Further Reading


Addictive Behaviors, 17:49-62, 1992
The afterword was added in September 1996 when the article was included in this website.

Alcoholism, Politics, and Bureaucracy

The Consensus Against Controlled-Drinking Therapy in America

Stanton Peele
Morristown, New Jersey



Controlled drinking has long been rejected as a therapeutic goal in traditional American alcoholism treatment. More striking has been the adoption of this position by behaviorists who once fostered controlled-drinking (CD) therapy, such as Peter Nathan, former director of the Rutgers Center of Alcohol Studies. This shift has occurred even while the evidence for rejecting the value of moderation training is unclear. Rather than being justified empirically, this rejection must instead be understood as a strategy behaviorists have used to jettison a politically explosive element in the behavioral package in order to gain overall acceptance for the behavioral approach to treating alcoholism. However, this strategy entails significant scientific, clinical, social, and ethical costs.

Again and again there comes a time in history when the man who dares to say that two and two make four is punished with death. The schoolteacher is well aware of this. But the question is not one of knowing what punishment or reward attends the making of this calculation. The question is that of knowing whether two and two do make four.

Albert Camus, The Plague

In a 1989 book applying social-learning and behavioral techniques to treatment of alcohol dependence, Monti, Abrams, Kadden, and Cooney (1989) dispatched the controlled-drinking issue quickly and firmly (p. 189).

We believe that the overwhelming data to date do not argue for controlled drinking as a viable treatment for individuals diagnosed with alcohol dependence (Foreyt, 1987). Nathan (1986) states:

The consensus among informed observers is that alcoholism treatment with controlled drinking as a prime treatment goal is neither efficacious nor ethical when offered to chronic alcoholics....the weight of the available data now suggests both that we have not developed treatment programs that can reliably teach chronic alcoholics to become controlled drinkers and that status as a controlled drinker is not in the best interests of most chronic alcoholics. (p. 44)

For a more comprehensive review of the issues and research studies involving controlled drinking, see Heather and Robertson (1983).

Monti et al. argue from authority, without discussing the evidence, in rejecting CD therapy. Their principal authority is Peter Nathan, a leading behaviorist and former controlled-drinking therapy advocate who became an outspoken critic of CD therapy after becoming director of the Rutgers Center of Alcohol Studies in 1983 (Nathan resigned this position in 1989). For his part, Nathan (1986) proposes as the primary standard for accepting CD treatment that such programs "reliably teach chronic alcoholics to become controlled drinkers," implying that programs exist for reliably teaching chronic alcoholics to abstain.

Monti et al. concluded their argument with a reference to Heather and Robertson (1983), British researchers who might seem to be part of "the consensus among informed observers" and to present some of the "overwhelming data" to refute the value of CD therapy. In fact, the referenced volume, Controlled Drinking, supports the application of controlled-drinking in alcoholism treatment, and not "only to those with less serious problems. The controlled drinking treatment goal has a much more important role to play than this" (p. viii). Heather and Robertson described 26 studies that have applied CD methods with clinical populations, and found strong evidence that, "Many seriously dependent alcoholics respond to certain controlled-drinking treatments by successfully controlling their drinking" (p. 200). The volume also added to the 74 publications cited by Pattison, Sobell, and Sobell (1977) that have uncovered a return to controlled drinking by former alcoholics who did not receive CD therapy.

Monti et al.'s confident assertion of a consensus against CD therapy for alcoholics, a consensus announced by Nathan, is the focus of the current article. This focus includes two topics: (a) Whether CD therapy has been broadly rejected in the United States on the basis of the evidence, or whether political and economic pressures have moved many behaviorists to reject CD therapy, and (b) as an example of how science does not occur in a vacuum, the forces at work in Peter Nathan's turnabout on CD therapy, from a position until about 1980 of openness to the evidence and endorsement of the therapy's value, to a position in the mid-1980s of actively discouraging the practice of CD therapy.

Why Are Behaviorists So Sensitive About Controlled Drinking?

That Monti et al. did not seriously evaluate the evidence on controlled drinking in a book on behavioral treatment of alcohol dependence indicates the highly charged nature of this issue for American behaviorists. The application to alcoholism of the social-learning approach, with its emphasis on self-regulation skills, would seemingly compel consideration of moderation training. Monti et al. acknowledged that: "Controlled or social drinking, rather than total abstinence, would be a logical extension of the coping skills approach." Yet, these authors relegated controlled drinking to a parenthetical discussion on page 189 of a 192-page book. In his foreword to the book, Marlatt confronted the issue more directly: "readers who are unfamiliar with the cognitive and behavioral skills training approach to alcohol dependence may be surprised to find these methods can be used in support of an abstinence goal [because] many popular accounts have equated behavioral approaches with a controlled drinking goal."

As Marlatt indicates, behaviorists have special reasons to be sensitive to being labelled as supporters of controlled drinking. The progress of Nathan's career traces this sensitivity as well as the rise and decline of controlled-drinking therapy. Nathan became the first behaviorist (or clinician of any kind) to head the Rutgers Center, the flagship of academic alcohol studies programs. Nathan's early research observed alcoholics' drinking in the laboratory setting (Nathan & O'Brien, 1971). This popular experimental paradigm was employed in the early 1970s by Mello and Mendelson (1971), Marlatt, Demming, and Reid (1973), the Baltimore City Hospital group (Bigelow, Liebson, & Griffiths, 1974; Cohen, Liebson, Faillace, & Allen, 1971), and the Jefferson Medical College group (Gottheil, Murphy, Skoloda, & Corbett, 1972). Such research conclusively demonstrated that (a) alcoholics do not drink excessively due to an inherent, biological inability to control their drinking and (b) that, instead, alcoholics drink purposely to attain desired feelings and that they respond to environmental cues and rewards even when they have been drinking or are withdrawing from previous intoxication.

Although this work yielded crucial insights into the nature of alcoholic drinking, it also drew a great deal of criticism. Today, such work is not supported by the NIAAA because of repugnance at the idea of providing alcohol to alcoholics. Other objections were raised by Meyer (1981, pp. 243-244) at a 1979 symposium in which Nathan and Nancy Mello participated.

Dr. Peter Nathan's work has involved systematic studies of drinking behavior...with an eye toward identifying environmental conditions associated with differential rates of alcohol intake.... Dr. Mello and Dr. Nathan both come out of the tradition of experimental psychology and its application to operant studies of alcohol consumption in human subjects. In general, this group of investigators has been loathe to talk about terms such as "loss of control" and "craving." They have associated this terminology with a medical disease orientation, which does not permit the possibility that alcoholics can be reeducated into social drinkers by using behaviorally oriented technology.

Meyer made clear in his comments that he disagreed with this viewpoint and that he was ethically opposed to naturalistic and experimental studies that allowed alcoholics to drink.

Nathan has maintained the essential points gleaned from this research--that alcoholic drinking is better understood in terms of cognitive and environmental influences than as a biological loss-of-control set off by the chemical properties of alcohol. Nathan and McCrady (1987, p. 121), for example, wrote:

...[I]t has become increasingly clear that, in many instances, what alcoholics think the effects of alcohol are on their behavior influences that behavior as much or more than the pharmacologic effects of the drug. In study after study, the important role of expectancies about the behavioral effects of alcohol have been demonstrated.... Expectancies are relevant to craving and loss of control because many alcoholics do in fact subscribe to the view that craving and loss of control are universal among alcohol dependent individuals.

But if alcoholics' expectations that they can't control their craving for alcohol are central to their excessive drinking, then telling them controlled drinking is impossible should produce a self-fulfilling prophecy. Conversely, informing them that controlled drinking is a realistic possibility should make this outcome more likely. Thus, the social-learning approach appears inherently to support controlled-drinking outcomes and strategies, particularly when the alcoholics is predisposed to believe this is possible.

What Does Research On Controlled Drinking Indicate?:
The Rand And Sobell And Sobell Studies

Despite the powerful link between social-learning ideas and the possibility of controlled drinking, Nathan and other prominent cognitive behaviorists have rejected controlled-drinking therapy. Yet this rejection has not been based on decisive evidence derived from clinical trials and other treatment outcome research.

Surprisingly, Nathan and his colleagues have defended the two most-attacked studies used to support the validity of controlled-drinking outcomes with heavily dependent alcoholics. These studies, the Rand report and the Sobells' clinical controlled-drinking trial, have for some time been assailed by advocates of the disease theory of alcoholism. In 1982, a follow-up study of the Sobell and Sobell (1973, 1976) controlled-drinking patients (but not the abstinence treatment comparison group) was published by Science (Pendery, Maltzman, & West, 1982). The authors of the Science study publicly denounced as fraudulent the Sobells' results, which found superior outcomes for subjects (all of whom were hospitalized alcoholics) who were taught controlled drinking techniques. Nonetheless, the Sobells were completely vindicated of any wrongdoing and their results accepted as legitimate outcomes. Subsequently, Nathan and McCrady (1987, p. 122), declared that, "the data from this [the Sobell's] study, which strongly suggest that behavioral treatment focussed on nonproblem drinking outcomes may have merits for some alcoholics, are convincing."

Both an initial Rand report and a four-year follow-up were also virulently attacked, beginning with press conferences convened by the National Council on Alcoholism when each of the reports was published. Nathan and Niaura (1985, p. 415) summarized what was learned from the second report (Polich, Armor, & Braiker, 1981) as follows: "(1) roughly one in five of patients followed through the four-year period who were both alive and could be interviewed were judged to be drinking without problems [this comprised 40% of the alcoholics in remission in the study], (2) nonproblem drinkers were no more likely than abstainers to manifest concurrent psychiatric problems, (3) nonproblem drinkers did not appear more likely than abstainers to relapse into problem drinking." Indeed, younger (under 40) unmarried alcoholics in this study, even when highly dependent on alcohol, were less likely to relapse when they adopted nonproblem drinking rather than trying too abstain.

Nathan and Niaura gave the Rand study high marks: "In terms of subject numbers, design scope, and follow-up intervals as well as sampling methods and procedures, the four-year Rand study continues at the state-of-the-art of survey research" (p. 416). But Nathan and Niaura discounted the relevance of the Rand results to clinical practice since the treatment centers that were studied did not practice controlled-drinking therapy (more than likely they actively discouraged it). This explanation for disregarding the Rand study contrasts with Nathan's (1984) reliance on Vaillant's (1983) natural history data as a principal reason for rejecting the value of controlled-drinking therapy. Indeed, considering only clinical outcome data, the conclusion of the Vaillant study is that abstinence treatment is useless. Vaillant compared eight-year outcomes in a clinical population which underwent abstinence and AA-oriented treatment and found them no better than those obtained with untreated alcoholics.

Interpreting The Body Of Evidence About Controlled-Drinking And Abstinence Therapy

A large number of studies in addition to the Sobells' study have found positive results from controlled-drinking therapy. Evaluations of the validity and import of this work differ. Nathan and Niaura (1985), Nathan (1986), and Nathan and McCrady (1987) indicated that this body of work suffers from methodological flaws that invalidate its results. Heather and Robertson (1983), on the other hand, quantified the strengths and weaknesses of these studies and found consistent and strong indications of beneficial impact from moderation treatment. Miller and Hester's (1986) review of all controlled studies of alcoholism treatment affirmed Heather and Robertson's conclusion that controlled-drinking therapy is a valuable technique. Indeed, Miller and Hester claimed, no other alcoholism therapy had been so frequently assessed and found to have positive results.

Despite their overall positive assessment, Miller and Hester concluded that controlled-drinking therapy "is not an effective method for chronic alcoholics who are severely dependent" (p. 148). They base this conclusion on two studies. Oddly, one of these--by Pendery, Maltzman, and West (1982)--contested the results of Sobell and Sobell (1973, 1976), the one study of the benefits of controlled-drinking therapy which Nathan and McCrady (1987) found persuasive. The other study cited by Miller and Hester is Foy, Nunn, and Rychtarik (1984), the most recent systematic clinical trial of CD therapy conducted in the United States with a severely alcoholic population. It is also the only American study in addition to Sobell and Sobell (1973, 1976) directly to compare CD and abstinence-oriented treatment for clinical alcoholics.

Foy et al. (1984) reported that a group of chronic alcoholics treated with behavioral techniques oriented toward moderation training were more likely to be abusing alcohol after six months than a comparison group treated with an abstinence goal. This finding, however, was not significant at the end of a year. Moreover, at a follow-up from years five to six reported by Rychtarik, Foy, Scott, Lokey, and Prue (1987, p. 106), "results showed no significant differences between [treatment] groups on any dependent [outcome] measure." Furthermore, at the longer follow-up, 18.4% of this group of chronic alcoholics were controlled drinkers by quite strict standards while 20.4% were abstaining.

Rychtarik et al. did not find that controlled drinking was any more likely for subjects treated with that goal than for subjects treated with an aim of abstinence. This study thus supported Nathan's (1986) assertion that behavioral techniques have not been shown reliably to produce moderation, or even to produce moderation more often than the absence of moderation training does. But the Foy/Rychtarik study also did not find that abstinence training reliably produced abstinence. Nor, in this study, did abstinence prove to be any more or less stable an outcome for alcoholics than controlled drinking. Only 4% of all subjects abstained and 4% moderated their drinking consistently over the five years of follow-up. The dominant pattern for alcoholics in the Foy/Rychtarik study during the follow-up was to shift among the abstinence, moderate drinking, and alcoholic categories.

This instability in drinking outcomes is similar to that found in the Rand study (where only 7% of subjects abstained throughout the four years of follow-up) and the Vaillant (1983) outcome study for hospitalized alcoholics, only 5% of whom did not relapse to alcoholic drinking over the eight-year follow-up period. Although the Foy/Rychtarik study was a clinical trial and the Rand study a survey of treatment outcomes, the two strongly agree in what they reveal about the course of alcoholism and its treatment (as do Sobell & Sobell, 1973, 1976). Alcoholism is not a progressive disease, but a persistent and recurrent problem. Small minorities of alcoholics resolve their drinking problem during treatment through moderation or through abstinence, whatever the goal of the therapy they receive is.

The research on treatment outcomes thus fails to find that any therapy yields reliable, consistent remission patterns for severely alcoholic individuals. Nathan and Niaura (1985), Nathan (1986), and Nathan and McCrady (1987) do not cite a single study that has found abstinence therapy to be more successful than moderation training for any group of alcoholics. On the other hand, Nathan and his colleagues accept the results of the Sobells' research, a direct comparison which found better outcomes for CD that from abstinence-oriented therapy for chronic alcoholics. There seems to be no reason for Nathan et al. to disparage CD therapy that does not apply to alcoholism therapy in general. Nonetheless, Nathan and Niaura (1985, p. 418) concluded: "abstinence-oriented therapy works for many persons while the success of controlled-drinking treatment, especially for alcoholics, is problematic...."

Is Abstinence The More Stable And Preferable Outcome For Severely Dependent Alcoholics?

In the late 1970s and early 1980s, a consensus emerged in the field of alcoholism (a consensus that includes not only Nathan, but those who advocate and practice controlled-drinking therapy such as William Miller and Alan Marlatt) that severely dependent alcoholics almost never achieve moderation, or else that they are so much less able to do so than they are to achieve abstinence that CD therapy is irrelevant to truly alcoholic individuals (Foreyt, 1987, Peele, 1984; 1987b). While Miller and Hester (1986) cited the Foy/Rychtarik study to support this consensus, the study actually disputes it. Nor is it the only study in the 1980s to do so.

Two studies of treated alcoholics conducted in Europe--McCabe (1986) in Scotland and Nordström and Berglund (1987) in Sweden--found that alcohol-dependent individuals displayed substantial moderation outcomes over follow-up periods of 16 years to two decades or more. Furthermore, alcoholics in these studies became better able to moderate their drinking the longer after their treatment they were examined. Nordström and Berglund found that for their subjects, all of whom had been hospitalized and were categorized as alcohol-dependent, controlled drinking was less likely to lead to relapse over time than was abstinence.

Vaillant's (1983) long-term American study of untreated alcohol abusers contradicts the results from the two European studies in finding that severely dependent drinkers were highly likely to relapse over time when they attempted to control their drinking. Nathan, for one, has found Vaillant's results convincing. One possible explanation for the different results in these three studies (aside from the European/American split) is definitional. Vaillant categorized an alcoholic with a single symptom of alcohol dependence (such as morning drinking) over the prior year as actively alcoholic even if he drank moderately the rest of the year. If an alcohol abuser abstained for most of the year (or drank less than once a month), and had a drinking binge lasting no more than a week in the year, Vaillant counted the subject as abstinent (cf. Peele, 1987b).

Nordström and Berglund found that abstinence versus controlled-drinking outcomes were not related to their subjects' level of alcohol dependence. Rychtarik et al. (1987) likewise reported no relationship between severity of alcohol dependence and a controlled-drinking or abstinence resolution of a drinking problem. Another long-term treatment outcome study (this one in England) by Edwards et al. (1988) also did not find that severity of dependence or number of drinking problems at intake predicted drinking behavior 10 years after treatment. (This important study instead showed that meaningful relationships only appeared when complex, multifactor outcomes were considered.)

Two European (British) clinical studies have tested the idea that CD therapy is appropriate only for less severely dependent alcoholics by comparing level of alcohol dependence or severity of alcoholism with patient expectations as predictors of abstinence and controlled drinking. Both these studies found subjects' beliefs and commitment to attaining an outcome to be superior to clinical measures of subjects' alcohol dependence at predicting which outcome the person achieved. Orford and Keddie (1986, p. 495) reported "no support was found for the dependence hypothesis:...there was no relationship between level of dependence/severity and the type of drinking outcome (ABST or CD)." Instead, patients' "persuasion" that they could achieve one form of remission over the other was more important for the outcome actually achieved.

Elal-Lawrence, Slade, and Dewey (1986) also found no relationship between severity of drinking problems and type of remission following treatment. The authors concluded "that alcoholism treatment outcome is more closely associated with the patients' own cognitive and attitudinal orientation, past behavioral expectations, the experience of abstinence and the freedom of having his or her own goal choice.... This may be the time to act with caution before reaching another... conclusion that only the less severely dependent problem drinkers can learn to control their drinking." Both the Orford and Keddie and Elal-Lawrence et al. results support an earlier British finding that alcoholics who rejected the axiom "one drink, one drunk" were more likely than those who accepted it to be drinking moderately at a six-month follow-up to treatment (Heather, Winton, & Rollnick, 1982).

Studies finding subjective beliefs about one's drinking to play an important--perhaps the most important--role in therapy outcomes do not seem surprising given the importance Nathan and his colleagues (and many other psychologists) attach to the role of expectations in alcoholic behavior. Yet, many psychologists do not extend these views of the importance of subjective states into the realm of therapy practice. Quite the opposite, Nathan and McCrady (1987) argued against considering patient preference or beliefs about controlled drinking: "The client's hesitance about abstinence and the potential benefits of abstinence....can be framed as an 'irrational' belief, similar to many irrational beliefs which therapists challenge in the course of therapy" (p. 124).

Why Reject Controlled Drinking In American Treatment?

Nathan has expressed positive attitudes towards controlled-drinking therapy as recently as a book chapter published in 1982. In describing his private practice, Nathan (1982) talked of his belief in the need for controlled-drinking treatment:

In central New Jersey there are many person, groups, and institutions that offer help to alcoholics who wish to stop drinking. Though they vary in competence, cost, and commitment, they are in abundance and they make themselves visible....

Given this richness of treatment resources for alcoholics, problem drinkers, and their families in central New Jersey, it is natural to ask whether any gaps in the range of services exist in this area. I think such a gap does exist--and I undertook to fill it...

Since I am thoroughly familiar with the field of nonproblem-drinking treatment, and since I want my very limited private practice to be stimulating and interesting, I have decided to restrict it to persons who want to control their drinking and have the resources to do so. (pp. 107--108, emphasis added).

Nathan then described his treatment of a woman whose drinking was endangering her marriage. This client, whom he called "Louise," successfully reduced her drinking while greatly improving her marriage. "Both the reduced drinking and the improved interpersonal habits appear to be stable." At the same time, Nathan ruminated, the woman probably still drank too much and had other personal problems--"It seems reasonable to assume, however, that Louise would not have changed her habitual ways of interacting with the world even if she had been able to stop drinking" (p. 123).

Nathan's endorsement and use of the controlled-drinking therapy option based on his informed view of the research (cf. Nathan, 1980) contrast markedly with his current attitudes. However, since Nathan primarily opposes CD therapy for "chronic alcoholics," it is possible that his treatment of this case is consistent with his current position, assuming that Louise is only mildly alcohol dependent. But (although Nathan presented no systematic assessment of her alcohol dependence) Louise did not seem to be a mild problem drinker: her father was "almost certainly alcoholic," she had had drinking problems throughout her life, she had been hospitalized for alcoholism after which she attended AA for several years, and she drank "between a pint and a fifth of vodka" nightly when Nathan began treating her.

Nathan agreed on a controlled-drinking goal for Louise because Louise rejected abstinence, because her husband supported her commitment to reduced drinking, and because she was not a totally abandoned alcoholic. But in exercising some control over her drinking, Louise resembled most alcoholics, including the heavily alcoholic populations in the Rand study and other research this article has examined. Such heavily dependent drinkers frequently moderate their drinking or abstain for varying lengths of time. In other words, saying only that alcoholics clearly worse off than Louise should eschew a controlled-drinking goal is to leave a very large alcoholic population for whom CD therapy can be appropriate. In Louise's case, Nathan accepted the claim of a client (one who had been active in Alcoholics Anonymous) that she would not quit drinking.

In a frank article entitled, "Ideal Mental Health Services for Alcoholics and Problem Drinkers: An Exercise in Pragmatics," Nathan (1980) described how political issues sabotaged a controlled-drinking program he tried to develop. Nathan was asked by the chairman of a hospital board--who was "himself a member of Alcoholics Anonymous" but who was "convinced that AA was not the only answer to alcoholism"--to create a controlled-drinking program for "problem drinkers." Nathan agreed to develop such a program with colleague Terry Wilson "because I felt that development of the program would be a worthwhile service and that a great deal of valuable data on the kinds of alcoholics and problem drinkers responding positively to controlled drinking treatment would be gathered."

Nathan (1980) reviewed evidenced on the effectiveness of CD therapy, carefully developed a treatment protocol and aftercare program, and then tried to implement the program. The hospital submitted the program to a review which resulted in requests for "a series of minor but annoying changes." Nathan and Wilson felt these changes expressed the ambivalence of the hospital toward a "treatment designed to convert...uncontrolled problem drinkers to controlled social drinkers, a treatment goal totally at variance with the abstinence goal of Alcoholics Anonymous," which was an important constituency for the hospital's inpatient program. Nathan and Wilson reiterated that the program was geared solely towards problem drinkers, and not alcoholics.

Finally, the program went ahead with a carefully worded press release describing the program and seeking clients. When this announcement produced only a small group of potential clients, Nathan and Wilson planned a more aggressive advertising effort. Before they could proceed, however, the hospital withdrew its support for the program because of threats by representatives of AA that they would stop referring patients to the hospital's profitable inpatient program. Nathan (1980, p. 296) concluded, "Above all, do not assume that an ideal mental health service for alcoholics and problem drinkers can survive the pragmatics of the marketplace simply by the force of its empirical justification."

Nathan's experience in this case illustrates that, although the idea of CD therapy for problem drinkers is apparently less controversial than it is for chronic alcoholics, it remains difficult to implement a program in the United States that applies CD therapy with any group of alcohol abusers. As Nathan (1980) made clear, co-existence of abstinence and controlled-drinking programs under the same aegis seems politically impossible. In addition, every attempt to identify problem drinkers for whom CD therapy is indicated also nets drinkers who are measured as alcohol dependent by standard diagnostic instruments (Heather, 1986; Sanchez-Craig & Lei, 1986). Furthermore, in the United States, people with any sort of drinking problem will usually be diagnosed and treated as alcoholics (Hansen & Emrick, 1983).

As a result, according to an interview Nathan gave in 1982 about the Sobell-Pendery dispute, "there is no alcoholism center in the United States using the technique [CD therapy] as official policy" (Fisher, 1982, p. 8). This situation may be unique to the United States. In Britain, Heather and Robertson (1982) found that over three-quarters of National Health Service alcohol treatment units offered CD therapy. Moreover, including even Councils on Alcoholism, over 90% of the surveyed units thought CD therapy was appropriate for some clients whether or not the units practiced such therapy. In 1990, the same attitudes prevailed in the United Kingdom (Rosenberg, Melville, Levell, & Hodge, 1992). Rush and Ogborne (1986) found 37% of Canadian treatment units offered CD therapy, including about a fifth of inpatient and detoxification facilities along with a majority of community outreach centers.

Institutional, Economic, And Legal Factors In The Rejection Of CD Therapy

Nathan's (1980) description of his failed attempt to establish a controlled-drinking program appeared in the same year the second volume of the Rand study reported a four-year follow-up of NIAAA-funded alcoholism treatment center outcomes. Like the initial 18-month follow-up, the second Rand report found substantial nonproblem drinking remission among alcoholics. Like the first report, it was loudly criticized by representatives of the alcoholism treatment industry, including a concerted attack organized by the National Council on Alcoholism. Moreover, the two different directors of the NIAAA at the time of the two reports both affirmed that, despite the Rand results, abstinence was the "appropriate goal in the treatment of alcoholism" (Peele, 1984). In 1982, when Nathan's description of his private CD therapy practice was published, the Pendery et al. assault on the Sobells' CD treatment research appeared.

On accepting the Rutgers Center directorship in 1983, Nathan had to deal with an important defection by a major financial supporter of the Center. According to Lender and Martin (1982), "when the Journal of Studies on Alcohol published a number of articles with conclusions similar to Rand's, one major private contributor to alcoholism research proceeded to cut the Rutgers Center of Alcohol Studies, the Journal's publisher, out of his will" (p. 193). This was R. Brinkley Smithers, a recovering alcoholic who was president of the National Council on Alcoholism until 1965, and who had donated $30 million to the alcoholism field through either personal donations or through the Christopher D. Smithers Foundation (named for his father).1

Smithers was the Center's largest private benefactor (Smithers Hall, the Center building, was named after the elder Smithers after the son gave the Center a grant when it moved from Yale in 1962). By 1983, when Nathan took over the reins at Rutgers, Smithers' relationship with Rutgers had cooled dramatically. By 1986, the relationship had been restored, and Smithers personally gave Rutgers $3.54 million as a part of an overall $6.7 million gift for alcohol studies.2 Smithers had proved himself to be a demanding donor in the past, willing to withhold funds from a program that did not adhere to his views on alcoholism (including that "the goal of treatment should be total abstinence from all addicting substances") in connection with a previous $10 million grant to Roosevelt Hospital (Minetree, 1986, p. 243).3

I don't believe that Nathan changed his views on controlled drinking (which had begun shifting prior to his assuming the directorship of the Center) in order to secure a large private donation for the Rutgers Center. Rather, I think Nathan was influenced by a range of pressures to jettison CD therapy. Given its less-than-decisive benefits, CD therapy had simply developed too many debits for Nathan to support. The pressures Nathan faced, which every behaviorist in the field experiences, had a double impact on Nathan as an individual researcher and therapist and as the director of a prominent alcoholism center whose constituency extended outside the usual academic institutional and funding boundaries.

Nathan's shift in views could hardly not have been influenced by factors other than the scientific evidence on the issue, given the climate of the times. These factors included his discouraging experience trying to establish a CD clinic, the outcry over the Rand study, the assault on CD therapy represented by Pendery et al. (1982), the clear disapproval of CD therapy expressed by the NIAAA hierarchy, and the exigencies of the Rutgers Center's funding needs. In his 1980 and 1982 articles, Nathan was positive about the value of CD therapy and felt that the research--with which he was thoroughly familiar--supported this position, wanted to make CD therapy more available, and was open to exploring the usefulness of CD therapy for a wide range of alcoholics. A few years later, while the balance of the research hardly had a chance to change, Nathan and Niaura (1985, p. 418) declared that, "for a variety of reasons, abstinence ought to be the goal of treatment for alcoholism."

Nathan and McCrady (1987, p. 127) described why they thought an abstinence goal in alcoholism treatment is necessary based on their "examination of the literature on controlled drinking, personal value issues, and the reality of the current system of delivery of health care services to alcoholics." One such reality is a legal one in the form of the therapist's

...liability for possible aversive [sic] consequences resulting from the client's continued drinking. Since the cultural and clinical norm in the alcoholism treatment field is for abstinence as the only treatment goal, it is possible that the therapist who works with a controlled drinking goal might be liable to civil suits.... Clearly, the goal of controlled drinking treatment is moderate, not excessive use, but this discrimination may be blurred in the eyes of the prosecutor or jury. [The authors' reference to "prosecutor" may indicate they wished to warn controlled-drinking therapists about criminal liability as well as civil suits.]

Consequences Of Current Trends In CD Therapy In America

Despite expressing negative attitudes towards CD therapy, Nathan and Niaura (1985, p. 418) concluded that "what is needed most, before rational decisions on treatment goals can be made for anyone, is additional comparative research on the applicability, efficacy, and permanence of the results both of abstinence-oriented and controlled-drinking-oriented treatment of alcoholic... [and] problem drinkers...." Such research is not forthcoming in the United States, and the NIAAA is currently not funding any CD therapy trials.4

What is more, behaviorists working in the area of alcoholism treatment often seem unwilling to consider the evidence on this question. For example, Monti et al. (1989) ignored Orford and Keddie's (1986) and Elal-Lawrence et al.'s (1986) evidence that patient preferences predict the most likely form remission will take. These authors' blind spot may well have been an asset, however, in the adoption of their manual for use in skills-training treatment trials sponsored by the NIAAA.

At a clinical level, the rejection of controlled-drinking as one goal for treatment promotes the heavily prescriptive approach Nathan and McCrady recommended (see above). Since controlled drinking is not an acceptable goal for them, Nathan and McCrady attack any desire by an alcoholic to continue drinking as "irrational." This approach has until now been associated with the disease (and, before that, the temperance) approach to alcoholism. An alternative approach is described by Sanchez-Craig and Wilkinson (1987, p. 307):

Our practice is to inform clients that there are divergent views about the nature of alcoholism.... we adhere to the view that excessive drinking is to a large extent a learned behavior that can be modified....[and that] total abstinence is not essential to solve the problem. We recommend to our clients that they evaluate these options in light of their own experiences, and that they decide whether the assumptions underlying our program are consistent with their beliefs.

This view might be termed client-centered, or respectful of the problem drinker's integrity and values, or as recognizing the reality and importance of coordinating client beliefs with the therapy approach.

The anti-controlled-drinking position as adopted by Nathan and colleagues and by other social-learning theorists in the alcoholism field has other implications. It represents a withdrawal from iconoclastic scientific positions that contradict prejudices and prevailing attitudes in the field. According to Nathan and McCrady (1987, p. 127), "instead of ... defending belief systems ... behavior therapists could more broadly influence the field, and learn a lot more about alcohol abuse, by looking for points of interaction." This point, however raises further issues, such as:

  1. Can the dominant disease-oriented alcoholism industry accommodate psychologically-based outlooks, techniques, and respect for research, or will psychologists be asked to modify more than their openmindedness towards controlled drinking. Nathan and McCrady (1987), for example, represent a serious turn away from the kind of client-centered psychological interviewing in which the therapist accepts and builds on the client's belief system, thereby maximizing the person's motivation to change. (See Miller, 1989, for an approach to clinical interviewing that runs counter to that recommended by Nathan and McCrady.) Another example of a basic conflict between disease and psychological notions is over the need to force problem drinkers to acknowledge they are alcoholics (overcome their "denial") before they can proceed in treatment. Traditionally, psychology has been wary of encouraging such self-labelling because people frequently live up to their labels. Moreover, the disease approach regards "alcoholic" as a life-long identity, and many psychologists (rightfully) oppose this type of identity-conversion-cum-therapy. Obviously, such self-labelling is tied to the possibility of controlled drinking, for if one is always an alcoholic one can never drink moderately.
  2. How successful are the prevailing approaches to which behaviorists are being asked to assimilate? What if behaviorists are asked to adopt positions which they suspect, or know, to be counterproductive? What if some subset of alcoholics stands a better chance to improve if they pursued a controlled-drinking goal? Are there legal and ethical issues involved in overlooking this possibility, which at least some recent research suggests may hold true? If a problem drinker with a long history of moderate drinking is taught to abstain and subsequently experiences life-threatening drinking episodes (one thinks of Kitty Dukakis in this regard), can the abstinence-oriented therapist be held legally liable?
  3. If the alcoholism treatment industry is finally challenged by insurers, HMOs, and government agencies and rejected in many of its fundamental beliefs and its overall efficacy, will psychology have lost its ability to provide an independent approach to replace discredited treatments?
  4. Who will be responsible in the United States for developing and supporting a research-based, scientific body of accurate information on alcoholism and addiction prevention and treatment, even when these ideas run counter to prevailing wisdom? Who will alert society to fresh ideas in the field that might help stem the tide of addiction?

While we wait for a change in the cultural climate, for treatment costs to become intolerable, and for the high failure rate and other costly drawbacks of the disease model to manifest themselves in undeniable ways, the best psychological work is not being incorporated into ongoing treatment and prevention efforts. One response to this dilemma is to mold our views and techniques to conform with conventional wisdom, or at least not to draw undue attention to our disagreements with current dogma. This will help psychologists to gain research funding, reimbursement for treatment, and career advancement--all estimable goals. But will psychology cease to welcome those who choose not to take this route, and become just one more branch of an intellectually bankrupt status quo?


  1. This identification was confirmed for me by a reliable source at Rutgers.
  2. Telephone interview with Peg Baker, July 16, 1990, and Rutgers University press release, April 22, 1986. (The complete history of these negotiations is yet to be written.)
  3. It is important to note that the Journal continued during Nathan's tenure to publish studies indicating controlled drinking is both a common outcome and worthwhile therapeutic goal (Elal-Lawrence et al., 1986; Nordström & Berglund, 1987). For what it is worth, the Journal also published my work (Peele, 1986; 1987a) and, indeed, presented me the 1989 Keller Award for the best article to appear during 1987-1988. At the same time, neither the Center nor any of its staff publicly conducted controlled-drinking therapy or research during Nathan's administration.
    The Smithers/Rutgers relationship calls to mind ethical questions that have been raised by private corporate donations to universities (Sykes, 1988). For example, corporations have funded laboratories conducting research in areas of concern to the companies. Even when the corporation does not dictate the research agenda, its financial stake influences the direction of research at the laboratory. Obviously, funding of university programs by groups with political agendas endangers the integrity of the university even more. Where on this spectrum does funding by an opinionated private donor on a highly controversial issue where researchers have previously been intimidated fall?
  4. When the National Heart, Lung, and Blood Institute sought to test a moderation-training program for heavy and problem drinkers who suffered from hypertension, the NIAAA selected for administration a CD program designed and supervised by Martha Sanchez-Craig, of the Ontario Addiction Research Foundation in Toronto, because no suitable program was available in the United States. This is the only research on a clinical application of CD therapy currently being funded by the NIAAA.


I am indebted to Harold Rosenberg for his careful reading of this manuscript, to an anonymous reviewer who gave me several valuable insights, and to Peg Baker and other Center of Alcohol Studies staff. Only I am responsible, however, for the ideas expressed in this paper.


Bigelow, G., Liebson I.A., & Griffiths, R. (1974). Alcoholic drinking: Suppression by a brief time-out procedure. Behavior Research and Therapy, 12, 107-15.

Cohen, M., Liebson, I.A., Faillace, L.A., & Allen, R.P. (1971). Moderate drinking by alcoholics: A schedule-dependent phenomenon. Journal of Nervous and Mental Disease, 153, 434-444.

Edwards, G., Brown, D., Oppenheimer. E., Sheehan, M., Taylor, C., & Duckitt, A. (1988). Long-term outcome for patients with drinking problems: The search for predictors. British Journal of Addiction, 83, 917-927.

Elal-Lawrence, G., Slade, P.D., & Dewey, M.E. (1986). Predictors of outcome type in treated problem drinkers. Journal of Studies on Alcohol, 47, 41-47.

Fisher, K. (1982, November). Debate rages on 1973 Sobell study. APA Monitor, 8-9.

Foreyt, J.P. (1987). The addictive disorders. In G.T. Wilson, C.M. Franks, P.C. Kendall, & J.P. Foreyt (Eds.), Review of behavior therapy (Vol. 11, pp. 187-223). New York: Guilford.

Foy, D.W., Nunn, L.B., & Rychtarik, R.G. (1984). Broad-spectrum behavioral treatment for chronic alcoholics: Effects of training controlled drinking skills. Journal of Consulting and Clinical Psychology, 52, 218-230.

Gottheil, E., Murphy, B.F., Skoloda, T.E., & Corbett, L.O. (1972). Fixed-interval drinking decisions: II. Drinking and discomfort in 25 alcoholics. Journal of Studies on Alcohol, 11, 325-340.

Hansen, J., & Emrick, C.D. (1983). Whom are we calling "alcoholic"? Bulletin of the Society of Psychologists in the Addictive Behaviors, 2, 164-178.

Heather, N. (1986). Change without therapists: The use of self-help manuals by problem drinkers. In W.R. Miller & N.K. Heather (Eds.), Treating addictive behaviors: Processes of change (pp. 331-359). New York: Plenum.

Heather, N., & Robertson, I. (1983). Controlled drinking. London: Methuen.

Heather, N., Winton, M., & Rollnick, S. (1982). An empirical test of "a cultural delusion of alcoholics." Psychological Reports, 50, 379-382.

Lender, M.E., & Martin, J.K. (1982). Drinking in America. New York: MacMillan.

Marlatt, G.A., Demming, B., & Reid, J.B. (1973). Loss of control drinking in alcoholics: An experimental analogue. Journal of Abnormal Psychology, 81, 233-241.

McCabe, R.J.R. (1986) Alcohol-dependent individuals sixteen years on. Alcohol & Alcoholism, 21, 85-91.

Mello, N.K., & Mendelson, J.H. (1971). A quantitative analysis of drinking patterns in alcoholics. Archives of General Psychiatry, 25, 527-539.

Meyer, W.R. (1981). Discussion of papers by Ludwig, Mello, and Nathan. In Evaluation of the alcoholic: Implications for research, theory, and treatment (DHHS Publication No. ADM 81-1033, pp. 243-248). Washington, DC: U.S. GPO.

Miller, W.R. (1989). Increasing motivation for change. In R.K. Hester & W.R. Miller (Eds.), Handbook of alcoholism treatment approaches: Effective alternatives (pp. 67-80). New York: Pergamon.

Miller, W.R., & Hester, R.K. (1986). The effectiveness of alcoholism treatment: What research reveals. In W.R. Miller & N.K. Heather (Eds.), Treating addictive behaviors: Processes of change (pp. 121-173). New York: Plenum.

Minetree, H. (1986, May). Alcoholism's sober philanthropist. Town & Country, pp. 184-186, 238-243.

Monti, P.M., Abrams, D.B., Kadden, R.M., & Cooney, N.L. (1989). Treating alcohol dependence. New York: Guilford.

Nathan, P.E. (1980) Ideal mental health services for alcoholics and problem drinkers: An exercise in pragmatics. In P.O. Davidson & S.M. Davidson (Eds.), Behavioral medicine: Changing health lifestyles (pp. 279-298). New York: Brunner/Mazel.

Nathan, P.E. (1982). Louise: The real and the ideal. In W.M. Hay & P.E. Nathan (Eds.), Clinical case studies in the behavioral treatment of alcoholism (pp. 107-125). New York: Plenum.

Nathan, P.E. (1984). The length and breadth of alcoholism. (Review of G.E. Vaillant, 1983, The natural history of alcoholism.) Contemporary Psychology, 29, 101-103.

Nathan, P.E. (1986). Outcome of treatment for alcoholism: Current data. Annals of Behavioral Medicine, 8, 40-46.

Nathan, P.E., & McCrady, B.S. (1987). Bases for use of abstinence as a goal in the behavioral treatment of alcohol abusers. Drugs & Society, 1, 109-131.

Nathan, P.E., & Niaura, R.S. (1985). Behavioral assessment and treatment of alcoholism. In J.H. Mendelson & N.K. Mello (Eds.), The diagnosis and treatment of alcoholism (2nd. ed., pp. 391-455). New York: McGraw-Hill.

Nathan, P.E., & O'Brien, J.S. (1971). An experimental analysis of the behavior of alcoholics and nonalcoholics during prolonged experimental drinking: A necessary precursor of behavior therapy? Behavior Therapy, 2, 455-476.

Nordström, G., & Berglund, M. (1987). A prospective study of successful long-term adjustment in alcohol dependence. Journal of Studies on Alcohol, 48, 95-103.

Orford, J., & Keddie, A. (1986). Abstinence or controlled drinking: A test of the dependence and persuasion hypothesis. British Journal of Addiction, 81, 495-504.

Pattison, E.M., Sobell, M.B., & Sobell, L.C., (1977). Emerging concepts of alcohol dependence. New York: Springer.

Peele, S. (1984). The cultural context of psychological approaches to alcoholism: Can we control the effects of alcohol?  American Psychologist, 39, 1337­1351. 

Peele, S. (1986). The implications and limitations of genetic models of alcoholism and other addictions. Journal of Studies on Alcohol, 47, 63­73.

Peele, S. (1987a).  The limitations of control­of­supply models for explaining and preventing alcoholism and drug addiction. Journal of Studies on Alcohol, 48, 61-77.

Peele, S. (1987b). Why do controlled-drinking outcomes vary by country, era, and investigator?: Cultural conceptions of relapse and remission in alcoholism. Drug and Alcohol Dependence, 20, 173-201.

Pendery, M.L., Maltzman, I.M., & West, L.J. (1982). Controlled drinking by alcoholics?: New findings and a reevaluation of a major affirmative study. Science, 217, 169-175.

Polich, J.M., Armor, D.J., & Braiker, H.B. (1981). The course of alcoholism: Four years after treatment. New York: Wiley.

Rosenberg, H., Melville, J., Levell, D., & Hodge, J.E. (1992). A 10-year follow-up survey of acceptability of controlled drinking in Britain. Journal of Studies on Alcohol, 53, 441-446.

Rychtarik, R.G., Foy, D.W., Scott, T., Lokey, L., & Prue, D.M. (1987). Five-six-year follow-up survey of broad-spectrum behavioral treatment for alcoholism: Effects of training controlled drinking skills, Journal of Consulting and Clinical Psychology, 55, 106-108.

Sanchez-Craig, M., & Lei, H. (1986). Disadvantages of imposing the goal of abstinence on problem drinkers: An empirical study. British Journal of Addiction, 81, 505-512.

Sanchez-Craig, M., & Wilkinson, D.A. (1987). Theory and methods for secondary prevention of alcohol problems. In W.M. Cox (Ed.). Treatment and prevention of alcohol problems: A resource manual (pp. 287-331). New York: Academic Press.

Sobell, M.B., & Sobell, L.C. (1973). Alcoholics treated by individualized behavior therapy: One year treatment outcomes. Behavior Research and Therapy, 11, 599-618.

Sobell, M.B., & Sobell, L.C. (1976). Second year treatment outcomes of alcoholics treated by individualized behavior therapy: Results. Behavior Research and Therapy, 14, 195-215.

Sykes, C.J. (1988). Profscam: Professors and the demise of higher education. Washington, DC: Regnery Gateway.

Vaillant, G.E. (1983). The natural history of alcoholism. Cambridge, MA: Harvard University Press.


The idea that Peter Nathan sold out CD therapy to Smithers as Director at Rutgers was not an original one with me -- Alan Marlatt mentioned it to me before I wrote the article. I later tried to gain access to any records the Center kept concerning the Smithers grant to Rutgers. Although Rutgers is a State institution (and I a resident of that State), Robert Pandina, as director of the Center, and Rutgers university counsel Beckman Rich refused me this access.

I have spoken with a number of graduate or post-graduate students at Rutgers, many of whom who had co-authored pieces with Peter Nathan indicating that CD therapy was dead. All these former students -- Ray Niaura, David Abrams, Jim Langenbucher, Stephen Lisman, and Mark Goldman -- told me that this was not their position. They apparently acquiesced in Nathan's view as a price for junior authorship in the various prestige publications that Nathan commanded.

Barbara McCrady, on the other hand, had always on her own opposed controlled drinking, even before coming to Rutgers as a junior faculty member. Thus she contributed some of the most unyielding expression of the anti-controlled-drinking position Nathan and McCrady expressed:

I question why training alcoholics in controlled drinking is so attractive. People who earn an alcoholism diagnosis have worked hard for it.... Instead of trying to help alcoholics drink, therapists might better view such a desire to drink as an "irrational belief." (McCrady, 1985, p. 370).

My article was a part of an exchange in Addictive Behaviors. Peter Nathan, Nick Heather, Bill Miller, Mark Sobell, and Barbara McCrady -- who was acting head of Rutgers at the time -- were asked to comment on my piece, and I to respond to their comments. McCrady, in her comment, made me out as someone who attacked abstinence therapists. Yet, at the time of the Addictive Behaviors exchange and since then, I have accepted and practiced abstinence therapy in addition to CD treatment. The converse apparently could not be said for McCrady.

Moreover, the idea that I as an individual without an institutional affiliation could oppress those supporting the dominant position in the field of alcoholism is a strange view of events in the eighties, when the forces against CD therapy smote Rand, the Sobells, and anyone else who dared challenge their hegemony over alcohol treatment. As I detailed elsewhere (Peele, 1986), my own career was endangered by my maintaining the value of controlled-drinking therapy. That I should have to insist on the integrity of research and clinical findings that an academic institution such as Rutgers denied seems a reversal of our appropriate roles.

This is all the more true since, within a few years of the AB exchange, Rutgers initiated a brief intervention clinic which accepted moderation outcomes. It also practiced the kind of client-centered therapy--where client beliefs and preferences are taken as indicators of the paths to follow in therapy--that I outlined in the above article and that I practice myself. (Actually, I understand from Fred Rotgers that the Rutgers program is not this open, and that it encourages nearly all clients to abstain.) As my article also predicted, the Smithers Foundation has now refused to entertain proposals for grants from Rutgers (R. Brinkley Smithers is dead and the Foundation is now headed by his wife, Adele).

But Barbara McCrady, in response to my claim that she had come around to my viewpoint, has insisted that my writings (which Barbara always refers to as "think" pieces) did not influence her thinking, but that instead the data dictated her change of course. The work McCrady specifically credits was by Harold Rosenberg (1993). It, however, is also a think piece in the sense of not comprising original research. Rosenberg's article, which was published shortly after mine, covers much similar ground, and refers to my own earlier analysis of the CD data (cf. Peele, 1987). Harold also supportively read and commented on my Addictive Behaviors article.

I wonder how Peter Nathan regards the change in Rutgers' position on CD therapy after he left to become Vice President for Academic Affairs at the University of Iowa. I thus sent him the following letter in September, 1996.

Dear Peter:

Perhaps you are aware of developments at Rutgers since you left there. The Center of Alcohol Studies has now created a brief intervention clinic which practices CD therapy! Moreover, rather than challenging client goals, the program seeks to work with them, which was the basis of our fundamental disagreement about alcoholism treatment expressed on the pages of Addictive Behaviors. In addition, in line with my article for AB, it seems that Smithers has ceased funding the Center. Meanwhile, Barbara says she hasn't really accepted my approach, but rather has responded to the data expressed in Rosenberg (1993), which is really the same data to which I refer in my article.

What I'm wondering is: (1) Why did Rutgers change its direction after you left? (2) Do you agree with this new direction, or do you think they have made an error? (3) What do you think about Smithers' rejection of the Center as a result of these changes? (4) Finally, don't you think that my point of view has won out, just as I predicted in my response to your and Barbara's comments on my article, when I wrote, "I fully expect that my views will be accepted eventually" (Peele, 1992, p. 89).


McCrady, B. (1985). Comments on the controlled-drinking controversy. American Psychologist, 40, 370-371.

Peele, S. (1986). Denial — of reality and freedom — in addiction research and treatment. Bulletin of the Society of Psychologists in the Addictive Behaviors, 5, 149-166.

Peele, S. (1987). Why do controlled-drinking outcomes vary by investigator, by country and by era? Drug and Alcohol Dependence, 20, 1783-201.

Peele, S. (1992). Why is everybody always pickin' on me? A response to comments. Addictive Behaviors, 17, 83-93.

Rosenberg, H. (1993). Prediction of controlled drinking by alcoholics and problem drinkers. Psychological Bulletin, 113, 129-139.