National Academies Press: OpenBook

Alcohol and Public Policy: Beyond the Shadow of Prohibition (1981)

Chapter:2. The Nature of Alcohol Problems

« Previous: 1. Simplifying Conceptions of Alcohol Problems
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page16
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page17
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page18
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page19
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page20
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page21
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page22
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page23
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page24
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page25
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page26
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page27
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page28
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page29
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page30
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page31
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page32
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page33
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page34
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page35
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page36
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page37
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page38
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page39
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page40
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page41
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page42
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page43
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page44
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page45
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page46
Suggested Citation:"2. The Nature of Alcohol Problems." National Research Council. 1981. Alcohol and Public Policy: Beyond the Shadow of Prohibition. Washington, DC: The National Academies Press. doi: 10.17226/114.
×
Page47

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

The Nature of Alcohol Problems For most people most of the time, even the immediate effects of drink- ing, to say nothing of the long-term effects, are hardly noticeable. When effects are noted, they vary across a broad range of kinds, degrees, and levels of desirability. As the rich vocabulary of descriptive terms, some of them centuries old, suggests, one can be mellow, tipsy, or pleasantly plastered; stoned, skunked, or sloshed; or dead drunk, wrecked, or under the table. Whatever the degree, however, intoxication alters the mental, behavioral, and physiological capacities of drinkers. An intox- icated person is generally less mentally alert in scanning the environment for hazards, less reliable in interpreting what is observed, and inclined to relax or forget about some things that might be the focus of intense concern in a sober state (Cappell and Herman 1972~. On the behavioral dimension, the drinker is likely to become distractible and clumsy. Of course, these effects of alcohol intoxication are all matters of degree, and the degree depends on various factors. Among these, the amount of alcohol consumption itself can be very important. Given enough alcohol, one can reliably produce a stupor even death—no matter what other factors are operating on the drinker (Poikolainen 1977~. In the more usual case, however, the effects depend on such factors as the spacing of drinks, the drinker's size and weight, how recently he or she has eaten, his or her own hopes and expectations about the effects of drinking, and even the expectations and demands of the people present in addition to the amount consumed. An excited, skinny teenager anticipating a big night with pals can become quite 16

The Nature of Alcohol Problems 17 exhilaratingly "drunk" on a quantity of alcohol that would produce no effect or perhaps a mild relaxation in a heavy, middle-aged man who had just finished a large dinner and had no greater aspiration than to pass the evening quietly. Of course, some people regard even the most temporary departure from sobriety as a significant moral and social problem a willful denial of individual responsibility in a society that both values and depends fundamentally on a sturdy, universal commitment to this standard. But for most people, short-lived infrequent periods of intoxication, to say nothing of drinking events that stop short of noticeable intoxication, do not create or indicate substantial harms (Cahalan et al. 1969~. Many people regard drinking and intoxication as beneficial a harmless, pleas- ant indulgence when they feel beset or entitled or a way of turning an ordinary event into a festive occasion. To create problems for drinkers and others, some special characteristics beyond drinking and drunken- ness must come into play. The idea of an alcohol problem brings some paradigmatic situations to mind: the drunk driver who causes a serious accident by ignoring a road sign or losing control of the car; the domestic fight that, fueled by alcohol and the ready availability of a weapon, flares into a bloody assault; the previously responsible husband, father, and employee whose ability to meet the needs and expectations of his family and employer deteriorates as a result of increasingly frequent and ill-timed drunken- ness and hangovers; the aging heavy drinker whose damaged liver be- comes a chronic health problem and ultimately a cause of death; and perhaps even the rowdy teenagers or the skid row resident whose be- havior offends others' sense of propriety and order. These paradigmatic situations may capture only a portion of an ac- curate accounting of alcohol problems. But, if we can take these as typical, we can make an important observation: few of the bad conse- quences result from drinking only. Instead, they emerge from relatively complex causal systems in which drinking is a major but only a single- contributing factor. Some of the consequences depend on unfortunate combinations of episodic drunkenness with dangerous or demanding environments. Others depend on being drunk frequently enough in sit- uations in which sobriety is expected and demanded that the drinker comes to be regarded as irresponsible and unreliable. In many cases, however, the environments in which the drinking is done as well as the frequency and degree of intoxication play major roles in determining whether bad effects occur. Figure 1 presents a schematic view of the hypothesized relationships. This observation should not be used to minimize the role that drinking

18 l In a) ~ ':n _ ~ ~ ~ ~ ~ ~ . ~ ._ Cam o ~ ~ ~ ; 1 14_ .= o Gm ~ ~ —-I= E ~— ' ~ 1 o Cal - 44_ o C) a: CON a: Ct of so: ._ ._ 3 D ._ a: o ._ I_ Cal - AL U C) _ o 3 ._ Cal ._ Cal TIC Cal U)

The Nature of Alcohol Problems 19 itself can play in creating problems for drinkers and others. Sustained frequent drinking can be directly responsible for some of the more important physiological consequences of drinking (e.g., fatty liver, phys- ical dependence) without regard to circumstance and even without no- table drunkenness. Moreover, for many important consequences (such as drunken driving and drunken falls), some contributing features of the environment (nighttime roads, steep stairways) may be so common and resistant to change that they might just as well be treated as inalterable features of nature, with all the variation in accident experience attributed to drinking. What the observation does suggest is that a simple equivalence of drinking with problems, and more drinking with more serious problems, while often true, can be somewhat misleading. Much seems to depend on how and where alcohol is absorbed in a drinker's life: how much drunkenness is generated and how that drunkenness is fitted into more or less demanding physical and social environments. To be sure, one would expect heavy drinkers to cause and suffer many of the bad con- sequences of drinking. But moderate and light drinkers could also cause and suffer problems if they consumed their smaller quantities of alcohol in unusually dangerous patterns. Moreover, some kinds and degrees of risk confronting drinkers of all kinds depend on certain characteristics of the social and physical environment. These considerations have two crucial implications for policy. One is that alcohol problems may be distributed over a large segment of the drinking population. They may not be only, or even mainly, problems of alcoholics. To the extent that this is true, focusing exclusively or even primarily on the treatment of alcoholics is not sufficient. A second implication is that there may be other avenues open to controlling al- cohol problems besides reducing alcohol consumption. For example, one can think of influencing the drinking practices of individuals so that a given quantity of alcohol could be distributed more safely and appro- priately within a given individual's activities, or one can think of altering the structural characteristics of the environment to make it safer for drunken people. We have now sketched the main outline of a complex conception of alcohol problems. In this view alcohol produces a variety of significant consequences some good and some bad. Moreover, the quantity of alcohol consumed is not the only cause producing the bad consequences; also significant are: (1) "drinking practices" that generate different de- grees and frequencies of drunkenness from given quantities of alcohol and place these episodes of drunkenness in different environments and (2) characteristics of social and physical environments that can make

20 REPORT OF THE PANEL drunkenness more or less risky. Since drunkenness, drinking practices, and environments all play a role in creating and shaping alcohol prob- lems, some of the problems may occur among drinkers who drink, not heavily and continuously, but simply unwisely or unluckily. The remainder of this chapter develops this conception with greater rigor and more detail, then tests the conception against observed facts. We first build an account of the important consequences commonly associated with alcohol and present some information about their mag- nitudes. We then analyze drinking practices: how much Americans drink; how drinking is distributed across the population; where current patterns stand in historical and cross-cultural perspective; and in what contexts drinking (and intoxication) typically occur. Finally, we analyze the relationship between the observed effects and the underlying pattern of drinking practices: specifically, the role that things other than alcohol itself play in producing the varied effects and how alcohol problems are distributed across the population of drinkers. We conclude with the major policy implications of this conception. EFFECTS OF ~ PINKING A complete, evenhanded accounting of the important consequences of current alcohol use is beyond current intellectual resources. We cannot report the simple frequencies of all the important events and conditions in which alcohol is implicated, to say nothing of gauging the infinitely more subtle issue of how much alcohol alone contributes to the character of the events and conditions we observe. But to facilitate the design of an effective policy, it is important that we develop an accounting scheme that strains to be comprehensive and orderly in arraying relevant effects, even if we know at the outset that the data will be incomplete. In particular, the scheme should be constructed in accordance with two . . prlnclp es. The first principle is that in identifying relevant effects of drinking, it is better to err on the side of inclusiveness rather than exclusiveness. Otherwise the scheme loses its value as a device for alerting us to (po- tentially) important aspects of the problem. The accounting scheme should include social as well as economic effects, collective as well as individual ones. It should include effects that are easily quantified as well as those that are not. It should include effects that have already been well measured, those that could in principle be studied but still await careful investigation, and those that cannot in principle be pre- cisely or reliably measured. This principle suggests that we include not only the effects of drinking in the accounting scheme, but also the effects created by the current

The Nature of Alcohol Problems 21 policies and programs designed to manage drinking. Drinking never occurs in a vacuum; it occurs in a social setting in which it is influenced by private attitudes and practices and by governmental policies and programs. Thus, when we observe the effects of current alcohol con- sumption, we are inevitably looking simultaneously at the effects of drinking and of the policies and programs we maintain to shape drinking practices and to cope with whatever problems emerge. This first principle also requires us to recognize good effects of drink- ing as well as bad. That benefit results from drinking is usually conceded even by those who are most appalled by the damages. Over $35 billion was spent in 1980 on alcohol by people who could have chosen to spend the money on better housing, new clothes, roast beef, or vacations. (For that matter, they could have chosen to spend the money on cocaine, marijuana, casino bets, or sexual services.) In describing the nature of alcohol problems the beneficial effects usually get short shrift: analysts are far more industrious in seeking out, enumerating, and marshalling impressive statistical evidence revealing the damages (e.g., Berry and Boland 1977; National Institute on Alcohol Abuse and Alcoholism 1978; Institute of Medicine 1980~. We are not major exceptions to this rule, first, because our concern is primarily with damage reduction and, sec- ond, because our discussion of beneficial effects is constrained by the dearth of available evidence. Still, we insist on the principle that pro- tection of benefits must be weighed if we are to know whether a policy's net effect is favorable (see Walsh and Walsh 1970, Makela and Osterberg 1979~. The second principle is that the accounting scheme should keep track of who is experiencing the varied effects as well as what the various effects are. For our purposes, it is useful to distinguish among the effects on (1) drinkers themselves, (2) people who are intimately connected to (or dependent on) the drinker, and (3) people who are relatively remote from the drinker (i.e., the general population). Such distinctions are important because they provide clues about who might be interested in controlling the drinking. For example, people who take losses from drinking problems might well be motivated to help solve at least the aspect of the problems that affects them. The distinctions are also important, however, because this society attaches different significance to effects that seem consciously chosen by individuals and those that individuals receive involuntarily from others. To be sure, the boundaries between state intervention and in- dividual choices are far from clearly established, and we do not propose to adjudicate them here. Our purpose is to array information so that one can see how such concerns are reflected in alcohol problems. Figure 2 presents the important effects of alcohol use and of relevant

22 in — ,_ — it, _ O .C ' ~ ~ ~ ~ ' ° ' ~ u, ~ O ~ ', " ~ ~ ° '' Z E Z = · 0 · ~ ~ ~ ' ~ , e ~ ~ E O C' ~ ~ O ~ c: ~ ~ m an A t11 uJ ~ O u. I ~ ~ ~ ~ ~n _ lL ~: ~ — ~ uJ ~ _ u' ~ z , 0 , O 0 2 O ~ _ u ~ O ~ O. ~ ~ O [_ ~ .~ ~ ~ s 0 I O E E ~ O ° O ~ E ~ ·~ E ~ E ~ ~ c~ ~ ~ i~ 0 u~ u, ~ ~ ~ ~ w. — Ul C~ LL :1 ~ (n u, ~ o z ~ ~n Y Ct Z o~ _ ~ O: · O: O s~ LL ~ ~ ~ z ~ 0 ~ z ~ "c o (5 uJ ,~ Z I ~ u g_ ~ ~ - ~ · ; ~ -, ' ~ _ u. I u~ cn ~ C: LL llt

The Nature of Alcohol Problems 23 policies for dealing with drinking. In principle, one can track any sig- nificant changes in the size and character of the alcohol problem by measuring these effects over time. Thus, if drinkers started to do ap- preciably less driving while drunk, one would see decreases in the num- ber of fatalities, injuries and, economic losses associated with traffic accidents: these categories (and possibly others) would all move in fa- vorable directions. However, if this change in drivers' habits had resulted from more aggressive enforcement of traffic laws and a major effort to mobilize public opinion, increases in public spending and losses in the social independence of certain drinkers would also be recorded. In prac- tice, it is difficult to find or develop reliable empirical estimates of all the phenomena of interest; the figure provides a basic orientation. (More detailed discussion and review of many of these effects are included in the paper by Gerstein in this volume.) It is important to understand that the detailed entries in each of the categories may respond negatively or positively—or both with respect to drinking. For example, some evidence now suggests that low or mod- erate levels of alcohol intake may be associated with reduced levels of ischemic heart disease and fatal coronary events; other evidence suggests that heavy drinking may be associated with elevated incidence of other cardiovascular diseases such as cardiomyopathy and hypertension (deLint and Schmidt 1976, Hennekins et al. 1978, Klatsky et al. 1978, Regan and Ettinger 1979~. Similarly, the effects of drinking on the ability to develop, sustain, and enjoy personal relationships include some pos- itive contributions of drinking, even though for many people drinking interferes with developing and enjoying relationships, inducing losses in this category. The positive gains may be judged to be small or large relative to the risks, yet the effects listed in Figure 2 do seem to include beneficial effects as well as harms and costs associated with drinking. Which effects of drinking should be the primary focus of our concern is an important but ultimately ambiguous issue. To a degree, this judg- ment can be usefully informed by empirical evidence on the frequencies and magnitudes of the various effects (e.g., the number of people who die of alcohol-induced cirrhosis, the frequency of alcohol-induced crim- inal assaults among strangers in public locations, the number of people who drink so much that their spouses seek divorces, etc.~. Ultimately, however, reliable information of this sort exists for only a few of the effects. In his paper in this volume, Gerstein finds a high degree of confidence in a figure between 20,000 and 25,000 for annual deaths from alcohol-induced cirrhosis of the liver. Reed (in this volume) provides a stable Bayesian estimate of about 12,000 annual traffic fatalities causally attributable to (not just "associated with") drunken driving and

24 REPORT OF THE PANEL somewhat less than $.5 billion in property damage from the same cause. Estimated medical expenditures for sequelae of alcohol abuse and al- coholism have been consistently placed in the neighborhood of $10 billion annually (Berry and Boland 1977), a figure that very roughly equals the annual receipts from state and federal excise taxes on alco- holic beverages (Hymen et al. 1980~. For other categories, the intervals of estimation are too open for numerical summary or do not now exist as operationally defined concepts. Even if this kind of information were available for all the effects of drinking, however, the question of which effects are most important would remain ambiguous. The reason is simply that "social importance" is ultimately a value question about which reasonable people can disagree. Thus, while Figure 2 does not pretend to give a precise, quantitative account of alcohol problems, it prevents one from thinking too narrowly about it. Drinking is not just a health problem or a social problem or an economic problem; it is not even just a problem. Moreover, it involves more than just drinkers and their intimates. Figure 2 alerts the reader to the variety of effects of choosing or having a given policy toward alcohol use. UNDERLYING PATTERNS OF DRINKING PRACTICES Beneath the effects outlined in Figure 2 lie drinking practices. Drinking practices produce these effects, not necessarily directly, but probabil- istically in combination with some important characteristics of the phys- ical and social environments. In order to see how the effects of drinking are associated with (or are the result of) drinking, one must begin with the drinking practices of the general population. To the extent that we expect to affect the current shape of the problem by altering drinking practices, it is important to understand what aspects of drinking behavior should be included in useful descriptions of drinking practices, what the current practices are, and what kinds of factors seem to be shaping them. DEFINITION OF DRINKING PRACTICES At the outset, it is important to be clear about what we mean by drinking practices. This may appear to be largely a semantic issue, but the def- inition is of crucial importance in structuring our conception of the problem and how it might be controlled. Specifically, it emphasizes the complexity of the link between the concepts of "drinking" and "alcohol problems." Moreover, the problem is not simply that characteristics of

The Nature of Alcohol Problems 25 the physical and social environment must operate to shape and create the consequences; choices about how much drinking is done, how much intoxication is generated, and how the periods of intoxication are woven into settings and activities that bring drinkers into contact with dangerous parts of the environment seem to be at least as important in creating risks for drinkers as the general level of hazards in their environment. We include these additional aspects of drinking behavior in the idea of drinking practices. If one could describe how the effects in the various dimensions of Figure 2 were linked to specific characteristics of drinking practices, our ability to design effective alcohol prevention policies would be signifi- cantly enhanced. For one thing, we would then be able to recommend or prescribe certain drinking practices and discourage others. In addi- tion, we would be able to see more precisely how risks and benefits of drinking are distributed across the population of drinkers. Unfortu- nately, at this stage the analytic or empirical work is still insufficient to fully define drinking practices as we want to use the concept. What can be done is to outline the dimensions that are likely to be linked impor- tantly to risks and benefits and report what is known about how the population is distributed across these dimensions. Drinking practices as we define them involve two distinct dimensions: one concerns the description of alcohol consumption and the other con- cerns settings and activities (contexts) commonly associated with drink- ing. With respect to consumption, as we have said, it is useful to dis- tinguish three characteristics. First, many effects depend on the degree of intoxication. At very high levels of blood alcohol content (BAC: above 0.30, absent any other drug effects), a drinker can die of acute alcohol poisoning. At somewhat lower levels of BAC, in the 0.15 to 0.30 range, the drinker does not risk death by poisoning but is likely to be so clumsy and inattentive that even benign physical environments become hazardous. At even lower levels, the drinker may still be suf- ficiently clumsy and inattentive to be unable to cope with moderately taxing physical or social demands; these lower levels generally are those at which psychological mood is most favorably affected (Mello 1972~. Thus, other things being equal, the degree of intoxication itself will probably increase the risks and decrease the benefits a drinker creates for others and himself or herself, beyond the threshold at which effects are noted. Second, certain effects of drinking depend on how often one is drunk (or how much time one spends above a given level of BAC). The chance of being drunk at the wrong place at the wrong time increases as a person is drunk more often. But it also seems plausible that some im-

26 REPORT OF THE PANEL portent consequences of drinking (specifically, bad social consequences in the areas defined in Figure 2) begin to occur only when a person is drunk sufficiently often (or at sufficiently awkward moments) to become "undependable." After all, in social terms, a person can be drunk in- frequently (or even fairly frequently if drunkenness is confined to "ap- propriate" times and places) without being seen as irresponsible and undependable. But if a person is drunk often and allows drunkenness to intrude on times when and in situations in which others need and expect him or her to be sober, he or she will become "undependable" and cause harm to those who count on him or her. The eventual result will be a loss of self-esteem for the drinker as well. Third, it seems fairly certain that the total quantity of alcohol con- sumed has important independent effects on certain attributes. It is fairly well established that the best-known medical risks of alcohol use (e.g., liver damage) are primarily linked to total quantity consumed and not the particular way in which it is consumed. The extent to which alcohol becomes a financial drain on the drinker or his or her dependents also depends in large part on the total quantity consumed. The specification of contexts involves greater complexity. The times and places one chooses for drinking are important in determining the effects of the practice. Drinking while at work is generally much riskier (in its health, social, and economic dimensions) than drinking at home during leisure time (Aarens et al. 1977~. Being drunk in a neighborhood bar where one is known well is probably safer and more satisfying than being drunk in a bar in which strangers fight (Bruun 1969~. Similarly, the activities one combines with drinking can be more or less risky or rewarding. Even a safe home can be a dangerous place if one combines late-night smoking or cleaning rain gutters with drinking. And it is especially dangerous to operate moving equipment when one is drunk. In general, to describe drinking contexts in ways that reveal riskiness, it is necessary to talk about how well a given degree of intoxication is welcomed or accommodated in a given context. It is important for our scheme to distinguish drinking episodes from drinking practices. A drinking episode describes a discrete period of time in which a given degree of intoxication is combined with a given context and associated activities. A drinking practice refers to a char- acteristic clustering of drinking episodes. The distinction is important for the simple reason that many of the effects defined in Figure 2 depend on accumulated experience. Therefore, a drinking practice not only generates the sum of effects associated with each episode, but also has effects that depend on the cumulative result of repeated episodes. To summarize, then, our definition of drinking practices is built on

The Nature of Alcohol Problems 27 the following concepts. A drinking episode combines a degree of alcohol use or intoxication with a given context (a physical and social setting and set of activities and attitudes). Cumulative consumption of alcohol emerges across time as one views the succession of drinking episodes. A drinking practice is a characteristic, more or less durable propensity to combine drinking episodes in certain frequencies and combinations over time. Drinking practices carry with them different probabilities of effects in the various dimensions of Figure 2. There is not a great deal of scientific data about drinking practices in the United States which is to say, the research territory on alcohol use is full of virgin or barely turned soil (Institute of Medicine 1980~. Nonetheless, some indicators are firm enough to elicit conclusions, par- ticularly, conclusions of relevance to our inquiry about policy alterna- tives. This information and summary views on what additional data resources would be most important to develop further are presented below. CONSUMPTION: QUANTITY AND FREQUENCY Current knowledge of drinking practices rests heavily on general sample surveys of households in the general population. Various limitations adhere to the use of such survey data. Since large-scale survey research is a post-World War II phenomenon, detailed quantitative data about drinking practices of the general population are confined to this period. In addition, the precision of the surveys depends fundamentally on how accurate respondents are about their drinking, how willing and able they are to report what they know about their own drinking to an interviewer, how skillfully the survey instrument is designed, and how adequately the sample represents all drinkers (see Midanik 1980~. Since drinking rates reported in such surveys do not account for all of the alcohol sold, it is likely that heavy drinkers are underrepresented and that some respondents tend to underestimate their consumption. Some informa- tion exists on how the tendency to underreport consumption is distrib- uted across drinkers, but the issue is not settled. Despite these diffi- culties, surveys remain the best way of projecting detailed quantitative information about how individuals are distributed across varying pat- terns of alcohol consumption. Quantity Consumed Table 1 summarizes the results of several recent surveys of the general household-based population. Alcohol consumption is expressed in terms

28 REPORT OF THE PANEL TABLE 1 Distribution of Alcohol Consumption by Adults in the United States Estimated Average Quantity of Pure Alcohol Consumed Daily in Ounces Mean Percentage (Standard Deviation) 1971-1976 1979 0.22-0.99 0 ("Abstainers9') 0.01~.21 ("Light Drinkers" up to three drinks weekly) ("Moderate Drinkers" up to two drinks daily) ("Heavy Drinkers") (2 - drinks daily) (4-10 drinks daily) (10 + drinks daily) 1.0+ 1.~2.0 2.01-5.0 5.01 + Number 35 (1.4) 32 (3.4) 22 (3.0) 33 34 24 11 (1.5) 6.5 (1 1) 3.4 (0.7) ** 1.2 (0.5) * 1,758 * 7 national surveys, total number = 12,139, mean number = 1,734, range 1,071-2,510. ** Breakdown not available. Sources: Adapted from Johnson et al. (1977), Clark and Midanik (1980). of the average daily consumption of pure alcohol. A typical drink of alcohol (e.g., a 12-ounce can of beer, a 4-ounce glass of wine, or a 1- ounce shot of distilled liquor) amounts to about half an ounce of pure alcohol, so one can conveniently relate the numbers presented in Table 1 to our natural experience with drinking: an ounce of ethanol a day amounts to about 2 drinks a day; 5 ounces implies about 10 drinks, etc. However, reported consumption from national survey populations amounts to at best two-thirds of all alcohol sold (Room 1971), indicating a certain degree of underrepresentation and underreporting. However, this does not make much difference with respect to the table, since only small changes in the percentage distribution would be sufficient to reflect this "missing" consumption.) We can draw three conclusions from these data. First, the aggregate ~ There are two alternative patterns that seem most likely to account for the missing third. One alternative is that a large fraction in the neighborhood of three-fourths—of the heaviest category of drinkers (5+ ounces daily) are virtually denying any alcohol use; this pattern of denial is supported by extensive clinical observation and a study in Ontario (Popham and Schmidt, 1981~. The other alternative is that a great many people in all consumption ranges are underreporting, but in a sufficiently uniform or predictable way such that the rank ordering of drinkers by quantity is reasonably well preserved (Boland and Roizen 1973, Fitzgerald and Mulford 1978~. For example, if 5 drinkers who consume 25, 16, 9, 4, and 1 gallants) annually were each underreporting by one-third (a simple linear transformation: 17, 11, 6, 2.7, and 0.7), or were even scaling their drinking claims in a more complicated nonlinear way (e.g., to read 14, 11, 7.5, 3.5, and 1), their rank

The Nature of Alcohol Problems 29 pattern of alcohol consumption has been fairly stable over the first half of the 1970s. Second, it is remarkable how much of the population either is completely abstinent or drinks very little. We can calculate that close to two-thirds of the adult population drinks three or fewer drinks per week. Third, the skewness of the distribution implies that much of the consumption of alcohol is concentrated in a small fraction of the drink- ers. The heaviest-drinking 5 percent of the population accounts for roughly 50 percent of total consumption, and the heaviest third accounts for over 95 percent of the total alcohol consumed. It is worth noting that among clinical alcoholic populations, daily consumption of 5 ounces or more is reported by virtually all patients (Schmidt and Popham 1975~. Hence, alcoholics are drawn from the upper 1-4 percent of the con- sumption range in the United States (Armor et al. 1976) but they are not equivalent to this fraction, since quantity of drinking per se is not sufficient qualification for a diagnosis of alcoholism. Frequency of Drunkenness These data on quantity provide important insights about the distribution of cumulative consumption. Moreover, at the extremes, the data allow us to make some informed guesses about the frequency of drunkenness as well as cumulative consumption. The 35 percent of the population that is abstinent (in a given year) is never drunk. The small percentage of the population whose reported average daily consumption is in excess of 4 ounces a day is probably drunk for some period of time most days of the year. Between these extremes, however, it is difficult to guess how often drinkers get drunk, since it depends crucially on how the order could still be preserved. even though the reported consumption would fall short by a third. If one recalculates the consumption aggregates in Table 1 using these three possible patterns (denial, linear, nonlinear), the result is a modest readjustment except in the very highest consumption category, in which a few percentage points represent a large relative error; in every other category the adjustment falls within the range of sampling error. The range of values for the 1971-1976 column of Table 1 and the three transformations are as follows: Daily Ounces of Alcohol o 0.01-0.21 0.22-0.99 1.0-2.0 2.01-5.0 5.0+ Percentage (Range) 33-35 30-32 20-22 5.9-8.5 3.4-4.7 1.2-4.3 (100)

30 REPORT OF THE PANEL TABLE 2 Average Daily Consumption and Number of "Drunk Days" Per Year in a Sample of Air Force Men Annual "Drunk Average Daily Consumption (Ounces of Ethanol) Days" (More Than 8 Dr~nks/Day) 0 0.1-1.0 1.1-2.0 2.1-3.0 3.1-5.0 5.0 Interval Midpoint O (0) 100% 40.0% 7.4~o 1.7% 2.4% Do 1-11 (5) 05,0 46.0~o 26.0~o 14.0~o 2.4~o 2.7~o 12-23 (18) 0% 6.1~o 8.1~o 5.2~o 7.1~o 2.7~o 24~7 (36) 0% 6.6% 26.0% 21.0% 9.5% 5.4% 4~77 (63) 0% 0.5~o 11.0% 8.6~o 2.4% 7.7% 7~181 (130) 0% 0.2% 21.0% 38.0~o 33.0% 16.0% 182 (273) 0% 0% 0% 12.0% 43.0% 70.0% Note: The term "drunk day" is defined here as more than 8 drinks/day. If one spaced these drinks evenly over a 16-hour waking period (e.g., a drink every two hours from 6 a.m. to 10 p.m.), one might never be visibly drunk, since the usual healthy adult's liver can detoxify alcohol fast enough to completely neutralize each drink before the next arrives. A more conservative term than "drunk day" might be "intensive drinking day." However, the evidence offered by Polich and Orvis indicates that little drinking occurred during the day (especially on weekends). The description "drunk for at least a few hours out of a day" is a fairly accurate interpretation of the 8-drink-daily category: hence, "drunk day." Source: Adapted from Polich and Orvis (1979~. light, moderate, and heavy drinkers distribute their consumption over time. Since many important consequences of drinking are related to frequency of drunkenness rather than cumulative quantity consumed, it is useful to know how the population is distributed with respect to drunkenness as well as total consumption. A recent survey of drinking practices among U.S. Air Force personnel (Polich and Orvis 1979) provides some very useful data on this issue. While the Air Force population cannot be taken as representative of the U.S. population as a whole, it is quite representative of the young employed male population. For this group, these survey results provide a good estimate of the relationship between the overall quantity of consumption (measured in terms of average daily ounces) and the fre- quency of drunkenness (measured in terms of the number of days in which more than eight drinks were consumed). It is notable that this survey, taken in a very "wet" drinking climate and based on a complete enumeration of personnel rather than a household census base, appeared to account for about 85 percent of alcohol consumed. Table 2 presents the relevant data. As one would expect, the frequency of drunkenness is highly corre- lated with total consumption: the more one drinks overall, the more

The Nature of Alcohol Problems 240 - 4 - ._ o . _ 200 it' 1 60 LU It 1 20 En Y 80 a: o 4o CC UJ m in O o J - 2.0 4.0 6.0 8.0 10.0 AVERAGE DAI LY CONSUMPTION FIGURE 3 Relationship between average daily consumption and "drunk days" in a sample of Air Force men. For an explanation of the term "drunk day," see the note to Table 2. Source: Based on data from Polich and Orvis (1979~. 31 days one is likely to spend being drunk for at least a part of the day (see note to Table 2~. A close inspection of these data reveals a more interesting and less obvious result. It appears that those who drink an average of 1.0-3.0 ounces per day have proportionately more "drunk days" than do very light drinkers or very heavy drinkers. Figure 3 il- lustrates this phenomenon by plotting the average number of drunk days per year against average daily consumption. The slope of this curve represents the "efficiency" with which cumulative consumption is con- verted to drunk days: the steeper the slope, the more drunk days are wrung out of a given level of consumption. Inspection suggests that this slope is at a maximum between 1.0 and 3.0 ounces of pure ethanol per day. The data suggest that moderate to heavy drinkers in this population tend to mass their drinking in binges rather than spacing their drinking evenly over time.

32 REPORT OF THE PANEL The degree to which drinking is massed rather than spaced by these drinkers is more typical of young men (not only those in the military) than characteristic of the rest of the population (Cahalan and Cisin 1968~. Still, this pattern reminds us that drunkenness is different from total consumption and is not necessarily proportional to total consump- tion. To the extent that social consequences of drinking are associated with drunkenness rather than with total consumption and to the extent that the many moderate to heavy (but hardly alcoholic) drinkers get drunk out of proportion to their consumption, these drinkers will figure more prominently in causing and experiencing the social consequences of alcohol than one might expect. Table 3 presents another useful calculation: the fraction of the total number of drunk days contributed by the different consuming groups. The very heavy drinkers (4 percent of the sample) account for about one-quarter of the total number of drunk days (26.1 percent). This proportion of drunk days is very large relative to their proportion in the community, of course. But if drunk days expose anyone to various risks, then three-fourths of the risks are borne by relatively unremarkable drinkers. In fact, even those consuming at the low rate of between 0.1 and 1.0 ounces per day generate about one-tenth of the total drunk days—more than a negligible contribution. There are two reasons why the lighter-drinking categories have an appreciable proportion of drunk days. First, there are so many more people in the lighter-drinking categories than there are in the very heavy- drinking categories that even small individual contributions to total drun- kenness will cause the lighter-drinking categories to rival the overall contributions of the smaller, very heavy-drinking group. Second, as we discussed above, the moderate to heavy drinkers are more "efficient" producers of drunk days than the very heavy drinkers. Both the number of these drinkers and their relative tendency to binge result in their contributing an unexpectedly large share of the total drunkenness ex- perienced by this population. In summary, the aggregate pattern of consumption in the United States has been relatively stable in the 1970s (Clark and Midanik 1980~. Approximately one-third of the adult population is abstinent. Another third averages no more (most average much less) than 3 drinks per week and seldom gets drunk. The next fifth or so of the population averages about 2 drinks a day, and of this fifth about 1 in 12 gets drunk (8 drinks or more in a day) more than 6 times a year. The next tenth or so of the population averages 3 drinks or more a day and gets drunk more than once a week. The remaining 1-5 percent of the population (a more exact percentage is difficult to ascertain) averages 10 drinks or more per day

The Nature of Alcohol Problems TABLE 3 "Drunk Days" Within Different Levels of Average Consumption in a Sample of Air Force Men Annual Number of "Drunk Days"a Evaluated at Midpoint 33 Average Daily Consumption of Ethanol (Midpoint) 0.0 O.S 1.5 2.5 4.0 8.0 O O O O O O O (N) (477) (710) (34) (3) (3) (1) 5 0 4,075 550 130 15 15 (N) (0) (815) (110) (26) (3) (3) 18 0 1,944 612 162 162 36 (N) (0) (108) (34) (9) (9) (2) 36 0 4,212 3,888 1,332 432 180 (N) (0) (117) (108) (37) (12) (5) 63 0 567 2,898 945 189 126 (N) (0) (9) (46) (15) (3) (2) 130 0 390 11,570 8,840 5,590 2,340 (N) (0) (3) (89) (68) (43) (18) 273 0 0 0 6,006 15,015 21,840 (N) (0) (0) (0) (22) (55) (80) Total People (3,078) 477 1,761 421 180 128 111 Total Drunk Days (94,061) 0 11,188 19,518 17,415 21,403 24,537 Percentage 11.9 20.7 18.5 22.7 26.1 a For an explanation of the term "drunk day," see the note to Table 2. Source: Adapted from Polich and Orvis (1979). and spends most days of the year above the level of 8 drinks. If the bad consequences of alcohol consumption began when one reached a yearly average of more than 8 drinks per day, we could calculate that virtually all of these bad effects would be concentrated in the worst 5 percent (or less) of the drinkers. On the other hand, if a large share of these effects are linked to days of actual drunkenness rather than the average quantity of consumption, the adverse consequences would be distributed more evenly across the population of drinkers. The 5 percent of the heaviest drinkers would account for only about 25 percent of the drunk days. The people averaging 2 drinks or more a day would account for 60 percent of the drunk days.

34 Historical and Cross-Cultural Comparisons REPORT OF THE PANEL While comparable information about the aggregate pattern of con- sumption for other times and other countries is scarce, we can say something about how typical our current position is by comparing our current average (per-capita) consumption with that of other times and other countries. While per-capita drinking statistics cannot fully substi- tute for precise information about the distribution of consumption prac- tices, they have been shown to capture more information about the distribution of drinking than one might expect. In particular, the pro- portion of very heavy drinkers (consuming 5 ounces or more of pure alcohol daily) varies roughly as the square of average consumption in the adult population as a whole (Bruun et al. 1975~. On this basis, the current position of the United States with respect to quantity of drinking is unremarkable in both historic and cross-cultural terms. Figure 4 presents a record of the 150-year trend of average annual per-capita consumption in the United States. Generalizing broadly, there are four periods. First, through 1850, the trend was downward from the high volume (6-7 gallons per capita) of drinking characteristic of revolutionary America. Second, from 1850 through 1915, the level varied between about 1.75 and 2.75 gallons, increasing to the high point in the decade just prior to World War I. Third, between the world wars, the dislocations associated with the prohibition laws and then the Great Depression served to lower the quantity to well below the level in the Victorian era. Finally, after 1945 the prewar increase was repeated, to about 2.75 gallons. For comparative purposes, Table 4 supplies a two-point time series for the postwar era for a number of industrial (mainly European) coun- tries. While some caution must be used in comparing consumption sta- tistics among countries, we may safely make two general observations. First, the United States is relatively unremarkable in its rate of mean alcohol use in this period: it is neither especially "dry" nor especially "wet." Second, the recent rise in U.S. consumption is also unremark- able: with little exception, increases have been common throughout the industrialized world and, for that matter, virtually every other country for which statistics have been compiled (Sulkunen 1976, Mosher 1979~. In both the static and comparative perspectives, the United States is in the middle of the pack formed by the array of mean consumption figures. Aggregate Patterns of Drinking Contexts In addition to facts about quantities and rates of alcohol consumption, we would like to know something about the contexts in which Americans

The Nature of Alcohol Problems 4.0—t' 3.6 - \ 3.2 - \ 2.8 _ \ `~' 2.4 _ of o J J 2.0 1.6 _ 1.2 _ 0.8 0.4 35 \ o 1 A / \ - Consumption Trend Cirrhosis Death Trend (1910-1945) - Alcoholism Death Trend (1910-1945) 1 1 1 1 1 1 'ail. 1 1 1 1 1 1 1 1 1920 1940 1960 1980 1840 1860 1880 1900 YEAR FIGURE 4 Time trends of U.S. consumption of absolute alcohol, cirrhosis death rate, and alcoholism death rate in the drinking age population, 1830-1977. Note: Data points are less dense and less reliable prior to 1900. All estimates adjusted to reflect the population 15 years and older. Sources: Consumption statistics from Hyman et al. (1980~; alcoholism death trend from Jellinek (1947-1948~; cirrhosis death trend from Bureau of the Census (1975~. In estimating Prohibition-era consumption Jellinek favored use of cirrhosis-based estimators; Warburton (1932) used principally alcoholism death rates and production estimates. do their drinking and how they have been changing over time. Do Americans do most of their drinking at home in the company of their families, with friends at private parties, or in bars and restaurants as part of a weekly (or daily) celebration? Do many people drink during working hours and has this changed over time? Similarly, how often do people combine drinking with driving, and has this been affected by the campaigns against drinking and driving? We have already stressed one reason for our interest in these issues: drinking contexts have a very important effect on the kinds of risks and benefits drinkers experience as a result of their drinking. It is worth noting that contexts have this effect on risks through two different mech- anisms. The context chosen by the drinker (or thrust on him or her by peers) defines the social and physical environment he or she will en-

36 REPORT OF THE PANEL TABLE 4 Changes in Apparent Consumption of Absolute Alcohol in 20 Countries by the Population Aged 15 Years and Overa Total Absolute Alcohol Total (Gallons Per Absolute Percentage Country Year Capita) Year Alcohol of Change France 1955 6.8 1972 5.9 - 13 West Germany 1959 2.2 1974 3.9 + 80 Belgium 1958 2.2 1973 3.8 + 72 Switzerland 195~1960 3.0 1971-1973 3.7 + 23 Italy 1958 2.8 1973 3.6 + 28 Australia 1957-1958 2.4 1972-1973 3.5 + 44 New Zealand 1957 2.5 1972 3.3 + 31 Denmark 1958 1.4 1973 2.9 + 112 Canada 1958 1.9 1974 2.8 + 51 Netherlands 1958 1.8 1974 2.8 + 53 USA 1960 2.1 1975 2.8 + 34 United Kingdom 1959 1.6 1974 2.7 + 71 USSR 1957 1.3 1972 2.3 + 83 Ireland 1959 1.1 1973 2.2 + 100 Poland 1959 1.7 1974 2.1 + 22 Finland 1957 0.8 1973 2.0 . + 157 Sweden 1958 1.3 1973 1.8 + 43 Norway 1959 0.9 1974 1.5 + 68 Iceland 1956 0.7 1973 1.3 + 84 Israel 1959 0.1 1974 0.9 - 12 a Arranged in order of consumption in most recent year available. Compared time is the available date nearest to 15 years prior in each country. See Gerstein (in this volume) on problems of comparisons between countries. Here even a 1-year difference in date could substantially alter any percentage change. The country in which consumption was highest in 1974, Portugal, is absent due to missing data for the earlier period. Source: Adapted from Keller and Gurioli (1976~. counter while intoxicated and thus the constellation of environmental risks. This is the mechanism we have so far emphasized. But another mechanism is at work as well. Substantial empirical evidence exists showing that the context exerts a separate, independent effect on the quantity and rate of alcohol consumption. Drinkers take cues from the people around them about how much and how fast they should drink. Sometimes (as is often the case when one drinks at home in the company of relatives) these influences tend to moderate drinking. Other times (as, for example, when one is out on the town with friends) the influences work to increase drinking. To the extent that the context shapes patterns of consumption and to the extent that the risks associated with given drinking episodes and practices depend on the combination of degrees

The Nature of Alcohol Problems 37 of intoxication with given features of the environment, the context in which drinkers do their drinking can shape the risks confronting drinkers by affecting amounts of consumption as well as by differentially exposing them to environmental hazards. We are interested in drinking contexts because they shape some of the effects associated with drinking. Despite the potential interest of the subject and despite frequent surveys and observational studies that capture some qualitative notions about U.S. drinking practices, we have surprisingly little quantitative information about how periods of drinking are distributed among given contexts and virtually nothing systematic on how the contexts are com- bined within a given individual's drinking history. Investigations of drinking contexts have been of two types. One type has focused on the micro question of whether the local contingencies of drinking contexts can shape short-term quantities and rates of alcohol consumption; the answer to this question has been a definite yes. Rates of drinking can be influenced in given contexts by such factors as price, the example of other drinkers in the same context, the relationship of the drinker to others in the same context, and the setting of the drinking (Babor et al. 1978, Harford 1979, Mass Observation 1970~. The other type has sought to establish differences in the contexts of drinkers who drank different amounts and suffered different kinds and degrees of problems (Cahalan and Room 1974~. Again, a fairly clear answer has emerged: heavier drinkers (and those drinkers with problems) use a much broader array of drinking contexts than light drinkers and are, in particular, more frequent users of bars and restaurants as drinking places (Clark 1977~. Such studies have their uses, but they fail to provide any clear answer to the questions of how different kinds of risks and benefits attach to different kinds of contexts and how the contexts of drinking episodes have been changing in quantitive density over time. All we have is a tantalizing suggestion that the context seems to shape alcohol con- sumption and that heavy drinkers seem to frequent on-premise estab- lishments. From our perspective, two important facts about drinking contexts are now available. The first is that sales for off-premise consumption now account for approximately three-quarters of total consumption- up from less than half at the end of World War II. This predominance of off-premise consumption shows up in survey results indicating that home is the most preferred drinking site and that drinking episodes at home or at friends' homes are much more common than drinking in bars and restaurants (Harford 1979~. It is hard to know whether this is good or bad news, however, from the point of view of reducing the bad consequences of drinking. To the extent that this trend takes drinking contexts out of the direct control of state-regulated enterprises, we may

38 REPORT OF THE PANEL TABLE 5 Where, When, and With Whom People Drink (Drinking Contexts), by Age of Drinker Weekdays Private Public ~ Relatives Friends Relatives Friends Weekends Private Public Relatives Friends Relatives Friends (Total Total* Events) 1~25 11% 14% 2% 16°/c ll~ic 20°/c 4°/c 23~c 100 °%c 2 107 2~35 17~o 12% 4% 13°fc 18°Xc 16°%c 7% 13°/c 100% 2. 172 3~48 28% 6% 3% 10°%c 25°/c 10°/c 5% 13% 100 °%c 2,513 49 62 33% 5% who 5°/c 29°/c 9% 6°Xc 6°Xc 100°%c 1,414 >63 43~o to 2% 3% 33°Xc 8% 2°/c 4% 100 °Xc 1.013 *Addition to less or more than lOO percent is due to rounding error. Solitary drinking episodes were quite rare and were coded under "friends'' (none present). Source: Adapted from Harford (1979, p. 171). judge that public capacity to shape drinking contexts is being reduced. On the other hand, privately created drinking contexts may tend to moderate drinking and insulate the drinker from certain dangerous con- sequences of drunkenness more effectively than publicly sanctioned drinking opportunities. To the extent that this is true, the shift toward off-premise sales and consumption is a good sign, but not enough is now known about this change (Partanen 1975~. The second fact is that preferred drinking contexts change dramati- cally with age (Table 5~. Drinking for all ages is heavily concentrated on weekends~about half of all drinking episodes are crammed into the relatively short weekend period—but this concentration tends to dimin- ish with age. In addition, young people tend to prefer public locations with friends, while older people do more of their drinking in private with relatives. Again, because we do not know exactly how these dif- ferent contexts are linked to different risks, these findings are difficult to interpret. One might hypothesize that this aggregate pattern makes young people more vulnerable to risks of auto accidents, fights, and other disturbances linked to public intoxication. This is speculation, however; the data presented in Table 5 are useful primarily to indicate the kind of information and data that must be developed to discover current aggregate patterns of drinking contexts and to explore the re- lationship among them and the risks and benefits of drinking. We think a major opportunity exists to do additional research on drinking contexts. In fact, we think much data could be reanalyzed to discover important facts about aggregate patterns of drinking contexts. The crucial questions involve how the contexts are linked to certain kinds of risks and whether the contexts are changing in more dangerous or safer directions.

The Nature of Alcohol Problems Drinking Status in College Abstinent M in College F Non-Problem M in College F Problem M in College F 39 Drinking Status 25 Years Later ~\\\\\\\\\\\~38~\\\\\\~ ~.i.~ 5 1 ~3 16~ ~~ :~: 73 ~ ~~ 11 <^,5~_~ 7 >8~ ~~ ~ ~ 22 2~ ... :. 17 ~ thy % Abstinent A\\\ 25 Years Later I..:. ~ % Non-Problem I. . ~ 25 Years Later ~ I % Problem 25 Years Later FIGURE 5 Reported drinking status of men and women in college and 25 years later. Source: Adapted from Fillmore et al. (1977~. STABILITY OF INDIVIDUAL DRINKING PRACTICES A crucial question affecting the feasibility of controlling the alcohol problem by shaping drinking practices is simply how stable these prac- tices are. It is common to think of drinking practices, particularly those that cause problems, as relatively deeply rooted and difficult to alter. There is evidence, however, that drinking practices and associated prob- lems are not so constant. A major source of evidence on changing drinking practices comes from a study in which individuals surveyed about their drinking practices in college during 1949-1952 were resurv- eyed 20 and 25 years later (Fillmore 1974, 1975; Fillmore et al. 1977~. Figure 5 presents one set of the results of the follow-up. The figure indicates a clear but modest relationship between drinking problems in college and drinking practices 25 years later. Certainly, the probability of later problems is considerably higher, in terms of relative risk? for problem drinkers in college than for others. But the vast majority of problem drinkers in college were not problem drinkers 25 years later, just as the majority of those who were abstinent in college were not problem drinkers 25 years later. Thus, while the drinking patterns of early adults do influence subsequent patterns, the influence is far from decisive. Such differences across time are apparent during shorter pe-

40 REPORT OF THE PANEL riods of a few years as well (Clark and Cahalan 1976~. Cahalan et al. (1969) found that retrospective reports of changes in consumption were common in an adult national sample, especially among men. Two-thirds of all male drinkers recalled broad variations in their quantity of con- sumption at different periods. The least lifetime variation was reported by abstinent women, 80 percent of whom were consistently so. Evidence about the volatility of drinking practices of clinical problem drinkers comes from follow-up studies of drinkers who have been in treatment. In a 4-year follow-up study, Polich et al. (1980, p.168) found that "although there is frequent improvement tin terms of drinking practices), there is also frequent relapse and much instability." Appar- ently, the drinking practices of even alcoholic drinkers are not etched permanently into their lives: there are good periods and bad periods lasting for several months or years. Moreover, the changes do not bear a necessary relationship to treatment. The fact of impermanence in drinking practices has several important policy implications. It suggests that effective influence might be brought to bear on current drinking practices: they are not so deeply embedded in personal characteristics that it is fruitless to think of trying to change them. On the other hand, it suggests that there are few permanent victories in the efforts to shape drinking practices. Nor can people with serious problems be treated then discharged with the expectation that the drinking (or nondrinking) practices of more than a few will retain for long or without further ado the pristine quality they may have had at the end of treatment. Apparently, the factors that shape drinking practices are varied, transient to some degree, and under the control of many different parts of society. They are neither permanent features of individual character nor monopolized by government agencies. CAUSES OF DRINKING PRACTICES To this point we have treated drinking practices descriptively, asking what they are rather than posing the issue in explanatory terms: i.e., why do people drink and why do they drink in these ways? The only certain answer to these latter questions is that no single explanation is adequate. The complexity of causes of drinking is indicated by Sulku- nen's (1976, 1978) discussions of the three "use values" of alcohol: as a food and beverage, an intoxicant, and a symbol of sociality. As a food, alcohol is itself a carbohydrate; most alcoholic beverages contain additional nutrients; these beverages are generally quite resist- ant to spoilage or infestation; and they are free of most water-borne diseases. Thus, fermented products have a long history as useful, durable basic commodities. People make and use them for reasons as diverse

The Nature of Alcohol Problems 41 as biological thirst and hunger, personal taste, and cuisinary custom (Gastineau et al. 1979~. In addition, however, alcohol is an intoxicant, and its multiple effects on psychological mood and alertness are well known. Alcohol is var- iously a stimulant and depressant, euphorigen and soporific, irritant and anxiety reducer~epending on a myriad of factors such as dose and schedule of use, individual metabolism, personality factors, and situa- tion. Alcohol, like other intoxicants, can produce such dependency phe- nomena as persistent search behavior, withdrawal, relapse, and loss of control, although these terms are controversial (see the forthcoming report on Commonalities in Substance Abuse and Habitual Behavior from the Committee on Substance Abuse and Habitual Behavior, Na- tional Research Council, National Academy of Sciences). Finally, alcohol carries symbolic values that are independent of its nutritional value or its intoxicating effects" for example, the religious symbolism of communion and sabbath wines; the celebration that calls for champagne; the slave-trade heritage of "demon rum"; and the laugh- ter in numberless jokes about drunkenness. The drinking practices we observe in society are shaped by a variety of factors- those noted above and others acting simultaneously. Some factors are individually based (e.g., genetic proclivities, psychological and developmental needs); some are rooted in intimate, informal social processes (e.g., family, ethnic, or religious traditions; the practices of one's spouse; the rituals of one's working companions); some are based on the marketing efforts of alcohol producers; and some are managed explicitly by government (e.g., taxes on alcohol, restrictions on avail- ability, laws regulating drinking conduct, and a variety of educational messages about drinking). We assume that these factors operate with varying degrees of force and varying degrees of persistence. Some factors (such as general personality traits and restrictions on availability) may operate continuously, with diffuse rather than precise impact on drinking practices. Other factors (events such as an illness, a divorce, or a firing) may operate with great force in the short run, with relatively transient effect. Still others (such as strong biological aversions or family drinking traditions) may be both powerful and persistent. Unfortunately, little is known about the relative power of the different kinds of factors shaping drinking practices, and discussion of such factors draws heavily on speculative assumptions (Institute of Medicine 1980~. Moreover, ex- pert views on the most important factors seem to depend more on disciplinary training than on persuasive facts currently in hand. Alcohol policy would benefit from the identification of those factors that are powerful but can be efficiently influenced by government a difficult combination to locate.

42 REPORT OF THE PANEL Although we have stressed observations of the impermanence of drinking practices over individual drinkers' lifetimes, it is clear that in the short run, at least, drinking practices do have some inertia. Past practices exercise important influence over future practices, particularly in the short run. This is true in part because some of the factors shaping practices are powerful, persistent, and not easy to manipulate or escape, and in part because people generally tend to maintain in their lives habits and routines that work. In general, then, we hypothesize that the longer a given practice has continued, the more meaning the individual attaches to that practice; and the more stable the social environment surrounding the individual is, the more stable the drinking practice will be. On the other hand, we assume that drinking practices are to some degree modifiable by changes in the experiences of the drinker. Often the experiences that lead to modification of drinking practices are the result of small-scale, private initiatives: e.g., a husband begins encour- aging his wife to drink more often with him in the evening; an employer complains about tardiness and threatens firing; a favorite neighborhood tavern closes. At other times, the experiences that cause individuals to change can be the result of broader efforts of social management (e.g., a persistent and persuasive mass media campaign designed to make people more self-conscious about drunken driving, a tax increase on alcohol, or a reduction in availability). In any event, we assume that drinking practices can change first as immediate circumstances change, then as persisting new conditions force continued adjustments and the development of new practices supported in part by the surrounding circumstances and in part by recent experience. THE PROBLEM OF ATTRIBUTING EFFECTS It is one thing to determine how much, how often, and how people drink. It is quite another to pinpoint how much the different dimensions of drinking practices and different characteristics of the environment contribute to the important social effects of drinking. For the purpose of informing policy judgments, the question of attribution seems fun- damental: we need to know what interventions targeted at different aspects of drinking practices or different features of the physical and social environment can be expected to yield in terms of positive and negative effects. Unfortunately for this purpose, a main conclusion of this chapter is that a variety of factors shape the effects associated with drinking. Drink- ing is a major factor, but the degrees and frequency of intoxication as well as cumulative consumption must be considered to develop any clear sense of the kinds of risks and benefits that a drinker may encounter.

The Nature of Alcohol Problems 43 In addition, the contexts in which drinkers do their drinking play an important role In creating effects, in part by differentially exposing drink- ers to the hazards of the environment and in part by exerting an inde- pendent influence on quantities and rates of alcohol consumption. Fi- nally, general features of the envirnments into which drinking may be fitted by the choice of context also play a role in producing different kinds of effects. The sheer complexity of this conception defeats any simple effort to attribute quantitative risks or benefits to specific factors. What our analysis suggests that is useful in guiding policy choices is a more general idea: namely, that various avenues for reducing the bad consequences of drinking exist, and that pursuit of each avenue results in slightly different gains (and losses) in terms of the dimensions of Figure 2. This suggests that policies toward alcohol use require complex choices about the relative importance of different consequences as well as choices about which instruments to select from a large and diverse array some targeted at cumulative consumption, some at drunkenness, some at contexts, and some at characteristics of the social and physical environment. A separate question about the attribution of risks and consequences of drinking can be usefully addressed: how the risks and consequences of Figure 2 are distributed across the population of drinkers. The adverse consequences are doubtlessly at highest concentration in the small frac- tion of the drinking population who drink a lot, get drunk frequently, and do so with little respect for time and place. This expectation has justified the current focus of policy on alcoholics and chronic, intensive users of alcohol. What is less obvious, however, is how much of the total social problem described in Figure 2 lies beyond this small fraction of the heaviest drinkers. Current answers to this question are preliminary but revealing. Table 6 presents data from the Air Force survey showing the fraction of people in given drinking patterns who experienced two or more serious incidents associated with alcohol in the past year. As we would expect, the fraction of people having these problems increases noticeably with given levels of consumption. But if we calculate the fraction of all the people ex- periencing two or more serious incidents related to alcohol who were in the highest consumption groups, we find that the heaviest drinkers account for only 24 percent of the population having problems. The reason is that there are so many more people in the lower-consuming groups that even small proportions in trouble can swamp the larger fraction of people who are in trouble from the higher-consuming groups. These qualitative results also stand up when we look at data from national surveys conducted in 1967 and 1977. Table 7 shows the fraction of people at different levels of consumption who experienced certain

44 REPORT OF THE PANEL TABLE 6 Alcohol Incidents by Total Alcohol Consumption Level in a Sample of Air Force Men Average Daily Number of People with Percentage of All Consumption Level in Total Two or More Serious People with Two or Ounces of Alcohol Number Incidents More Serious Incidents 0.1-1.0 1,761 37 26.8 1.1-2.0 418 33 23.9 2.1-3.0 181 13 9.4 3.1~.0 79 14 10.1 4.1-5.0 49 8 5.8 5.0+ 111 33 23.9 TOTAL 2,599 138 99.9 Source: Adapted from Polich and Orvis (1979~. kinds of problems. Again, it is clear that the chance of experiencing difficulties associated with drinking increases as one drinks more. But Table 8 shows how much each portion of the drinking population con- tributed to people who have problems associated with drinking. Again, it turns out that the heavy drinkers typically account for less than half of the people with problems. Thus, in attributing the consequences of alcohol consumption to var- ious causal factors and types of drinkers, we can make two important observations. First, alcohol consumption itself is important as a sufficient cause for some effects. Other consequences of drinking can be attacked by reducing drunkenness, motivating people to change drinking con- texts, or changing the external environment as well as by reducing cu- mulative individual consumption. Second, while chronic drinkers with high consumption both cause and suffer far more than their numerical share of the adverse consequences of drinking, their share of alcohol problems is still only a fraction- typically less than half the total. Alcohol problems occur throughout the drinking population. They occur at lower rates but among much greater numbers as one moves from the heaviest drinkers to more moderate drinkers. CONCLUSION: AVENUES FOR AFFECTING ALCOHOL PROBLEMS The main conclusions of this chapter are that controlling alcohol and treating relatively small numbers (even as many as 1 or 2 million) of the most troubled drinkers are not the only, and may not be very effective, ways of coping with alcohol problems. We have developed these conclusions largely by analyzing the complex

The Nature of Alcohol Problems TABLE 7 Percentage of U.S. Population Experiencing Different Problems Associated with Drinking, by Level of Consumption 45 Prob- Prob- lems lems with with Friends Spouse Total Number Job Accidents/ of Prob- Legal Respon- lems Problems dents 1967 National Survey Light drinkers 6 1 Moderate drinkers 8 5 Heavy drinkers 14 20 1979 National Survey Light drinkers Moderate * * * * 394 2 . . ~ a ~ 499 8 5 11 0 1o2 1 1 * 1 * 444 drinkers 5 10 4 2 6 2 ss7 Heavy drinkers 10 31 6 6 24 7 124 * Less than 0.5 percent. Source: Clark and Midanik (1980). TABLE 8 Contribution of Different Consuming Groups to Reported Experience of Problems Problems Problems Job Accidents/ Health Bellig- with with Prob- Legal Problems erence Friends Spouse lems Problems 1967 National Survey Light drinkers 26~o Ho * * * * Moderate drinkers 51 45 42~c 30% 54% 100% Heavy drinkers 23 46 58 70 45 0 Total percentage 100 100 100 100 99 100 (Total number) (83) (56) (19) (10) (33) (6) 1977 National Survey Light drinkers 13~o Ho 12~o 5~c 9% 0% Moderate drinkers 59 59 66 62 49 55 Heavy drinkers 28 37 22 33 43 45 Total percentage 100 100 100 lOO 101** 100 (Total number) (46) (103) (32, (21) (70) (20) * Less than 0 5 percent. ** Totals do not add to 100 percent due to rounding error. Source: Clark and Midanik (1980~.

46 ·~ =1 ~ o ol I ul cn ~ ~, E `,, ° w ~ ~ | ~__ ° ~o . r---~ r- - -~ ~ Ir ~ G ·C~ c: ~ ~0 L___J L___J r ' I _ .. E x O I O; .O .O L_J ~ L___J L___J cn ~ 1 ~ c c= ._ .O O _ ~ r ~ 5~ . . . I .O 1 .m :~ 1 q' c 1 ~ c~ a) Q ~ 1 C: <: O ° oc m C) C) o ,~ C~ C) a~ - I - o o ._ Ct Ct - x C) Ct L~° D , C~ — _ C~ a~ ._ Ct r~ — O O o ._ . C) ;> X C) ._ _ ~ ~_ Ct ~ 'V ·— =: C~ ._ ~0 C~ ~ .. r~ ~ — :?

The Nature of Alcohol Problems 47 causal machinery that linked drinking with the important consequences of drinking. Figure 6 presents a schematic view of our conception of the broad causal system. In this scheme, drinking is incorporated into the concept of a drinking practice, which specifies drinking in several di- mensions (degree of intoxication, frequency of intoxication, and cu- mulative consumption) and includes commonly chosen contexts for drinking as well. Drinking practices mediate the effects of drinking but, for many effects, do so only in combination with some features of the physical and social environment. Note that the environment appears in this scheme twice: as something that combines with a given drinking practice to produce consequences (see Figure 2) and as something that shapes drinking practices by making alcohol available, by defining at- titudes toward drinking practices, and so on. Often, the same features of the environment play a role in both shaping practices and producing consequences; still, one can analytically distinguish among the different causal roles. Contexts appear in this scheme as a description of how drinkers incorporate pieces of the environment into their drinking ac- tivities. We assume that the contexts exert an independent influence on consumption and that the environment influences the frequency with which certain contexts appear in drinking practices. The sheer complexity of this system argues for the existence of al- ternative routes for controlling the dimensions and shape of the alcohol problem, apart frown reducing total consumption. At a minimum, one can conceive of reducing consumption levels, shaping drinking practices, and changing features of the physical and social environment. Although the theoretical existence of these alternative approaches can hardly be taken alone as a strong argument supporting their use, their existence does stir our curiosity about whether they might be exploited. The conclusion that treating alcoholics is not sufficient response to the problem is based on an empirical result: namely, that the problems associated with drinking are distributed rather broadly through the pop- ulation of drinkers. It is true that the small proportion of the most troubled drinkers cause and suffer far more than their numerical share of alcohol problems; it is also true that if all the clinically diagnosable alcoholics were to stop drinking tomorrow, a substantial fraction of what we understand as alcohol problems would still remain. The general alcohol problem, then, is more than what the standard view of "alco- holism" suggests, because it affects many more of us than we are ac- customed to believing. The problem may have less to do with alcohol than the alcoholism concept suggests and have much more to do with how alcohol is fitted into daily life and the external environment.

Next: 3. Perspectives on Current Alcohol Policies »
Alcohol and Public Policy: Beyond the Shadow of Prohibition Get This Book
×
Buy Paperback | $125.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF
  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!