Does Marijuana Lead to Dangerous Drugs?
The Appendix to Drugs in American Society, First Edition, by Erich Goode,
Professor of Sociology at the State University of New York at Stony Brook.
©1972 Alfred A. Knopf
During 1971 and 1972 I was a consultant for the National Commission
on Marihuana and Drug Abuse, which was mandated in 1970, under
the provisions of the Comprehensive Drug Abuse Prevention and
Control Act, to conduct an extensive study on the effects of marijuana.
I was asked by the commission to conduct research and write monographs
on two fundamental marijuana issueswhether marijuana use "caused"
violent and criminal behavior, and whether it "led to"
the use of more dangerous drugs, notably heroin. A number of other
studies were conducted by other researchers. I would like to summarize
the findings and conclusions from my monograph on the use of marijuana
and dangerous drugs, which
was itself a summary of dozens of surveys and studies on this
issue. Since the original monograph was almost volume-length,
I will present only its conclusions.
Of the many surveys on the use of marijuana and other drugs, probably
all have demonstrated an empirical association. The marijuana
user has a statistically higher likelihood of trying and using
a wide range of dangerous drugs than is true of the nonuser; frequency
of marijuana use is similarly associated with dangerous drug use.
The basic problem, however, is not the existence of this relationship
but just what it means in a causal sense. There are many statistical
associations that are in no way causally related. For instance,
one observer has pointed out that the drinking of tomato juice
is statistically correlated with becoming intoxicated (Kupperstein
1971). Clearly, tomato juice has no intoxicating properties whatsoever;
the fact that it is often mixed with vodka in "Bloody Marys"
produces an accidental or spurious association between
the ingestion of tomato juice and intoxication. Another study
found that coffee drinkers are more likely to use dangerous drugs
than individuals who do not drink coffee (Blum et al. 1969, pp.
57-58). Does this indicate a causal relationship between the ingestion
of coffee and using drugs? Many studies have shown a correlation
between liberal politics and ideology and using marijuana. Does
using marijuana "cause" someone to adopt liberal politics?
Certainly agreeing with these propositions would be fallacious.
An amusing book of two decades ago, entitled How to Lie with
Statistics (Huff 1954), discussed dozens of such methodological
and factual blunders in reasoning. We must be careful not to make
similar errors in exploring the marijuana-dangerous drug link.
It would be erroneous to automatically translate the statistical
correlation into a causal process. Thus we want to know whether
the relationship between marijuana and dangerous drugs is spurious
that is, an incidental feature of some other relationshipor
whether marijuana use actually causes the use of other,
more dangerous drugs.
There are a number of different and to some degree contradictory
explanations of the cause behind this statistical association.
The first is the biochemical, or "effects," theory,
which claims that it is the experience of the marijuana intoxication
itself that provides the dynamic force in drug "escalation"
This biochemical explanation has several variants. One is that
being high on marijuana provides an introduction to the drug experience,
thus stimulating a "thirst" for bigger and better "thrills."
A former director of the Bureau of Narcotics and Dangerous Drugs
has written: "The evidence is strong that the use of marihuana
develops a taste for drug intoxication which, in turn, leads many
people to the use of more potent drugseven heroin" (Giordano
1968, p. 5).
A variant of the biochemical theory is the tolerance disillusionment
explanation. According to this line of reasoning, initial experimentation
with marijuana leads to ever greater involvement with the drug;
this in turn leads to diminishing effects. Eventually, the marijuana
user finds the experience of becoming intoxicated on the drug
unexciting, banal, and less pleasurable; hence he turns to more
potent drugs to achieve the high he initially felt, but can no
longer achieve, with marijuana. A proponent of this view is the
Attorney General of the United States, John Mitchell, who was
quoted as saying that marijuana
... can be a dangerous and damaging drug.... l think we'll find
physical and chemical evidence of that.... If we have a national
commission to undertake a study and report its findings, then
we have... more information and evidence that will be acceptable
to the public. For example,... a commission could make clear
the distinction between addiction and dependency. A kid gets into
steady use of marijuana. After a while he gets less of a charge
from it, and this psychological dependency causes him to move
on to the harder stuff. ... We have got to get proof that
it does create this dependency [my emphasis] (Newsweek,
September 7,1970, p. 22).
The basic metaphor indicated by these two variations of the biochemical
explanation is that of a conveyor belt. Heroin addiction
is seen as a later stage of a process that begins with marijuana
use. If marijuana use is halted, slowed down, or diminished, fewer
drug users will be produced at the other end. John Ingersoll,
the present director of the Bureau of Narcotics and Dangerous
Drugs, has been quoted as saying:
... it is a matter of record that the explosion in marijuana
use has been accompanied by a sharp upturn in heroin use.... We
know that the overwhelming majority of those who use heroin or
LSD... have had prior experience with either marijuana or hashish.
Thus it seems reasonable to assume that if many individuals
did not get involved with marijuana, they would never get around
to using the more potent dangerous drug [my emphasis] (U.S.
News and World Report, May 25, 1 970).
The view that a certain number of marijuana users translates into
a certain proportion of heroin addicts was the guiding principle
underlying the government's "Operation Intercept" and
"Operation Cooperation." It also dominates the theory
that the decriminalization and ready availability of marijuana
would eventually and inevitably produce a greater number of heroin
addictsbecause more people would be smoking marijuana, which
leads to heroin use.
Another theory explaining the statistical relationship between
marijuana use and the use of dangerous drugs, especially heroin,
is the personality abnormality approach. According to this
view, it is only, or mainly, the psychiatrically inadequate who
"escalate" from marijuana to dangerous drugs. Among
those with sound and healthy personalities, marijuana does not
necessarily or typically produce this progression. To put it another
way, the greater the ego inadequacy of the individual, the greater
the likelihood of drug escalation. Experimenting with marijuana
may be the sign of a minor, or even possibly of no, psychiatric
problem. Becoming seriously involved with marijuana may be the
sign of a severe problem. And experimenting with and becoming
involved with the heavy use of drugs such as heroin, methedrine,
and LSD is almost certainly a sign of something seriously wrong
with the user's psyche. Individuals with a severe problem will
not find in marijuana a sufficient "solution" to their
problem. They will therefore "shop around" for some
drug that will answer or mask their psychiatric difficulties.
Thus, this theory holds, both marijuana use (to a minor degree)
and the use of dangerous drugs (to a major degree) are external
manifestations of an internal turmoil. One implication of this
theory is that marijuana use per se has relatively little to do
with the escalation process. In fact, marijuana use is a kind
of "dead end" search: those with serious psychiatric
problems will discard that particular solution and look about
for another one. In this sense, then, rather than "leading
to" the use of more dangerous drugs, marijuana use may actually
retard the process.
The third theory or explanation for the progression from marijuana
to dangerous drugs is the social, or subcultural, model.
This view holds that marijuana use per se has little or nothing
to do with why some users "go on" to the use of more
dangerous drugs. The answer lies almost entirely within the kinds
of social relationships and friendships that users establish in
the course of taking drugs. Drug use inevitably entails making
and sustaining social interactions and human relationships in
conjunction with using drugsit is not simply the ingestion
of a chemical substance. Making drug-related friendships involves
making friendships of a particular type. This will influence one's
further behaviorparticularly in relation to drug use. These
social relationships are variable, not constant. Within certain
social and cultural milieus, the marijuana-to-heroin escalation
may existbecause of the nature of the groups using these drugs
and because of social patterns arbitrarily related to drug use.
But in different groups this escalation may be totally lacking,
because the friendships or relationships established do not have
anything to do with the use of dangerous drugs. It does not depend
on the biochemical effects of the drugs themselves, but on the
characteristics of the groups using drugs, the norms and values
and beliefs held within subcultures. The escalation process (where
and when it does occur) does not come about as a result of anything
intrinsic to the drug itself or to the experience of being high;
rather it relates to the nature of the personal associations one
makes in the course of using drugs.
What evidence do we have to support or refute these various theories
Although personality and psychiatric factors may play a role in
drug escalation, it is clear from available systematic studies
that (I) the biochemical theory of the progression from marijuana
to dangerous drugs is completely false and invalid, and (2) the
sociocultural theory explains more of the variation in drug escalation,
if and when it does occur, than any other model currently proposed.
The disenchantment process proposed by the proponents of the biochemical
theory does not even remotely occur; in fact, precisely the opposite
takes placethe more that an individual smokes marijuana, the
more that he feels and enjoys its effects. Rather than a desensitization,
what actually occurs is an increased sensitization. From
the results of dozens of studies on the subjective effects of
marijuana among experienced, long-term users, a few of which were
described in Chapter 2, it is clear that under the influence of
the drug users experience, and continue to experience, effects
such as euphoria, relaxation, heightened sexual feeling, greater
sensitivity, increased pleasure in listening to music, and so
on. When users are compared with respect to frequency of use,
exactly the opposite occurs as would be predicted by the disillusionment
theory; it is not the occasional user who describes the most frequent
and most pleasurable effects, but rather the heavy, frequent,
chronic user. (Of course, a dialectical relationship is taking
place here; those who enjoy the experience tend to repeat it.)
In one study only 23 percent of the occasional users reported
always feeling an "increased sexual pleasure"
while high; this was true of 39 percent of the chronic users.
Of the occasional users 35 percent reported always feeling an
"intensified sense of taste," but 49 percent of the
chronic users said this (Hochman and Brill 1971).
My own study verified this basic finding. Effects such as paranoia,
depersonalization, and depression become less commonly reported
the more that the individual used marijuana. Pleasurable effects
become increasingly prominent. For instance, when I divided my
sample into three use levelsthree times a week or more, once
or twice a week, and less than once a week52 percent of the
frequent marijuana users said that the drug stimulated their sexual
desires, but less than 30 percent of the infrequent users agreed.
Similarly, of those who had had sex while under the influence
of marijuana, 77 percent of the frequent users claimed that marijuana
increased their sexual enjoyment, and only 49 percent of the infrequent
users agreed (Goode 1970, pp. 162-167).
What the data on the relationship between degree of involvement
with marijuana and subjective effects show is that heavy, chronic,
extended use does not conform to the classic tolerance-disillusionment
model proposed by the adherents of the biochemical hypothesis.
If a drug progression does occur, it is quite decidedly not because
of tolerance, of a reduction in chemical or psychological responsiveness
to the drug. It is specifically the chronic user who tends to
enjoy marijuana's effects the most, and it is also the chronic
user who is most likely to experiment with other drugs. It
is therefore necessary to abandon the tolerance explanation for
drug escalation, if and when it does occur.
The other biochemical modelthe "bigger kicks, greater
thrills" theoryis inadequate simply because it does not
specify any variables. It is known that not all marijuana users
"go on" to the use of heroin or other dangerous drugs;
in fact, only a very small proportion do so. Thus we are still
left with the question of why some (a minority) do and others
(the majority) do not. The view that marijuana itself develops
in the user a "thirst" to try bigger and better things
is invalidated by the fact that all marijuana users go through
the same experiencegetting highyet there is enormous variability
in whether or not they "go on" to the use of more dangerous
drugs. A constant condition cannot explain a variable outcome.
Consequently, again, we have to seek an explanation elsewhere.
In my own study of multiple-drug use (Goode 1969), I found a remarkable
"concatenance of many factors relating to marijuana use."
The study showed that a given user's likelihood of experimenting
with more dangerous drugs was closely related to his involvement
with and in a drug-using subculture. The greater the proportion
of marijuana-using friends an individual had (which was itself
highly correlated with frequency of smoking marijuana)and the
more that an individual was "insulated" from nonusing
friendsthe more likely he was to use other, more dangerous
drugs. About 64 percent of the respondents in my study who had
60 percent or more marijuana-using friends had themselves tried
LSD; but only 26 percent of the respondents who had less than
30 percent marijuana-using friends had tried LSD themselves. Thus
marijuana use can be seen as a kind of index of one's involvement
with a drug-using subculture, with a community of fellow drug
users. The more that one uses marijuana, (1) the more favorable
will be the views of one's friends toward using drugs other than,
and in addition to, marijuana (most users of illegal dangerous
drugs continue to use marijuanaheroin addicts are a
major exception); (2) the greater the likelihood that one will
become involved in selling marijuana; (3) the more one will learn
about the workings of the underground drug market; (4) the further
one will become involved with the drug-using community; and (5)
the greater will be one's opportunities for trying and using other
dangerous drugs. Thus it is not the biochemical properties of
marijuana itself that "lead to" the use of dangerous
drugs; rather it is the social and cultural milieus in which marijuana
useand especially frequent marijuana usetakes place.
This analysis was taken a step further by sociologist Bruce Johnson,
whose work has been cited earlier in this volume. In 1970 Johnson
distributed questionnaires to 3,500 college students in and around
the New York metropolitan area. His findings demonstrated the
crucial role of the marijuana user's place in the drug subculture
as a determinant of dangerous drug use. Johnson divided his sample
into four levels of marijuana useabstainers, experimental users
(less than monthly), moderate users (monthly or more but less
than weekly), and regular users (weekly or more). The outcome,
or dependent variable, that which is to be explained, was the
use of heroin (although fundamentally the same relationship was
found with the use of other dangerous drugs). Johnson found, as
I did, and as do all researchers investigating this phenomenon,
a positive, linear, and significant statistical association between
marijuana use and the use of dangerous drugs. Less than I percent
of the marijuana abstainers in Johnson's survey had tried heroin.
This was true of 1 percent of the experimental users, 5 percent
of the moderate users, and 17 percent of the regular users. The
same relationship held for the use of the other dangerous drugshallucinogens,
amphetamines, barbiturates, methedrine, and cocaine. Johnson concluded
that "the more frequent the use of marijuana, the greater
the likelihood of using any and all other dangerous drugs"
Johnson then set out to determine what variables influenced the
relationship between frequency of marijuana smoking and the use
of dangerous drugs. A crucial explanatory variable in Johnson's
study was differential exposure to dangerous drug-using friends.
By holding marijuana use constant, Johnson was able to test
the relative strength of frequency of marijuana use and having
heroin-using friends as determinants of heroin use. He found that
almost none of the marijuana abstainers in his survey had tried
heroin, whether or not they had heroin-using friends. But among
experimental marijuana smokers, 5 percent of those with at least
one heroin-using friend had tried heroin themselves; this was
true of only.5 percent of those marijuana experimenters with
no heroin-using friends. Among regular marijuana users, only 5
percent of those without heroin-using friends had tried heroin
themselves, but 45 percent of the regular users with at least
one heroin-using friend had tried heroin. As one moved from experimental
to regular use of marijuana, the concomitant use of heroin rose
only slightly, when heroin-using friends is controlled. But as
one moved from not having to having heroin-using friends, the
likelihood of using heroin rose markedly. It is obvious that having
heroin-using friends is far more potent in influencing heroin
experimentation than is the level of marijuana use. Even the
regular use of marijuana does not "lead to" the use
of heroin in the absence of having heroin-using friends.
What variables, then, influence making friends with those who
use heroin? One researcher has suggested that buying and selling
drugs play an independent role in determining involvement in the
drug subculture (Carey 1968). Johnson tested this lead and found
that involvement in the buying and selling of illegal drugsparticularly
drugs other than marijuanawas the single most potent variable
determining whether or not one had heroin-using friends. Among
the respondents in his survey who had only bought (but not sold)
marijuana (and no other drugs), 4 percent of the experimental
marijuana users and 8 percent of the regular users had at least
one heroin-using frienda very small increase. Among those who
had both bought and sold marijuana as well as other drugs, 25
percent of the experimental users of marijuana and 36 percent
of the regular users had at least one intimate heroin-using friend.
This relationship can be looked at in two ways. The jumps from
4 to 8 percent and from 25 to 36 percent (that is, the impact
of degree of marijuana use on having heroin-using friends) are
fairly small. But the jumps from 4 to 25 percent and from 8 to
36 percent are quite significant. Thus it is clear that involvement
in selling drugs is considerably more potent and influential in
having heroin-using friends, and hence using heroin, than is the
factor of frequency of use of marijuana. The original correlation
between frequency of use and the use of dangerous drugs is largely
due to involvement in selling drugs, not to use itself. Buying
but not selling only marijuana did not influence the relationship
very much at all, but selling, and especially selling drugs other
than marijuana, did influence the relationship powerfully. Thus
the causal link between marijuana use and the use of dangerous
drugs does not appear to be the use of marijuana at all. Use of
marijuana is merely an external manifestation of something that
underlies itnamely, involvement with and in a drug-using subculture,
especially in the form of buying and selling illegal drugs, and
having friends who use other dangerous drugs.
These data do not answer the question of whether or not a psychiatric
abnormality plays a role in the escalation process. But if it
alone is determinative, then quite clearly the independent role
of marijuana is inconsequential. If heroin experimentation and
addiction (as well as the use of other dangerous drugs) answer
a deep psychic need for oblivion, degradation, escape, or whatever,
marijuana is merely a fairly easily available drug that presents
itself at a relatively early age and is eventually discarded as
a "solution" to the problems of the user. Thus marijuana
is actually a "digression," to adopt one observer's
phrase (Joyce 1971); it serves to slow down the progress to heroin
rather than to facilitate it (Kaplan 1970). If anything, the personality
abnormality model discounts marijuana's independent role
in the escalation process.
Thus two factors account for the greater use of dangerous drugs
among marijuana users in comparison with nonusers: (1) a process
of selective recruitment, and (2) a process of selective interaction
and socialization. The selective recruitment process largely explains
why young alcohol drinkers and cigarette smokers are more likely
to "go on" to the use of marijuana and dangerous drugs.
This same process determines why marijuana users are more likely
to engage in liberal politics and to be more permissive sexually.
It also underlies the apparently anomalous correlation, noted
earlier, between coffee drinking and illegal drug use. The explanation
is simply that almost no social group or category, almost no set
of participants in any activity, forms a random selection of a
total society. There will be a wide range of differences in participants
versus nonparticipants in any activitynot because the activity
necessarily has anything to do with the differences observed but
because some other factor or variable is meaningfully associated
with the two together, producing an apparent, or spurious, relationship
between them. Knowing this fundamental fact will insulate us from
making absurd causal inferences. This does not mean that no relationships
are meaningfulonly that we have to search carefully to see
whether they are in fact significant, systematic, and causal.
There is a relationship between the use of any and all drugs and
the use of any and all other drugs, whether they are legal
or illegal. This relationship was verified by the Narcotic Addiction
Control Commission's study of 7,500 randomly selected residents
of New York State (Chambers 1971). Regular users of barbiturates
(four-fifths of whom obtained all their supply legally, via prescription)
were ten times as likely to use amphetamines, heroin, methedrine,
and LSD than was the general population. Does using barbiturates
legally "lead to" the use of heroin? Users of diet pills
were five times as likely to be regular users of heroin, and twelve
times as likely to use LSD and methedrine, than was true of the
general population. Does taking diet pills "lead to"
the use of illegal dangerous drugs? The point is that there is
to some extent a drug-taking orientation or "disposition,"
just as there is an abstention orientation. Even before someone
tries a drug, legal or illegal, for the first time, he is already
different from a peer who will never use drugs. Individuals who
use drugs tend to be selectively recruited from segments of the
population that are oriented toward the use of drugs. Illegal
nonmedical drug use is correlated even with frequency of taking
aspirin (Estes and Johnson 1971), not because there is
any direct causal link between the two but because there are some
points of similarity in social characteristics between the kinds
of people who use drugs and the kinds of people who take aspirin.
Marijuana use is not spread evenly across the population. It is
necessary to pay attention to the social characteristics and personality
make-up of marijuana smokers to remind ourselves that many relationships
between marijuana use and almost anything else could be
accounted for solely by the fact that marijuana users are simply
differentwith or without marijuana use. We have to hold these
differences constant before making any inferences as to the "effects"
of the drug, or its "causal" impact.
The second processthat of selective social interaction and
socializationalso operates with marijuana use. Not only are
marijuana users different even before they smoke their first marijuana
cigarette, they also become different by the distinctive
social relations they engage in during the course of their marijuana-related
activities. I have emphasized two such activities and interactions
that have special relevance for the escalation process: buying
and selling drugs, and making friendships with others who use
dangerous drugs. (There are, of course, other indicators of one's
involvement with the drug-using subculture.) The crucial difference
between marijuana "leading to" the use of more dangerous
drugs and legal drugs "leading to" the use of dangerous
drugs is that there is a definite subculture of marijuana usersand
even more so of users of such drugs as methedrine, LSD, and heroin
whereas the term "subculture" cannot be meaningfully
applied to those who use legal drugs such as alcohol and sedatives.
The strong parallels between the use of legal drugs and illegal
drugs can be carried only so far. The subterranean status of marijuana
makes its use more than simply a question of selective recruitment
into a specific activity by a certain "deviant" segment
of society. The very criminal status of marijuana gives its use,
possession, and sale an added socialization and subcultural power
not evidenced by the possession and use of the legal drugs. One
of the reasons that marijuana users are to some degree insulated
from conventional society is the fact that the drug is illegal.
And it is this very insulation that gives the drug-using subculture
its socializing power, and that provides some of the force behind
the progression from marijuana to more dangerous drugs.
The data I have presented point very strongly to the conclusion
that the "conveyor belt" metaphor of drug escalation
is invalid. If marijuana possession were decriminalized and placed
under legal controls similar to those for alcohol, the drug would
eventually lose its subterranean character. This would give the
marijuana subculture less "specialness" and less
socializing power. And the link with the use of other dangerous
drugs would be attenuated. The correlation would not, of
course, disappearit still exists with the legal drugs. But
the marijuana-dangerous drug correlation would become no
different from that between the current legal and illegal drugs.
As John Kaplan has pointed out, it is legitimate to ask "whether
or not the criminalization of marijuana is part of the problem,
rather than the solution" (Kaplan 1970, p. 260). In the case
of the escalation process, this appears to be quite clearly
true. Far from producing more addicts, as the "conveyor belt"
metaphor indicates, it is likely that legalization would produce
fewer, by taking the young marijuana user out of a criminal drug-using
subculture. (This is not to be regarded as a "solution"
to the heroin problem as well; there are other, massive causes
for heroin addiction, and the decriminalization of marijuana would
not make much of a difference.) The specific patterns that the
escalation process follows show the invalidity of retaining marijuana's
criminal status. Our legal policies in regard to marijuana have
been a gigantic mistake. It is time to correct that error.
1. I do not discuss the monograph on marijuana
use and criminal behavior because the issue is less likely to
be taken seriously today than is the question of whether marijuana
leads to dangerous drugs.(back)
Blum, Richard H., et al. 1969. Students and Drugs. San
Carey, James T. 1968. The College Drug Scene. Englewood
Cliffs, N.J.: Prentice-Hall.
Chambers, Carl D. 1971. An Assessment of Drug Use in the General
Population. New York: Narcotic Addiction Control Commission.
Estes, J. W., and Johnson, Malcolm. 1971. "Relationships
Among Medical and Nonmedical Uses of Pharmacologically Active
Agents." Clinical Pharmacology and Therapeutics 12:
Giordano, Henry L. 1968. "MarihuanaA Calling Card to Narcotic
Addiction." FBI Law Enforcement Bulletin 37 (November
1968): 2-5, 16.
Goode, Erich, 1969. "Multiple Drug Use Among Marijuana Smokers."
Social Problems 17: 48-64.
Goode, Erich, 1970. The Marijuana Smokers. New York: Basic
Hochman, Joel S., and Brill, Norman Q. 1971. "Marijuana Use
and Psychosocial Adaptation." Unpublished manuscript.
Huff, Darrell. 1954. How to Lie with Statistics. New York:
Johnson, Bruce D. 1972. Social Determinants of the Use of Dangerous
Drugs by College Students. New York: Wiley.
Joyce, C. R. B. 1971. "From Hard to Soft DrugsProgression,
Regression, or Digression?" Unpublished manuscript.
Kaplan, John. 1970. MarijuanaThe New Prohibition. New
York: World Publishing.
Kupperstein, Lenore. 1971. "The Role of Pornography in the
Etiology of Juvenile Delinquency." In Technical Report
of the Commission on Obscenity and Pornography, Vol. I, pp.
103-111. Washington, D.C.: U.S. Government Printing Office.