles by vehicle class - in 2000. Note that Pickups (2WD and 4WD), SUVs (2WD) and Cargo and Large Passenger Vans are associated with far more traffic causalities (not to mention injuries) than cars as a whole or 4WD SUVs as a whole. Also note that 4WD SUVs are only marginally safer to their occupants than cars as a whole. Dissecting this data further as shown in Figure 26, one notices that large cars are safest to their occupants. What Figures 25 and 26 do not show is the effect of SUVs, pickups and vans on other vehicles on the road. We turn to that next.

As shown in Figure 27, the NHTSA "aggressivity ranking" is much higher for vans, pickups and SUVs than cars. For example, SUVs cause 1.91 deaths in the vehicles they strike per 1000 crashes (as of the reported year - 1999 - in Ref. 19) - which is 3 times more than the deaths caused by a compact car when it strikes another vehicle. The higher risk posed by pickups and SUVs to other cars is also shown in Fig. 28. 

Indeed, in a more recent (MUST-READ) article discussing the book " High and Mighty: SUVs--The World's Most Dangerous Vehicles and How They Got That Way" ( by Keith Bradsher), Gregg Easterbrook points out that (bold text is our emphasis), 
"...The most recent comprehensive study of SUV performance and safety, published last July by the National Research Council (NRC), an affiliate of the National Academy of Sciences, found that occupant deaths were slightly higher in SUVs as a class than in cars as a class. That's right: in an accident, you and your family are more likely to die if you are riding in an SUV rather than in a car. During last year's Senate debate on blocking SUV mileage standards, Senator Barbara Mikulski of Maryland--a state with SUV component plants--declared that "women love their SUVs ... because of their safety." This is political feminism degraded into a sales pitch for bad cars. There is no gainsaying that women, and others, feel secure in SUVs, but the important fact is that this feeling of security is an illusion. The NRC study found that large cars such as the Buick Riviera and mid-size models such as the Toyota Camry are notably safer for the people inside them than are SUVs...
The study also confirmed the popular belief that large cars are safer than compacts. But safety design, not tonnage, is the significant factor. SUVs weigh far more than full-size and mid-size cars, but the latter are safer for occupants because they are designed in accordance with the strictest safety standards, while SUVs and pickups are not. Minivans are large, but do well on safety ratings: their buyers tend to be safety-conscious because the vehicles are intended to haul children, and so automakers, even in Detroit, have applied to minivans the safety-design emphasis missing from SUVs and pickups.
The NRC's findings have been relentlessly misrepresented by Detroit's lobbyists and by the White House, which seeks to justify inaction on SUV fuel waste at the same time that it may ask American soldiers to die in Iraq in part for the politics of oil. Lobbyists and administration officials have repeatedly stated that the NRC study concluded that if the MPG rules for SUVs were made stricter, 2,000 people would die from weight reduction in future SUV designs. Yet the NRC asserted no inescapable relationship between weight and occupant safety..."

As Easterbrook also points out, "...Why did the reduction in auto deaths stop while new safety devices came into use and drunk driving declined? The "kill rate" caused by the presence of ever more SUVs on the road was swamping all safety gains. Bradsher cites an engineering study: "When a car strikes another car in the side, the driver of the struck car is 6.6 times as likely to die as the driver of the striking car. But when an SUV hits a car in the side, the death ratio rises to 30 to 1." Bradsher adds that the "kill rate" for pickups is worst of all, "because pickups are more likely to be driven by reckless young men, and because the bulk of pickups on the road are full-sized models while the bulk of SUVs on the road are still midsized..."

Note, as shown in Fig. 29, SUV sales have been increasing quite rapidly in recent years, becoming a gradually larger portion of the vehicle population (while pickups occupy a fairly steady percentage of the new car market). Thus, it behooves states and the Federal Government to work on tighter safety rules for these vehicles. The more drunk drivers use SUVs, the more likely it is that we will see fatalities increase. The only way to reverse this is to make SUVs substantially safer. 

F. OTHER FACTORS

Other factors like parental views on alcohol, and peer pressure surely play some role in shaping the alcohol-behaviors of individuals, especially youth. The more parents and role models display the need for moderation, the more society as a whole seeks to balance the "coolness" of alcohol portrayed by ads and peer pressure, by the counter of powerful visual images of alcohol-related deaths and the causes for it, the more likely it will be that we can continue to bring the scourge of alcohol-related fatalities down further.

References

1. MADD, "Total Traffic Fatality vs. Alcohol Related Traffic Fatality", http://www.madd.org/stats/0,1056,1298,00.html

2. National Transportation Statistics 2001 Table 1-29 (Bureau of Transportation Statistics), http://www.bts.gov/publications/nts/html/table_01_29.html

3. National Institutes of Health, National Institute on Alcohol Abuse and Alcoholism, http://www.niaaa.nih.gov/databases/consum01.txt

4. National Institutes of Health, National Institute on Alcohol Abuse and Alcoholism, http://www.niaaa.nih.gov/databases/crash02.txt. Note that the crash fatalities statistics here are outdated compared to the MAAD data in Ref. 1. See the bottom of Ref. 1 for details. Note that the qualitative relationship between alcohol-related traffic fatalities per 100 Million miles driven and the alcohol consumption per capita remains the same regardless of whether the data from Ref. 1 or Ref. 4 are used. 

5. Bureau of Justice Statistics, Alcohol and Crime: An Analysis of National Data on the Prevalence of Alcohol Involvement in Crime, 1998,  http://www.ojp.usdoj.gov/bjs/pub/pdf/ac.pdf

6. NHTSA, Traffic Safety Facts 2001: Alcohol, http://www-nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSF2001/2001alcohol.pdf

7. Insurance Institute for Highway Safety (IIHS), Fatality Facts: Alcohol, October 2001, http://www.iihs.org/safety_facts/fatality_facts/alcohol.htm

8. Insurance Institute for Highway Safety (IIHS), Fatality Facts: General, October 2001, http://www.iihs.org/safety_facts/fatality_facts/general.htm

9. Insurance Institute for Highway Safety (IIHS), Fatality Facts: Passenger Vehicles, October 2001, http://www.iihs.org/safety_facts/fatality_facts/passveh.htm

10. Insurance Institute for Highway Safety (IIHS), Fatality Facts: Teenagers, October 2001, http://www.iihs.org/safety_facts/fatality_facts/teens.htm

11. J. P. Nelson, Alcohol Advertising and Advertising Bans: A Survey of Research Methods, Results, and Policy Implications, (2001?), http://econ.la.psu.edu/Research/wpapers/jpn(7-01-2).pdf (or HTML)

12. FTC Reports - Alcohol Appendix B, http://www.ftc.gov/reports/alcohol/appendixb.htm. As stated here, the advertising dollars in measured media may constitute only a fraction of the total advertising expenditures of alcohol producers.

13. NHTSA, Traffic Safety Facts 2000, http://www-nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSFAnn/TSF2000.pdf

14. NHTSA, Traffic Safety Facts 1997, http://www-nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSFAnn/TSF97.pdf

15. T. M. Nephew, et al., NIAAA, Surveillance Report #55, Apparent Per Capita Alcohol Consumption: National, State and Regional Trends, 1977-98, http://www.niaaa.nih.gov/publications/Cons98.pdf

16. M. Thornton, "Policy Analysis: Alcohol Prohibition was a Failure," Cato Institute Policy Analysis No. 157, July 1991, http://www.cato.org/pubs/pas/pa-157.html

17. F. Chaloupka, The Effect of Price on Alcohol Use, Abuse and Consequences, ImpacTeen, Oct. 2002

18. MADD, Estimated Lives saved by Minimum Age Drinking Laws, 2000, http://www.madd.org/stats/0,1056,4562,00.html

19. Insurance Institute for Highway Safety (IIHS), Status Report, Vol. 34, No. 9, Oct 1999, http://www.hwysafety.org/srpdfs/sr3409.pdf

20. Office of Transportation Technologies, Department of Energy, U.S., U.S. Light Vehicle Sales Shares, 1976-1999,   http://www.ott.doe.gov/facts/archives/fotw108.shtml

21. G. Easterbrook, America's Twisted Love Affair with Sociopathic Cars: Axle of Evil, January 2003, http://www.tnr.com/doc.mhtml?i=20030120&s=easterbrook012003&c=1
 (Easterbrook also reports also on: High and Mighty: SUVs--The World's Most Dangerous Vehicles and How They Got That Way, by Keith Bradsher (PublicAffairs, 468 pp., $28); Click here to purchase the book)

22. Insurance Institute for Highway Safety (IIHS), Status Report, Vol. 36, No. 6, June 2001, http://www.iihs.org/srpdfs/sr3606.pdf

23. MADD, Rating the States - Report Card, 2002, (a) Main page: http://www.madd.org/stats/0,1056,5546,00.html, (b) Rating by State: http://www.madd.org/docs/rts2002/rts_section2.pdf

24. H. Saffer, Alcohol Advertising and Youth, Journal of Studies on Alcohol 14 (2002) 173, http://www.collegedrinkingprevention.gov/reports/Journal/saffer.aspx, PDF