TO ALCOHOL SOCIOLOGY HOME PAGE
Citation: Roizen, Ron, "Comment on the 'Rand Report,'" Journal of Studies on Alcohol 38:170-178, 1977.
Section 102 of the Public Law that created the National Institute on Alcohol Abuse and Alcoholism asked NIAAA to "submit an annual report to Congress which shall include a description of the actions taken, services provided, and funds expended under this Act....an evaluation of the effectiveness of such actions, services, and expenditures of funds, and such other information as the Secretary of Health, Education, and Welfare considers appropriate" (1). Thus evaluation research of one sort or another was an integral part of NIAAA as it was conceived by the Congress. In its broadest form this evaluative effort was to be addressed to the general goals set down by Congress for the Institute: namely, to "develop and conduct comprehensive health, education, training, research, and planning programs for the prevention and treatment of alcohol abuse and alcoholism and for the rehabilitation of alcohol abusers and alcoholics" (1). In practice, a large fraction of NIAAA's resources was devoted to the treatment of "alcohol abusers and alcoholics" in a system of federally supported Alcoholism Treatment Centers, and much of the evaluation effort was directed toward this enterprise. Before too long NIAAA was busy establishing a system of alcoholism-treatment evaluation that is surely unequaled in terms of size and ambitiousness in the history of alcoholism treatment in the United States or elsewhere.
Instead of funding many little evaluation studies or a series of yearly larger studies, NIAAA created an ongoing monitoring system which could produce up-to-date data in a conveniently standardized form on a regular basis. By March 1973 the system had produced a preliminary report (2) on the characteristics of treatment center patients and various cost-oriented analyses of treatment center functioning. In mid-1975 Stanford Research Institute (3) published a more complete account of the monitoring system data and of a special 18-month follow-up study in which attrition from the sample had been kept to a minimum. The data from the evaluation system also trickled out in a variety of smaller papers and news reports: The Alcoholism Report carried highlights of the Stanford report on one or two occasions;2 Chafetz (4) published a slightly more detailed account of some of the major findings. Towle (5) presented summaries of the findings at NIAAA's Fourth Annual Conference on Alcoholism and at the Epidemiological Section meetings of the 21st International Institute on the Prevention and Treatment of Alcoholism in Helsinki in 1975.3 Some of the system data were used in NIAAA's second report to the Congress (6). Until the publication of the Rand report (7), 1 suspect that the readership of the large reports (2, 3) was confined to a small circle of government officials, consultants to the evaluation project, and prominent scholars with a pipeline to reports like this. The shorter papers and articles may have had a broader readership but lacked both details and the spark to ignite controversy.
The public controversy that began after the release of the Rand report in June 1976 seems to have been foreshadowed by an intense private controversy extending for over a year, about which the record is still incomplete. It is a matter of public record that both officials at Rand and a United States Senator3 were asked to delay publication of the report until its contents could be reviewed and its data, perhaps, reanalyzed. The central concern in this private controversy and in the public controversy that followed has been the report's apparent implication that at least some and perhaps a respectable fraction of alcoholics do and can return to normal, controlled or social drinking. The belief that this finding is wrong and its publication dangerous has been responsible for most of the debate, and it seems to have created the vigorous effort in many quarters to examine with a fine toothcomb the data, assumptions and implications of the report.
Of course, neither the evaluation system nor the Rand Report was solely focused on the abstinence question. The evaluation system was designed to provide data on a number of dimensions of alcoholism treatment including management information on services rendered and their costs, on the extent to which treatment centers were meeting local demands for treatment services, and on patient change between the time of entering treatment and several follow-ups. Thus, even the "patient change" problem was only one of several that the evaluation system addressed. And "patient change' was measured in several different ways in the longer reports that concerned this subject (3, 7).
This exclusive focus on the abstinence question has caused reviewers to miss some of the other interesting findings in the report. For example:
For some years now, attempts to measure the prevalence of alcoholism with epidemiological surveys have been hampered by the absence of comparable data on clinical samples. Without clinical norms to compare them with, the findings of general population surveys alone could not indicate the prevalence of cases that were similar to or equally severe as alcoholics in treatment. As Clark (8) pointed out a decade ago, prevalence estimates could be varied from very high to very low (or zero) depending on where the cutting points were laid. And without comparable clinical data or some other method for devising them, these cutting points were left arbitrary. Armor et al. (7, ch. 3, app. A) analyzed problem drinkers in the general population (drawn from data collected in four Harris surveys commissioned by NIAAA) and alcoholics in alcoholism treatment centers. The findings suggested that current estimates of the prevalence of alcoholism in the U.S.A. may be much too high if what is meant by the estimate is the size of a general population group with drinking practices and alcohol-related impairments similar to those of persons entering treatment programs.
The alcoholism treatment center studies yielded comparable data on patients who were in treatment as a consequence of being arrested for driving while intoxicated (DWI) and on non-DWIS. Although the authors did not conduct a detailed analysis of these two groups, the data suggest that DWIS and non-DWIS had substantially different alcohol-related characteristics. Thus these data raise questions about the application of the concept of alcoholism to the DWI population.
The data on spontaneous remission from alcoholism have never been adequate (9, 10)6 in spite of the fact that this subject is critically important to a number of concerns in the alcohol problems field. The implication of the monitoring system data seems to be that a substantial fraction of untreated persons undergo improvements in their conditions without the aid of treatment.
The study found that treatment types and patient characteristics predict very little of the remission variance.
Related to this, the finding that predictors of problem drinking in the general population may not perform very well as predictors of remission in the clinical population suggests the need for a careful look at the logic of our attempts to account for and to treat drinking problems.
The Abstinence Fixation
Most of the public commentary on the Rand report has focused on the abstinence issue. This is an interesting and telling concentration of public attention. As I mentioned, the report itself was not addressed solely to the abstinence issue but was a comprehensive analysis of the data from NIAAA'S monitoring system for alcoholism treatment centers. But the abstinence issue seems to burn brightly in alcohologists' campfires, so that raising the issue is likely to create a debate in which at least some participants regard the subject as a vital concern. But why does the abstinence question have this excitatory power? What are the things at stake when the question is raised? And why has a question that would seem to be quite amenable to empirical resolution nevertheless remained an unresolved matter for so long?
The abstinence issue is controversial because this question is intimately tied to a full paradigm or gestalt about alcohol-related problems. It is not the fate of the abstinence question alone that is at stake in the debate but the fate of this larger paradigm. In order to create a picture of the abstinence question it is necessary, first, to discuss a few points concerning that paradigm.
In the classical model, an alcoholic is 'diseased," first and foremost, in that his drinking behavior is outside his own volitional control. His troubles with drinking, as Jellinek suggested (11), stem from the fact that he is ignorant of his own constitutional difference from other people and because he sees no reason to suspect that such a difference exists. Strong rationalizations may surround his drinking behavior, and these may serve to "explain" his drinking to himself to his own satisfaction, thereby making it more difficult for the therapist to convince him that he is different. But it is the education of the alcoholic--convincing him that he is different--that is the major element of the classical treatment ideology. Thus, this ideology does not treat "the alcoholism" in a patient, which is to say it does not attempt to affect the putative constitutional difference that makes him an alcoholic, but rather attempts to alter his conscious knowledge of his own physiology or psyche. The treatment attempts to educate him to the acceptance of the belief that he is different, and that this difference is not a piecemeal or transitory thing. In short, alcoholism treatment, as it is seen the classical model, is largely a didactic and persuasional effort that attempts to treat an alcoholic by getting him to accept a particular theory of alcoholism, and his acceptance of that theory is itself the essence of alcoholism treatment. It is a process organized around the conscious half of the mind-body dualism that is implicit in the Alcoholics Anonymous image of man; and to that extent it is more aptly called an educational rather than a treatment program. The outcome criterion of "abstinence" is not solely a measure of the patient's improvement, but a sacred and essential element in the "treatment' process. The treatment is itself the inculcation of the disease conception of alcoholism, and, in this domain, the disease concept means that a constitutional difference in alcoholics is at the root of troubles with drinking. If the constitutional basis theory is accepted, the alcoholic may be persuaded to give up drinking. Embracing abstinence is thus a sign that the model of alcoholism has been accepted by the patient. Thus, measuring whether or not the alcoholic internalizes this theory and makes a commitment to abstinence becomes the only really important outcome criterion.
We usually think of a theory concerning a disease as "coming before" and "informing" the treatment regimen that it suggests. In the case of the classical alcoholism model, we have a theory whose acceptance-by-the-patient is the treatment. The close fit between theory and practice creates a number of difficulties. For one thing, it creates a confusion in the manner by which the theory should be judged and the relevant observations that should judge it. Most of us, of course, are familiar with any number of different criteria by which theories may be evaluated--we judge them by whether or not they fit the facts, whether they can be tested or falsified, whether they make sense or clear up old questions that have been nagging at us for some time, and so on. For our purposes, it is useful to distinguish two broad classes of judgmental criteria: those pertaining to the correctness of the theory, and those pertaining to its utility. While we usually think of these criteria as going together (that is to say, correct theories are more useful than incorrect ones, and useful theories are more likely to be correct than useless ones), the history of science is replete with theoretical systems in which the two criteria varied independently. The assumptions of Ptolemy's system of astronomy, in spite of being "wrong," provide a perfectly good basis for navigation of the seas--indeed teaching the use of the sextant is more simple in a Ptolemaic than a Copernican universe. Likewise some theories about which there is substantial consensus regarding their scientific correctness (e.g., Darwinian theory) seem to have few utilities either with regard to prediction or control of the things they pertain to.
In the case of the classical "alcoholism theory" it is essential that we keep separate issues of "correctness" and "utility." The classical "alcoholism" theory, as Keller has described its history (12), was adopted from A.A. ideology primarily because "it worked"--the success of A.A. was well known and impressive to the scholarly community that in the early 1940s was studying "the problem":
"At first glance it may seem surprising that much of the contemporary understanding of a disease, with which medical and allied therapeutic professionals are heavily engaged, should derive from a fellowship of laymen. Especially so when, if one re-examines the exhaustive review of the psychiatric literature published in 1941 by Karl M. Bowman and E.M. Jellinek, it is obvious what a vast amount of observation, study, theorizing and writing had been done in the effort to understand alcoholism. Why, then, in spite of all the sophisticated synthesis that came out of that review, did the medical and paramedical world, and Jellinek himself, soon after capitulate, as it were, to the lay wisdom of Alcoholics Anonymous? This problem merits a deeper consideration than I can give it in the present aside, but I would like to suggest that it was a very practical and understandable capitulation. For all the wisdom of the older medical-psychiatric writings, in the beautifully organized Bowman-and-Jellinek synthesis, made good sense in theory, but offered small help in practice. That is, in medical practice, in the practical business of successful treatment. On the other hand, at the time when that review was published, Alcoholics Anonymous began to become famous, the story of its success was then for the first time widely publicized. The medical world had to look, at first with surprise, and finally with conviction, at a way of dealing with alcoholism that worked."
Thus the theoretical stage is set by viewing the theory from a utilitarian (rather than correctness) perspective. In a sense, moreover, the utility of the theory was quite independent of its truth content. If in-coming patients can be convinced that they are "alcoholics," with all of the implications that this word carries in the classical model, then by all rights they should give up drinking and commence an abstinent life. If the theory were true, all the better; but if the theory were false it would not necessarily undercut the therapeutic usefulness that it might possess--unless patients were exposed to disconfirming evidence. Thus regardless of its truth content, the classical alcoholism theory provides a basis for treating the alcoholic and providing him with a number of things we regard as benefits: it gives an "explanation" of alcoholismic behavior that avoids ladening the alcoholic with too much guilt for past drinking excesses; it wins his release from the rnorally based criminal justice system of social management; it provides rationale for abstinence; it opens a route for his reintegration into conventional roles and obligations; it provides the basis for medically oriented research on the etiology of the condition; it, when an alcoholic embraces it, arrests his condition.
Of course from a purely intellectual standpoint the theory is surprisingly vacuous; it seems only to assert that alcoholics are different without anything more than speculation regarding the source or even the evidence for that difference. What is important to recognize, however, is that the theory is ideal from a utilitarian standpoint, and provides social managers of alcoholism with a seeming intellectual basis for carrying out their mission. This utility is intact so long as therapists can present the theory honestly and openly without fear of contradiction. Only when a considerable body of contrary evidence is around does the theory and the entire mode of approach to alcoholism begin to weaken. Without the ability, in other words, to create a genuine conviction in the classical disease concept of alcoholism, the theory, its treatment implications, and its authority and legitimacy dissolve. Thus in a crucial sense, the utility of the alcoholism theory--which both historically and contemporarily its best selling point--is absolutely premised on the degree of commitment or belief in the theory that the research and treatment communities are willing to invest. And thus the abstinence criterion is far more than a free-standing "outcome criterion"--it is at the foundation of the entire gestalt that the alcoholism theory reflects and at the foundation of its utility.
From a traditionalist's standpoint, an attack on the abstinence criterion is an attack on the classical disease concept of alcoholism. Small wonder that traditionalists balk at seeing the abstinence criterion too easily dispensed with in favor of some more eclectic or behaviorist success criteria.
One cannot report data that many or most alcoholics return to normal drinking without undercutting the fundamental "truth" that alcoholics in most treatment centers are told. And undercutting that truth is only done at great peril because the embracing of that truth proves to be the most successful treatment known for the condition. At a minimum, it cannot be undercut without supplying some other truth, either for the "telling to" of alcoholics or for the foundation of an equally efficacious treatment system.
Of course, proponents of controlled drinking have cited the ostensible social utilities of their position (e.g., 13):7 alcoholics will be more agreeable to entering treatment if they know that permanent abstinence will not be mandatory; that treatment will become possible for patients "for whom abstinence has not been attainable" (13, p. 820); or that treatment will be possible for alcoholics who do not care to affiliate with groups that specifically reinforce nondrinking. But these sorts of arguments, again, refer to the social utilities rather than the correctness of the outcome criteria and the theoretical picture of alcoholism that they imply.
The Declining Utility of the Disease Concept
I imagine that if a report showing that a substantial fraction of alcoholics were capable of controlled drinking had been released in the mid-1940s, the impact on the alcoholism movement would have been substantial. The argument for treatment-instead-of-punishment of chronic court offenders and treatment-instead-of-hiding for earlier-stage alcoholics was closely tied to the classical disease conception of alcoholism. To find that a substantial fraction of labeled alcoholics returned to controlled drinking or that a substantial friction of those diagnosed as alcoholics were mistakenly diagnosed would have undercut the claims to expertise that were crucial to the foundation of a transition to a medical or public health system for the management of alcoholism.
But the news falls on a different world three decades later. For instance, that 10, 20, 50 or 75% of the patients in alcoholism treatment centers return to controlled drinking is a finding that is unlikely to suggest that alcohol problems should be returned to the criminal justice system because the alcoholismic conditions are, thereby, shown to be volitional. We usually regard the disease concept as a means of softening the opprobrium toward alcoholism, but it is also possible that a declining sense of opprobrium in a society, for whatever reasons, may make the disease concept more or less superfluous. Perhaps the social history of alcohol problems will show that the ideal of medical-style social management of alcoholics survived the declining influence of the disease concept simply because no one had any interest in the punitive social controls of the past. I do not know what the sources of this diminished opprobrium are, but one can guess at several factors: Alcoholics have been upstaged, perhaps, by other-substance misusers and the larger household of social problems that attract public attention these days, and seem to pose more dramatic threats to the community. A chronically below-full-employment economy and a social structure that increasingly employs a vast sector of its population in health and social services are less likely to resent alcoholics for their absence from the labor market and perhaps will welcome their presence in one or another 'target" group or 'catchment population." Even the stereotypical Skid Row alcoholic's assault on public decorum has been diluted by a decade of change in hair, dress and cleanliness fashions that make the chronic court offender only one among other recognized social types about whom it has been said that "a haircut and a bath" are needed. One wonders about the extent to which the wartime circumstances of the early 1940s contributed a public sense of the need for every man and woman at some productive enterprise. This public sentiment may have created a special animosity toward alcoholics and the special need for a vigorous means for defending them from undue rejection. But, again, the time is gone and the images are faded.
Of course, the utilities of the classical disease concept are still alive in Alcoholics Anonymous and in other treatment enterprises where the classical model provides the basis for treatment. But where research rather than treatment (that is, where intellectual utility more than treatment efficacy) defines the situation, the abstinence issue is likely attract attention only grudgingly.
On this broader explanatory plane, the Rand report seems to have signaled some of the serious weaknesses in conventional perspectives on alcoholism.
Writing from the perspective of a would-be science of alcohol problems, the over-all impression I get from the report is that the theoretical paradigm that is implied in the "alcoholism" concept has not served to sort out and make intelligible the realities out there. Little of conventional wisdom seems to predict much, and that which does a little predicting (e.g., the worse one's initial condition, the less likely that one will change to a problem-free status) does not seem to the sort of "discovery" or "finding" that increases our knowledge over that of commonsense knowledge. We are left with the impression that an entire vocabulary drawn from theory and clinical experience does not order the world for us. The report constitutes a culmination the alcoholism perspective as it was put into practice by NIAAA; this perspective, and its language, do not constitute a demonstrably effect way to make sense of the data that are reported. This is not to say that treatment programs do no good or do good, but merely that, as scientists, we end up knowing little about what is going on.
In summary, the public controversy over the report has been confined to the abstinence question so that several interesting findings have received little or no commentary. Thus in a sense it is not Rand report that is the source of the furor but the abstinence question; the Rand report merely is the most recent occasion for the continuation of the abstinence debate. The debate over the abstinence question both in the past and as it has developed around this report has been too much narrowed to one question, namely, whether "true" alcoholics can return to controlled drinking. The intensity and the duration the debate about this question should be clues to the fact that it is controversial because the answer to it is crucial to the coherent and complete paradigm of deviant drinking. Thus in order to make sense of the debate, and to make some progress toward its resolution, we need to examine not just the abstinence question but the paradigm that is at stake when that question is raised. I have argued that the paradigm owes its continued popularity and the allegiance of its proponents primarily to the social and therapeutic utilities that have been ascribed to it (rather than to its powers of explanation or powers to advance research in the area). And I have suggested that the social climate of opinion regarding alcoholism has changed in the last decade or so in such ways that some of the social utilities of the paradigm are no longer quite as important as they once were.
1Social Research Group, School of Public Health, University of California, 1912 Bonita Ave., Berkeley, CA 94704.
2For example, in the issue of 28 March 1975, pp. 4-5.
3TOWLE, L. H. Routine monitoring of alcoholism treatment services and client follow-up as an input to national program planning and policy. Presented at the Epidemiological Section, 21st International Institute on the Prevention and Treatment of Alcoholism, Helsinki, June 1975.
4PENDERY, M. NCA, press conference, Shoreham Hotel, Washington, DC, 1 July 1976.
5The Alcoholism Report, 12 September 1976, p. 7.
6 Also ROIZEN, R., CAHALAN, D. and SHANKS, P. Spontaneous remission among untreated problem drinkers. Presented at the conference on Strategies of Longitudinal Research on Drug Abuse, San Juan, Puerto Rico, April 1976.
7It should be made clear that Lloyd and Salzberg (13), in the article which these examples are drawn, did clearly distinguish and discussed separately the theoretical and the utilitarian advantages (and disadvantages) of "controlled drinking" as a treatment goal.
REFERENCES1. U.S. Congress. Senate. Comprehensive alcohol abuse and alcoholism prevention treatment and rehabilitation act. S. 3835, 91st Cong. 2d sess.
2. Towle, L.H. Alcoholism, program monitoring system development; phase II. Menlo Park, CA; Stanford Research Institute; 1972.
3. Ruggles, W. L., Armor, D. J., Polich, J. M., Mothershead, A. and Stephan, M. A follow-up study of clients at selected alcoholism treatment centers funded by NIAAA. Prepared for the U.S. National Institute on Alcohol Abuse and Alcoholism. Menlo Park, CA; Stanford Research Institute; 1975.
4. Chafetz, M.E. Monitoring and evaluation at NIAAA. Evaluation 2 (No. I): 49-52, 1974.
5. Towle, L.H. Alcoholism treatment outcomes in different populations. Proc. 4th Annu. Alcsm Conf. NIAAA, pp. 112-133, 1975.
6. U.S. National Institute on Alcohol Abuse and Alcoholism. Second special report to the Congress. Alcohol and health; new knowledge. (DHEW Publ. No. ADM-75-212.) Washington, DC; U.S. Govt Print. Off.; 1975.
7. Armor, D.J., Polich, J.M. and Stambul, H.B. Alcoholism and treatment. Prepared for the U.S. National Institute on Alcohol Abuse and Alcoholism. Santa Monica, CA; Rand Corp.; 1976.
8. Clark, W.B. Operational definitions of drinking problems and associated prevalence rates. Quart. J. Stud. Alc. 27: 648-668, 1966.
9. Smart, R.G. Spontaneous recovery in alcoholics; a review and analysis of the available research. Drug Alc. Depend., Lausanne 1: 277-285, 1975.
10. Emrick, C.D. A review of psychologically oriented treatment of alcoholism. 11. The relative effectiveness of different treatment approaches and the effectiveness of treatment versus no treatment. J. Stud. Alc. 36: 88-108, 1975.
11. Jellinek, E.M. Phases of alcohol addiction. Quart. J. Stud. Alc 13: 673-684, 1952.
12. Keller, M. On the loss-of-control phenomenon in alcoholism. Brit. J. Addict. 67: 153-166, 1972.
13. Lloyd, R.W., Jr. and Salzberg, H.C.. Controlled social drinking; an alternative to abstinence as a treatment goal for some alcohol abusers. Psychol. Bull. 82: 815-842, 1975.