From: Engs, Ruth C. [Ed.], " Controversies in the Addition's Field". Dwight B. Heath, Ph.D. CHAPTER 9: Flawed Policies from Flawed Premises: Pseudo-Science about Alcohol and Drugs

Back to table of contents
Home Page | Article List | Questionnaires | Books | Search my Files | Health Hints | Resume

CHAPTER 9
Flawed Policies from Flawed Premises: Pseudo-Science about Alcohol and Drugs

Dwight B. Heath, Ph.D

Public health policies are often influenced by political considerations that are extraneous, or even counterproductive, in terms of the scientific and humanitarian criteria that ideally would apply. We all recognize that economic necessity, unrelenting pressures of time, lack of personnel or facilities, the unavoidable need for compromise, emotional reactions, or various other reasons can be invoked to account for some such discrepancies, although unalloyed political expediency is doubtless also a consideration in at least some instances. An increasingly influential factor in recent years has been the promulgation of quasi-scientific "data" and "principles" that come to be accepted as justification for a variety of policies but that have little factual basis.

In this paper, I will briefly discuss different kinds of representation and misrepresentation that occur concerning alcohol and drugs; although the issue is undoubtedly relevant elsewhere, I will focus on agencies of the United States and the United Nations.

The Fallacy of "the Control Model" of Prevention

Throughout the world, people are increasingly being told by public health "authorities" and "experts" that it is necessary to reduce everyone's drinking if we are to reduce a wide range of alcohol-related problems. For example, within the context of a W.H.O. publication, it is baldly asserted that "because of the overwhelming evidence that consumption levels are closely related to the extent of alcohol-induced harm, the cornerstone of primary prevention must be the control and, where necessary, the reduction of national per capita alcohol consumption" (Walsh, 1982:79).

Unfortunately, the evidence is by no means overwhelming, so such a simplistic formula is grossly distorted and far removed from what is often cited as a factual basis. The intellectual cornerstone of the recent wave to control the availability of alcohol, cited as such by all of those who have any scholarly credentials or pretensions in the field of alcohol studies, is a monograph compiled by an international group who met in Finland some

76


years ago (Bruun et al., 1975). The crucial wording of their original report is extremely tentative. What they said, after comparing historical trends concerning death from liver cirrhosis in various countries was simply that "... changes in the overall consumption of alcoholic beverages have a bearing on the health of the people in any society. Alcohol control measures can be used to limit consumption: thus control of alcohol availability becomes a public health issue" (Bruun et al., 1975:90).

Successive attempts to find out just what kind of "bearing" such changes in consumption have on public health and social welfare have met with a remarkable lack of uniformity. Lest it appear that I selectively emphasize the views of colleagues who are outspoken critics of increasing legislative and regulatory controls, I will first cite people who had been coauthors with Bruun in the original study. For example, a multinational study by Makela et al., (1981), including fourof Bruun's co-authors, offers the following generalization: "as each society has its peculiar drinking habits, the mixture of problems varies accordingly and, in cross-sectional comparisons, there are few positive relationships between consumption level and the incidence of problems..." (Makela et al.,1981 :62). Similarly, "relations of alcohol consumption level and patterns to casualties and social problems associated with drinking are far less clear and universal correlations of these problems with consumption level are frequently negligible or negative" (ibid: 90). A close inspection of the case-studies reveals those inconsistencies. For example, both Poland and Finland had "very low" overall consumption but both had high rates of social conflicts related to drinking, crimes of violence, arrests for drunkenness, and fatal alcohol poisoning. The Netherlands, with "very similar" consumption, showed "few signs of disruptive drinking, " while Switzerland, with "the highest level of aggregate consumption" among the seven societies studied, had "few signs of social conflicts related to disruptive drunken comportment" (ibid: 45-46).

Despite such evidence and their own apparently contradictory generalizations, they asserted at the end that "one of the basic findings of our study, however, is that the growth in alcohol consumption was accompanied by an increase in a broad variety of problems related to drinking. When all the case studies are taken together, this holds true not only for consequences of prolonged drinking, but for social and health problems related to single-drinking episodes" (Makela et al., 1981: lO9).

That kind of oversimplification has tended to recur in a variety of contexts, and to influence legislators, administrators, and others who are not familiar in detail with the studies. For example, addressing an international policy-oriented readership, "...alcohol-related health problems can be expected to vary with the level of overall consumption" (Sulkunen, 1985: 123).

77


Writing to a broad lay readership in the U.S., to introduce a set of policy recommendations: "while only a fraction of all alcohol users—the alcoholics—have most of the alcohol related illness, there is a distinct correlation between the total consumption of alcohol by a society and the prevalence of all alcohol related problems and diseases including alcoholism per se" (West, 1984: 4). When the Director General of the World Health Organization asserted unequivocally that "any reduction in per capita consumption will be attended by a significant decrease in alcohol related problems" (W.H.O.,1978: 4), too many jounalists, bureaucrats, and other non-specialists unquestioningly accepted that apparently authoritative judgment.

One might well wonder how it can be that such cautious and nuanced conclusions as those originally offered by Brunn et al. (1975) became converted or "translated" into such dogmatic and slanted premises, which now serve as guides to policy-making.

Whereas the premises of "the control model" appear, in cross-cultural perspective, to be flawed, the alternative "sociocultural model" (Heath, 1988a) fits with the experience of many populations throughout history and around the world. That is, teaching young people honestly about the nature and effects of alcohol, and instilling them with norms about the limits of appropriate and inappropriate behavior, serves as a kind of "immunization" against alcoholism and drinking problems. No immunization is foolproof, but the reactions against prohibition when it was tried in India, Britain, Sweden, Russia, the United States, and elsewhere, or the "forbidden-fruit" appeal among American adolescents today, suggest that increasing emphasis on social control—peer-pressure, adherence to norms, etc.—is likely to be more effective in the long run than increasing emphasis an legal and regulatory controls—such as shorter hours of sale, higher purchase-age, server-liability, and so forth.

Playing the Numbers: How Many Alcoholics?

A recent article in Science took testimony before the U.S. Senate Committee on Governmental Affairs as a distressing example that "govemment research institutes and individual scientists are generally careful to include the best information in their reports to Congress, but occasionally the data are not statistically sound or are extrapolated beyond the limitations of the survey" (Barnes, 1988: 1729). It has been a standing joke among researchers that, during the two decades since Congress (in a unanimous vote) established the National Institute on Alcohol Abuse and Alcoholism, through which hundreds of millions of dollars have been channeled for treatment, prevention, and research, the estimated number of alcoholics has continued to rise at a rapid rate. Rather than question the

78


efficiency and effectiveness of the agency, let us consider the basis for such estimates.

Robin Room provides an anecdotal account of how the original "9 million" was extrapolated, on the basis of a survey that included no questions about alcoholism (quoted in Barnes 1988: 1729). Don Cahalan, Kaye Fillmore, Mark Keller, Selden Bacon, and other knowledgeable researchers have, in diverse contexts, provided variant, but similarly plausible, versions of what might be viewed as an origin myth for this quasi-scientific pseudo-statistic.

Over the years, successive publications by NIAAA have revised the number upward, usually with no source cited, or with reference to an unpublished survey that was not available through the Institute.The terminology has changed often in succeeding years, so that, in the most recent report, "an estimated 18 million adults 18 years old and older currently experience problems as a result of alcohol use" (NIAAA, 1987:12). Repeated efforts by several interested epidemiologists and social scientists to ascertain an empirical basis for the rapidly evolving estimates of prevalence have been unsuccessful.

The Fabrication of "Consensus Documents"

Levine (1984) described his concern, based on first-hand experience in which he found many discrepancies between the so-called "consensus document" that was promulgated by the organizers of an international transdisciplinary conference on alcohol, and what had actually transpired at the conference. As a fellow participant in the same conference, I share his consternation that the document stressed some things that had never been mentioned during the conference, distorted much of what had been said, and proposed generalizations that virtually all of the speakers had explicitly warned were unwarranted (Heath, 1988b,c). Delegates to the World Health Assembly who saw the "consensus document" but not the proceedings could not have avoided misunderstanding, as the document that purported to summarize our scientific deliberations went in very different directions from what actually transpired.

Propaganda in the "War on Drugs"

In attempting to muster support for his tough enforcement-oriented policies, the Director of the Office of National Drug Control Policy provided "...terrible proof that our current drug epidemic has far from run its course. Estimated 'frequent' use of cocaine in any form (measured by the number of survey respondents who report ingesting that drug one or more times each week, and calculated as a percentage of the total cocaine using population) has doubled since 1985" (Bennett, 1989:3). That

79


statistic is not nearly so alarming when one remembers that, citing the same survey just 2 pages earlier, he had noted that "current use" of cocaine was down 48 percent in the same time. This means that only about half as many people were using cocaine in any given month (in 1989, as compared with 1985). Halving the denominator—a far more important index of a diminishing "drug epidemic"—would automatically double the rate of "frequent use," even with no change in the actual number of those confirmed addicts. If our concern is with the prevalence of drug use and its related problems, the significantly smaller number of "current users" is, as he himself noted, "good news—very good news" (Bennett: 1989:1). That kind of numerical shell-game may be effective as propaganda in a "Drug Czar's" declared "war on drugs," but it grossly distorts the findings of the ninth periodic National Household Survey on Drug Abuse, and is seriously flawed as a basis for developing or justifying any kind of policy.

A Warning about Warning Labels

Recent federal legislation that mandates the inclusion of a health warning label on every container of alcoholic beverages illustrates another instance in which a policy was instituted on the basis of virtually no scientific evidence, although media coverage and political pronouncements often implied the opposite. Senator Strom Thurmond had repeatedly proposed such labeling since 1967, and had repeatedly failed to gain sufficient support. Part of the hesitancy on the part of many Congresspersons may well have been the fact that testimony in successive years demonstrated that there was no support for assertions that such labels would prevent alcohol-related problems by educating consumers. Finally, in 1986, Congress charged the Public Health Service to conduct a search of the literature on this controversial subject. The project was contracted to Micro Systems, Inc., who concluded: "The literature search did not discover any empirical research studies which examined the impact of product labeling upon the health consequences associated with [any kind of ] product use. [Furthermore,] the literature search did not identify any studies of the effects of alcohol warning labels in countries that currently require such labels" (D.H.H.S 1987: Appendix B(1)). In short, no immediately relevant data exist in print!

But they went on to review the literature that was available on other aspects of warning labels: 2 studies of warnings on soft drinks about saccharin, and one study of warnings on cigarette packages (in Sweden). On the basis of those 3 cases—none on alcohol, and one in another country— they did not hesitate to make policy recommendations. They concluded that "health warning labels can have an impact on the consumer if. . .certain conditions be met, so the review of the research literature on health

80


warning labels suggests the following major implications for alcohol warning labels: ..." (D.H.H.S.1987:3-5). The ensuing web of conjecture was treated as if it comprised substantive findings of the sort that had been sought but could not be found.

The Dangers of Pseudo- Science

Presumably the P.H.S.'s report to Congress was interpreted as "proving" the efficacy of labeling. The Omnibus Drug Act of 1988 (P.L. 100-610: Title VIII) included, with no fanfare, a provision that drinks packaged after November 1989 carry a warning from the Surgeon General about possible fetal damage and about risk while driving or operating machinery.

It seems unlikely that anyone seriously believes a warning-label would deter an alcoholic from taking a drink, but there does appear to be a widespread belief that "it wouldn't do any harm, so why not require a label just on the off-chance that it might do some good?" The answer to that question is that taking ineffective—or, at best, untested—action seems likely to divert our attention from the need for tested and effective action.

Instead of gaining a false sense of security from having mandated warning-labels, legislators might better have supported further research on why it is that a small portion of those who drink encounter various drinking problems, how young people can be taught to abstain (like fully one-third of the adult population in the U.S. today), or to drink moderately (like fully 90% of the rest), thereby avoiding such problems. Or another appropriate action would be to learn more about how those who have such problems can be helped to overcome them.

It is not that we lack such knowledge today. We do not have all the answers, but great progress has been made on all of these subjects. Perhaps more to the point is the fact that such knowledge is being effectively channeled into appropriate programs of prevention and treatment, but at a distressingly slow rate. In recent years, such practice efforts are, ironically, increasingly hampered by the diversion of funds to support an increasingly militant but ineffectual "war on drugs." As in this country's several previous attempts at stemming drug-use by military, police, and judicial actions (Brecher, 1987), drugs are becoming not scarcer but rather more widely available, not more attenuated but ever stronger, and not more expensive, but cheaper yet, as government expenditures on countermeasures escalate and as abuses of civil liberties (in the name of law enforcement) proliferate.

Similarly, the fallacy of the control model of prevention offers the illusion of providing a sure solution to alcohol-related problems far more quickly and easily than could be done with the sociocultural model.

81


Revision of norms is a slow and gradual process, requiring education in the broadest sense, and involving all sectors of the community. But the contrasts between Spaniards who drink every day and Swedes who do so only occasionally, Jews who are taught at an early age to drink ceremonially, and Irishmen whose convivial drinking starts much later in life, Camba who praise drunkenness and East Indians who fear it, all show that problems stem less from how much one drinks than from how one drinks and what one expects the outcomes to be (Heath, 1982).

Inflated estimates of the prevalence of alcoholism, or of the "growing" use of cocaine, are crude appeals to the compelling quality of numbers. They should neither distract us from addressing the far more important qualitative dimensions of public and private harms that are suffered nor should they be taken as quantitative justification for continuing policies that have proven to be flawed.

Perhaps the most tragic risk in all such manipulations is that the growing use of pseudo-science might eventually discredit the scientific enterprise itself. We would all be poorer if policy-makers came to view science as irrelevant and empirical evidence as unimportant. Let us hope that more people will learn to recognize the difference.

REFERENCES

Barnes, D.M. (1988). Drugs: Running the numbers. Science, 240 1729-1731

Bennett, WJ. (1989). Introduction, NationalDrug Control Strategy (Office of National Drug Control Policy), U.S. Government Printing Office: Washington.

Brecher, E.M. (1987) Drug laws and drug law enforcement: A review and evaluation based on 111 years of experience. Drugs and Society, 1 (1) 1-27.

Bruun, K, Edwards, G. Lummio, M, Makela, K., Pan, L., Popham, R.E., Room, R., Schmidt, W., Skog, OJ., Sulkunen, P., and (Osterberg, E. (1975). Alcohol Control Policies in Public Health Perspective. Finnish Foundation for Alcohol Studies: Helsinki.

Department of Health and Human Services (June,1987). Review of the Research Literature on the Effects of Health Warning Labels: A Report to the United States Congress (pursuant to PL. 99-570, sec. 4017, Anti-Drug Abuse Act of 1986), [D.H.H.S.: Washington].

Heath, D.B. (1982). Sociocultural variants in alcoholism. Encyclopedic Handbook of Alcoholism (Eds. E.M. Pattison and E. Kaufman). Gardner Press: New York.

Heath, D.B. (1988a). Emerging anthropological theory and models of

82


alcohol use and alcoholism. Theories on Alcoholism (Eds. C.D. Chaudron and D.A. Wilkinson). Addiction Research Foundation: Toronto.

Heath, D.B. (1988b) Alcohol control policies and drinking pattems: An international game of politics against science. Journal of Substance Abuse, 1 109-115.

Heath D.B. (1988c) Quasi-science and public policy: A reply to Robin Room about details and misrepresentations in science. Journal of Substance Abuse, 1 121-125.

Levine, H.G. (1984). What is an alcohol-related problem? (Or what are people talking about when they refer to alcohol problems?). Journal of Drug Issues, 14 45-60.

Makela, K., Room, R., Single, E., Sulkunen, P., and Walsh, B. (1981). Alcohol, Society and the State, Vol. 1: A Comparative Study of Alcohol Control. Addiction Research Foundation: Tronoto.

National Institute on Alcohol Abuse and Alcoholism (1987). Sixth Special Report to the U.S. Congress on Alcohol and Health, (D.H.H.S. Publication (ADM) 87-1519), N.I.A.A.A.: Rockville, MD.

Sulkunen, P. (1985). International aspects of the prevention of alcohol problems: Research experiences and perspectives. Alcohol Policies (Ed. M. Grant). World Health Organization Regional Publications, European Series 18: Copenhagen.

Walsh, D. (1982). Alcohol-Related Medicosocial Problems and their Prevention. World Health Organization Regional Publications, Public Health in Europe 17. Copenhagen.

West, L.I. (1984).Alcoholism and Related Problems: Issues for theAmerican Public. The American Assembly, Columbia University, and Prentice-Hall, Inc.: Englewood Cliffs, NJ.

World Health Organization (1978). Report by the Director General, Executive Board, Sixty-third Session, AlcoholRelated Problems: The Need to Develop Further the WHO Initiative (Provisional Agenda Item 23, Attachment E). World Health Organization: Geneva, (27 Nov).

83

Back to table of contents/