This file is also available in Adobe Acrobat PDF format
We all pay the tab for cheap drinks.1
EXCESS DRINKING IS A problem for millions of Americans and their families. It is also a problem for the communities in which they live, degrading public health and safety and ultimately lowering our standard of living. The public response to this problem has varied over time, but always with some mix of two general approaches. On the one hand are efforts, both public and private, to reduce excess drinking directly—education, persuasion, counseling, treatment, sanctions of various sorts. On the other hand are measures to reduce excess drinking by restricting availability or raising the price—licensing, product and sales regulation, liability rules, taxes, partial or complete bans. Both approaches can be effective. Yet during the last half century the public policy “mix” has largely neglected the second approach.
In fact, the post–World War II history has seen a long downward drift in prices (adjusted for inflation) together with a general weakening of other restrictions on availability, all without much discussion of what is being lost. To the extent that alcohol has had any prominence on the policy agenda, it has been primarily in connection with drunk driving. A grassroots movement led by Mothers Against Drunk Driving (MADD) and other such groups has succeeded in promoting tougher laws and stronger enforcement. As one consequence of this movement, Congress adopted a national minimum drinking age of twenty-one in 1984 primarily in the hope of reducing the fatal accident rate for teen drivers. These efforts have been quite successful in reducing drunk driving for both teens and adults. However, the “alcohol problem” is by no means limited to the highways. Our policy portfolio should respond to the full array of harms. The regulation of price and availability—alcohol-control measures—has great albeit neglected potential in that role.
Fortunately there has been enough interest in alcohol-control measures to support a research program by economists, epidemiologists, and other scientists. We now have a quarter century of research results on the causal effects of alcohol-control measures on drinking, abuse, and a wide array of consequences. This research has helped develop the case that the price and availability (both commercial and social) of alcohol affect the amount of alcohol-related harm to society. Although researchers in this area are often labeled “neo-prohibitionists” by industry spokespersons, in fact our work is more closely aligned with the alcohol-beverage-control movement of the 1930s that was the successor to Prohibition. The evidence suggests that higher excise taxes and some additional restrictions on marketing would save lives and pass the cost-benefit test. The goal is not prohibition, but moderation.
A BRIEF HISTORY
What is the problem to which alcohol policy is responding? The predominant answer has changed over the course of American history, and policy choices have followed. The first Congress met at a time (the 1790s) when heavy drinking was the norm. Alexander Hamilton, our first Secretary of the Treasury, advocated a domestic whiskey tax to raise revenue, noting that if it also promoted moderation in drinking, that would be all to the good. During the nineteenth century, alcohol became the matter for a broad moral crusade, which waxed and waned and waxed again. By the 1880s, the Women’s Christian Temperance Union was denouncing alcohol itself (rather than the abuse of alcohol) as the problem, and actually persuaded most of the state legislatures to mandate WCTU-approved textbooks that labeled alcohol a poison and described at length its various deleterious effects on the body. Advocacy groups of that day also denounced the industry that supplied alcohol, saying that it lured working men into spending money on drinking and other vices, depriving their families of desperately needed necessities. The Anti-Saloon League eventually led the way to adoption of the prohibition amendment, implemented in 1920. But this Great Experiment proved a considerable disappointment. After a dozen years of ambivalent enforcement by federal and state governments, coupled with corruption and gang violence, most of the public and particularly the business community became disillusioned. The new voice of reason, led by business tycoon John D. Rockefeller Jr. and other plutocrats, favored repeal of national prohibition, to be replaced by state alcoholic-beverage-control systems and taxes that would, they believed, create an orderly legal market that was conducive to moderate drinking.
Along with repeal, the 1930s also saw the beginnings of an entirely different conception of the alcohol problem (Moore 1990). The founding and great success of Alcoholics Anonymous, coupled with well-publicized research by Yale scientist Edward Jellinek and the efforts of the National Council of Alcoholism, engendered a new “scientific” understanding. The locus of the problem was shifted from alcohol itself to that small fraction of the population who were vulnerable to alcoholism. Most people, it was thought, could drink safely and moderately—for them alcohol was not a problem—but for those relative few who were vulnerable, drinking posed a great risk. Given this new definition of the alcohol problem, the logical policy solution was not to control supply of the substance, but rather to help the unlucky ones who couldn’t handle it. Alcoholism was the problem, alcoholism was a disease, and the right response was to encourage abstinence by those so afflicted. The liquor industry gladly bought into this perspective, since it exonerated their product from blame. Alcoholism research and treatment were established as national health priorities with the creation of the National Institute on Alcoholism and Alcohol Abuse (NIAAA) in 1970.
The alcoholism agenda is worthy but narrow. Programs to identify and treat individuals who have become dependent on alcohol inevitably miss a large portion of youthful abuse, drunk driving, alcohol-fueled domestic violence, and other problems, and bypass important opportunities for the prevention of alcohol dependence and abuse. Even Dr. Jellinek, the godfather of the alcoholism movement, recognized the importance of the social context; he suggested that whether someone with an innate propensity for alcoholism would actually develop the disease depends in part on whether he was living in a wet or dry environment. Thus, an effective prevention program must address the community’s involvement in alcohol, not just the involvement of those members who have become dependent. In support of this perspective, an international group of researchers began writing about alcohol as a public health problem during the 1960s. The leaders included Milton Terris and Robin Room in the United States, Griffith Edwards in Britain, Wolfgang Schmidt and Robert Popham in Ontario, Kettil Bruun and Ole-Jørgen Skog in Scandinavia, and others. A number of these pioneers published a brief, elegant book in 1975 with sponsorship from the World Health Organization, making the case for a population-based, evidence-driven, multifaceted, pragmatic approach (Bruun et al. 1975). In the United States, a similar case was made in the report Alcohol and Public Policy by an expert panel assembled by the National Academy of Sciences under the leadership of Harvard public policy professor Mark H. Moore. The report was subtitled Beyond the Shadow of Prohibition (Moore and Gerstein 1981), but it could just as well have been subtitled Beyond the Shadow of Alcoholism.
A comprehensive community-oriented approach continues to be touted by the public health community, but perhaps with more success in Ontario and northern Europe than in the United States. Here public attention came to focus more narrowly on the particular concern of drunk driving. During the 1980s, MADD emerged as a force to be reckoned with by judges and legislatures, advocating effectively for the sensible position that DUI (driving under the influence) was a crime that should be taken more seriously by law enforcement and the courts. One important sidebar to this effort was a law establishing a national minimum drinking age of twenty-one, which Congress, moved by evidence on drunk driving fatalities caused by teenagers, accomplished in 1984. Since then the alcohol “problem” has been equated in most public discourse with drunk driving and underage drinking; whatever political energy exists behind alcohol control has focused on these issues. The fate of alcohol control contrasts markedly with the recent success of advocates promoting tobacco control and taxation.
In an era where the public is generally hostile to tax increases, the tobacco tax has become fair game for legislatures around the country. In 2002, for instance, New York State raised its tax to $1.50 per pack, and New York City added another $1.50, up from just 8 cents. Since then, the New York street vendors who import their packs (illegally) from lower-tax areas are known as the “$5 men,” in reference to the price they charge; remarkably, it is a multiple of the legal price just a few years earlier. Other jurisdictions have joined in the tax-raising jamboree. Between 2000 and 2005, forty-two states and the District of Columbia enacted sixty-three cigarette-tax increases, with nineteen states exceeding $1 per pack.2 Besides the obvious attraction of bringing in revenue, these taxes are touted as improving the public health by encouraging confirmed smokers to quit or cut back, and by discouraging youths from developing the habit in the first place. These claims do encounter some skepticism. Indeed, it wasn’t so long ago that most savvy commentators insisted that smokers’ addiction would survive any sort of price increase, and adolescents would always find a way to experiment. But the evidence that price matters when it comes to smoking, accumulated over many years of careful study, is compelling. Death and taxes are both inevitable, but in this case there is a trade-off (Phelps 1988).
Meanwhile, alcohol taxes have received much less attention from state legislatures. State legislatures raised alcohol excise taxes just eight times during the period 2002 to 2005, and all of these increases were modest in magnitude.3 Why the difference with tobacco? True, fewer people die from drinking than smoking each year, but the alcohol-related death count is still considerable, on the order of 75,000 to 100,000, as compared with 435,000 deaths due to tobacco (Mokdad, et al. 2004; Midanik et al. 2004).4 Most of these drinking deaths are the result of injury, taking a heavy toll on children and young adults. And unlike most smoking deaths, a sizable share of the drinking deaths involve innocent bystanders—victims of drunk drivers, victims of arguments transformed into deadly assaults due to intoxication, victims who are children abused or neglected by alcoholic parents.
Furthermore, the evidence supporting the public health benefits of increased alcohol taxes is every bit as strong as for cigarette taxes. An increase in the average prices of beer or liquor induced by a tax increase would reduce per capita alcohol consumption, the incidence of alcohol abuse by youths and adults, and the rate of accidents and crimes resulting from abuse. Over the long run, higher alcohol prices would reduce the prevalence of alcoholism and organ damage associated with chronic excess consumption. As in the case of tobacco, a small increase in tax would have a small effect on the public health, a large increase in tax a larger effect. These conclusions are solidly grounded in the evidence compiled by economists and public health researchers.
The divergence between alcohol and tobacco policy is not just for excise taxes, but for other regulations as well. In the late 1990s, at the insistence of the state attorneys general, the tobacco industry agreed to eliminate vending machines and restrict marketing practices in a number of ways. The National Association for Stock Car Auto Racing (NASCAR) Winston Cup series has changed its name and tobacco companies no longer are allowed to sponsor cars and drivers; meanwhile, NASCAR allowed liquor companies to serve as sponsors for the first time in 2004.
Why, then, have legislators pushed up tobacco taxes and tightened controls while neglecting alcohol policy? It is no doubt relevant politically that over twice as many Americans drink as smoke, and the drinkers are as a group more influential politically—they tend to be better educated, richer, and less ambivalent. (Most smokers say that they want to quit. Most drinkers express no such aspiration.) Beyond that political reality, the public health argument for seeking a reduction in drinking has been confused in recent years by evidence that drinking is not all bad. Alcohol appears to be a potent anticholesterol drug that, “taken” in moderation in middle age, may actually extend life. On the other hand, smoking in any amount at any age is harmful. And finally, the public and the legislators they elect are not necessarily reading the latest research on alcohol price effects. Ordinary people (not including economists) may remember that we tried national Prohibition a while back and it proved disastrous. Does it make sense that an increase in taxes and tighter regulations would succeed where a total ban failed?
In what follows, I make an effort to explain the nature of this research evidence in some detail, and in a number of instances provide new results. Because it is not possible to run experiments with alcohol-control measures, making reliable causal inferences is a challenge. Much of the evidence that is most persuasive comes from analyzing the “quasi-experiment” generated by states’ changing their policies. For example, the effect of the minimum drinking age law on highway fatality rates can be estimated by comparing changes in youthful highway fatalities in states that changed their minimum with states that didn’t; between 1970 and 1988 there were scores of such changes, which together provide the basis for reliable inferences about what difference minimum age laws can make.
The goal here is to get beyond intuition. There is a powerful tendency to assess social-science research findings on the basis of prior beliefs, since in a sense everyone’s an expert. If we are already inclined to believe a new research result, then we welcome it as support for our views. If it contradicts our prior beliefs, then we feel free to ignore it; after all, it may be based on faulty methods or poor data, and in any event it will probably be contradicted by another social scientist in the near future. In my experience most people’s intuition tells them that prices and regulations simply don’t matter, or don’t matter much, when it comes to drinking—that drinkers decide whether and how much and in what circumstances to drink for reasons that are scarcely affected by prices and availability. A summary of research results is not going to persuade most people to abandon this intuition, since that would require that they do the hard work of reinterpreting their personal observations. For that reason I aspire to provide enough detail about data and research methods that it cannot be dismissed too easily.
The evidence provides one key argument for putting alcohol control and taxation back on the policy agenda. But the evidence on what works in reducing alcohol-related problems is not a sufficient guide. Policy choices do and should reflect multiple objectives. Improving health and safety are surely on the list, but so are other concerns; our Declaration of Independence sums it up as “life, liberty, and the pursuit of happiness,” rather than as just “life.” America’s history is full of struggle over finding the right balance among conflicting principles when it comes to alcohol policy as in much else. The policy assessment here is explicit about the important values at stake. For adults, the principle that deserves considerable weight in assessments of alcohol control is consumer sovereignty, which is to say respect for individual preferences. Using government powers to compel adults to change their behavior requires strong justification, and usually an adequate justification requires reason to believe that the behavior harms others. Liberty from government interference is also desirable for adolescents, but for that age group it is easier to make the case that they need protection from self-destructive choices. Reasonable perspectives range from the libertarian’s to the public health advocate’s, a spectrum of increasing value placed on life as opposed to liberty. These and other fundamental differences of opinion are not going to be resolved by any amount of research, but it is informative to lay out the arguments. By most definitions of the public interest, alcohol-control measures have an important role to play.
A ROAD MAP
Understanding modern policy debates over alcohol regulation requires some understanding of how those debates have played out in American history. I begin with a chapter on the history of the federal excises and alcohol regulation through Prohibition and Repeal. That whiskey tax promoted by Alexander Hamilton was the first inland revenue measure of our republic, and it was even more contentious then than now, leading to armed rebellion against federal “tyranny.” A failure the first time around, it was reinstated later and was an important source of funding for the Union during the Civil War and even more so thereafter. It was only replaced as the dominant source of domestic revenue when the Sixteenth Amendment established the constitutionality of the federal income tax, making it possible not only to finance World War I but to continue financing the federal government during national Prohibition—which of course ended all tax collections from alcohol. Thus, the Sixteenth Amendment made the Eighteenth (Prohibition) feasible. Prohibition of course did not end drinking, but it did lead to a substantial reduction, especially in the consumption of beer. Although Prohibition was in that sense a success, it was surely a failure politically. There was little enthusiasm for enforcing it, and after a dozen years it was repealed through yet another amendment. What followed was the invention of the modern-day state alcohol-control apparatus, much influenced by the efforts of John D. Rockefeller Jr. to develop a rational policy that would moderate drinking without the corruption and violence of the Prohibition era.
In fact there was little increase in drinking through the 1930s, so the alcohol-control effort appears to have been successful—although the Great Depression of that era gets much of the credit. But with the founding of Alcoholics Anonymous and its development into perhaps the most successful voluntary organization of our time, attention shifted to the science of alcoholism. Instead of the broad population-based prevention approach of taxes and controls, the new focus was identifying and treating alcoholics. Although AA has remained the predominant “treatment” for alcoholism, federal funding for research and medical treatment expanded and became institutionalized with the creation of NIAAA. The research is now producing payoffs in the form of drugs that reduce craving and help people who are inclined to go on the wagon. But no matter how effective, alcoholism treatment now and in the future will have limited scope relative to the full array of alcohol problems. The reason is the “preventive paradox”—that while problems are concentrated among long-time heavy drinkers who are in enough trouble that they might be persuaded (or coerced) into seeking treatment, the bulk of alcohol-related problems are diffused among the much larger group.
Part II begins with a primer on alcohol and drinking, and then goes on to review the evidence on alcohol control. Chapter 4 describes trends and patterns in drinking in America, with some attention to just how those trends and patterns are estimated. Among the possibly surprising facts: a majority of adults either don’t drink at all or drink less than once per month, while the heavy drinkers at the top 10 percent of the distribution account for the bulk of sales and consumption. Greatly over-represented among those heavy drinkers are young adults under age thirty, and especially those of the male persuasion. What the American public is drinking these days is beer—that’s the form in which over half the ethanol is imbibed—while liquor is far less popular than it was a generation ago.
The health and social consequences of drinking are what make it worthy of special attention. Alcohol is a complex drug that produces a variety of effects on the body. The most alluring effect is intoxication, but that can and does engender serious lapses of judgment and coordination which in turn lead to injury, crime, violence, unprotected and unplanned sex, embarrassment—not to mention a hangover the next morning. Over the long term a routine of heavy drinking can produce dependence, organ damage, loss of friends and family, and poverty. On the other hand, there is considerable evidence now (still controversial) that light or moderate drinking in middle age actually improves health and extends life expectancy, due to its anticholesterol effect.
The core findings on the effects of alcohol control are presented in chapters 5 through 8. Alcohol sales, drinking, and abuse respond to prices and other controls. Whereas for most people the decision whether to drink is not much affected by cost or availability—there are a number of other influences—for some individuals prices influence the decision of whether to drink. For a larger group, the decision of how much to drink is influenced by price and availability. Contrary to conventional wisdom, prices even matter for chronic heavy drinkers with a well-developed habit.
The “bottom line” question is whether alcohol-control measures affect the safety and health consequences of drinking. Researchers have extracted compelling evidence from the frequent changes in state laws to establish the importance of minimum drinking age, taxes, vendor liability, and other regulations. The outcomes of interest here include highway safety, sexually transmitted diseases, liver cirrhosis mortality, and violent crime. Also important is worker productivity, a topic that has proven exceptionally difficult to research. But new results presented here indicate that higher alcohol prices increase the earnings of workers, a finding that John D. Rockefeller Jr. would have had no trouble believing.
Part III turns to an assessment of alcohol-control options in light of some important normative considerations. Higher alcohol taxes would improve health and safety and save lives. But they also impinge on the legitimate enjoyment of alcoholic beverages. John Stuart Mill’s “harm principle” provides an important distinction between self-harm and harm to others; his view is that the risks to the drinkers themselves are not sufficient to justify restricting freedom. This argument underpins a focus on “negative externalities,” such as the risks posed by a drunk driver to others who share the road with that driver. But many commentators believe that focus is too narrow—that adults and especially youths are prone to ignore the more remote consequences of their actions, and as a result may benefit from a somewhat restrictive government policy. Modern research on “behavioral economics” offers some support for this perspective.
Mill’s case for limited government appears to have carried the day. There has been a progressive easing of 1930s-era restrictions on alcohol commerce, most obviously as local jurisdictions that took advantage of local-option laws to stay dry in the early years have now elected to become “wet” for the most part. In addition, the eighteen states that chose to institute a state monopoly over part of the alcohol business have tended toward liberalization. There have also been pronounced trends in industry structure that have influenced production and prices. Most notable is the increasing concentration in production—over half of all beer sold in this country, for example, is now manufactured by a single company. The private industry structure to some extent has been influenced by the concern from the saloon era about preventing too close a connection between manufacturer and outlet. The resulting three-tier system of manufacturer, distributor, and retailer remains in place, although it is being challenged in the courts.
The great variety of government regulations could in principle be evaluated one by one, but it seems clear that the most important and comprehensive possibility is raising taxes on alcoholic beverages. Given the social cost of drinking, higher taxes are readily justified. Critics of this approach point out that taxation is a crude measure that does not distinguish between harmless and harmful consumption. While true, the fact is that more refined interventions leave much of the problem untouched; even if we do adopt (and pay for) tougher measures against drunk driving and domestic violence and so forth, there will remain a worthwhile opportunity to do still more through a general reduction in drinking. The fact that some of the reduction will come out of heart-healthy consumption by middle-aged drinkers is of some concern, but is balanced by other effects on health.
Youths present a special sort of challenge for alcohol regulation. The minimum drinking age has beneficial effects but also unintended consequences, beginning with the fact that it turns millions of our youths into criminals. Enforcing this prohibition has become a priority for colleges, the military, and other authorities, with some success. But it is an uphill battle to effectively limit youthful drinking in an environment that accommodates adult drinking so freely. This is further reason to embrace a more comprehensive strategy.
Alcohol is the source of great enjoyment and also great harm. In the face of this complex challenge, we should take advantage of the full array of cost-effective policy options. The neglect of alcohol-control measures by policymakers during the last few decades is a wasted opportunity to reduce that harm while preserving much of the enjoyment, and thereby improve our collective standard of living.
Return to Book Description
File created: 8/7/2007
Questions and comments to: firstname.lastname@example.org
Princeton University Press