Schaffer Online Library of Drug Policy Sign the Resolution for a Federal Commission on Drug Policy


Contents | Feedback | Search | DRCNet Home Page | Join DRCNet

DRCNet Library | Schaffer Library | Major Studies

Marihuana, A Signal of Misunderstanding


The Report of the National Commission on Marihuana and Drug Abuse

Marihuana and (Non-Violent) Crime

Much that has been said with respect to aggressive behavior and violent crime also applies to the more general proposition that marihuana causes or leads to (non-violent) crime and delinquency. The popular and professional literature abounds with claims and counter-claims. Public and professional opinion surveys demonstrate a widespread belief in the existence of a marihuana-crime relationship but also reveal considerable uncertainty about the existence of such a relationship.

The empirical evidence is somewhat more consistent. Laboratory studies provide no evidence that marihuana produces effects which can be interpreted as criminogenic. Although some studies of offender populations purport to demonstrate a causal relationship between marihuana and crime, they reveal, at the most, a significant statistical association. Closer examination of these data or more sophisticated analysis, however, generally shows the purported relationship to be spurious. The original relationship is usually found to derive not from the chemical effects of the drug but from the operation of social and cultural variables unrelated to either the drug or its use.

In the following pages the available evidence bearing on the relationship between marihuana. and crime will be reviewed in an effort to determine whether marihuana itself or the use of the drug plays a significant precipitating or contributory role in the commission of criminal or delinquent acts.


The formulation and expression of strong opinions about the relationship of marihuana to criminal and delinquent behavior have persisted despite the inherent complexities of the issue, the relative absence of conclusive empirical evidence and the general lack of knowledge and understanding about the effects of the drug. For the most part, however, neither public nor professional opinion about marihuana and its impact on public safety has been explored in any systematic fashion. As a consequence, public policy with respect to the drug has been shaped by the most vociferous advocates of one position or another at any given time.

Probably the first official body to explore prevailing opinion about the relationship between marihuana use and crime was the Indian Hemp Drugs Commission of 1893-1894. The Commission spent over a year in making field trips to 30 cities, in receiving evidence from almost 1,200 expert witnesses (335 of whom were medical practitioners) and in reviewing judicial proceedings and the case files from India's mental hospitals. The Commission's Report, reprinted in 1969, contains the following information gathered with respect to the relationship of marihuana use to crime.

First, the Commission found that the majority of hemp drug consumers were moderate rather than excessive users, and that the drug users were rarely regarded as offensive or potentially dangerous by their neighbors (a few objected to the smell of the smoke or the example set by the users for the neighborhood children).

With respect to the drug's possible long-term or chronic criminogenic effects (producing "bad characters"), the Commission reported that two thirds of the witnesses did not believe that marihuana would produce, over time, a large proportion of "bad characters" among the moderate users. A majority felt that even excessive use was unrelated to the production of "bad characters." When the possible relationship was framed more precisely in terms of cause-effect rather than statistical association, a ratio of 8 to 1 of the witnesses held that moderate consumption of these drugs had no connection with crime and a ratio of 4 to 1 denied a causal connection between excessive consumption and being a "bad character."

With respect to the more acute effects of hemp drugs, the Commission set out to determine whether criminals use the drug to fortify their courage prior to the commission of their crimes, whether the drugs were used by criminals "to stupefy their victims," and whether the drugs incited the user to commit unpremeditated crimes. The Commission's conclusion regarding the first question was that "criminals like any other consumers of these drugs go to them for that assistance when they feel they require it" (p. 256).

To the second question, the Commission responded that although some persons had alleged the commission of "thefts of ornaments from children stupefied by sweet meats" containing marihuana, the fact of other readily available, more effective and more disabling drugs, considerably more conducive to surreptitious administration, cast doubt on the use of hemp drugs by criminals for this purpose.

To the third question, the Commission said that the majority of witnesses saw no connection between either the moderate or excessive use of hemp drugs and the commission of unpremeditated crimes, including crimes of violence. The Commission therefore concluded that "for all practical purposes it may be laid down that there is little or no connection between the use of hemp drugs and crime," (p. 264).

The Mayor's Committee on Marihuana (1944) also interviewed law enforcement officers (federal, state and local police) about the purported link between marihuana and crime. The Committee reported that:

In most instances [the police officers] unhesitatingly stated that there is no proof that major crimes are associated with the practice of smoking marihuana. They did state that many marihuana smokers are guilty of petty crimes, but that the criminal career usually existed prior to the time the individual smoked his first marihuana cigarette (Schoenfeld, 1944: 14-15).

Reference has already been made to the results of a small, unpublished survey of the opinions of police officers about the relationship of marihuana to aggression (Schofield, 1968), All of the respondents reported observing a variety of conduct exhibited by users under the influence of marihuana and emphasized that an individual's reaction to the drug depends on his particular personality and, in some instances, the strength of the dose. Some, respondents observed a recent shift in the type of individual or personality prone to using marihuana.

In response to a question regarding the use of marihuana prior to engaging in property crimes, 10 out of the 16 officers interviewed believed that this often occurred. Seven cited specific examples of offenders who reported using marihuana before committing their offenses to bolster their courage and sharpen their senses; the views of three rested on hearsay evidence; and the remaining six officers responded that they did not believe this was a common practice and that they had never personally observed such examples.

In the Commission-sponsored National Survey of a representative sample of more than 3,000 American youth and adults (Abelson, et al., 1972), respondents were asked to indicate their agreement or disagreement with the statement that "many crimes are committed by persons who show that 56% of the adults and 41% of the youth agree with the statement. The extent of agreement, however, varies significantly according to age, education, geographic location and the marihuana-using experience of the respondent. Those who are early adolescents (12-13 years), over 25 years of age, have not completed high school, live in the South or North Central regions of the country and have no experience with marihuana are, significantly more likely to agree with the statement than are those who are between 14 and 25 years of age, are at least high school graduates, live in the Northeast or West and have had experience with marihuana. Table 5 below, shows the percentage of respondents agreeing to the statement according to age, education, geographic location and experience with marihuana.

The survey also showed the existence of considerable uncertainty with respect to the proposition that many crimes are committed by persons under the influence of marihuana; 25%, of the youth and 17% of the adults were either unsure of its relative truth or failed to respond to the question (p. 69). Adults were considerably more certain, however, about the relationship between alcohol and crime (youth were not asked the question) ; 7% were either unsure or did not respond and 69% agreed with the statement that many crimes are committed by persons who were under the influence of liquor (pp. 28, 31).


Percent agreeing

Youth Adults

Total respondents agreeing 40.7 56.0

Age (years):

12-13 54.5

14-15 35.2

16-17 31.8

18-25 35.1

26-34 48.5

35-49 59.0

50 and over 68.8


Less than high school graduate 70.9

Eighth grade or less 49.5

Ninth-twelfth grades 31.8

High school graduate 56.8

College or more 39.2

Geographic location:

Northeast 34.1 46.9

North-Central 42.1 54.9

South 48.5 66.8

West 34.0 52.3

Marihuana experience (ever used):

Yes 8.9 24.1

No 46.1 62.0

The surveys of prosecuting attorneys, judges, probation officers and court clinicians also revealed considerable doubt about the existence of a causal connection between marihuana and crime. Although the survey of prosecuting attorneys conducted by the Commission staff asked only about aggressive behavior and violent crime, the data showed that 52% of the respondents either denied or were uncertain that marihuana causes such behavior.

The Commission-sponsored survey of nationally representative samples of judges, probation officers and court clinicians, however, posed several questions bearing on the more general proposition that marihuana is related to crime and delinquency (InTech, 1971). Respondents were first asked whether or not they had witnessed an increase in the incidence of drug or drug-related offenses in their caseloads during the past five years. The large majority of all three groups (86%) responded affirmatively and reported that marihuana was the type of drug most often seen in this connection (83%). However, about two thirds of the respondents reported that marihuana offenders (those arrested for using or selling marihuana) constituted less than 20% of their caseloads; 56% of the respondents reported that those arrested for non-drug offenses but incidentally found to possess marihuana comprised less than 10% of their caseloads; and 73% of the respondents reported that less than 10% of their caseloads attributed their offenses to marihuana (pp. 31, 34).

To determine better the nature and direction of a possible relationship between marihuana use and crime, respondents were asked to indicate which -of four statements most reflected their own professional experience. On the whole, the respondents were more likely to postulate a statistical association (36.6%) rather than a direct cause-effect relationship (26.9%) ; 18% of the respondents thought that involvement in a criminal or delinquent subculture caused or led to the use of marihuana (in contrast to the. more prevalent belief that marihuana use, leads to crime) and 9.1% believed there was absolutely no relationship between marihuana use and other criminal or delinquent behavior. The extent to which uncertainty prevails even among these practicing professionals is reflected in the 9.3% no response category (p. 39).

Table 6 shows the responses of the judges, probation officers and clinicians to each of the four propositions (InTech, 1971: Appendix 1, Section II, Question 3). The data show that judges are significantly more, likely to believe that marihuana causes or leads to criminal and delinquent behavior and significantly less likely to believe in either a statistical relationship or in no relationship whatsoever than either the probation officers or the clinicians.

The researchers also attempted to determine these, professionals' opinions about crimes actually committed by marihuana users. More specifically, did the respondents think that most non-drug crimes committed by persons who were known users of marihuana occurred when the individuals were, actually under the influence of the drug or when they attempted to obtain it? Again, the responses indicated considerable uncertainty in all three, groups. One-third of the, respondents were either unsure or did not respond to either proposition. About twice the number of respondents thought that the crimes occurred while in the attempt to obtain marihuana (31.0%) rather than while tinder the influence of marihuana (16.8%).

Table 7 shows the percentage of respondents indicating the probable truth and probable nontruth of the two propositions (Appendix 1, Section 11, Questions 2a, 2b).

The data show that half of the respondents denied the commission of non-drug crimes while under the influence of the drug, suggesting that marihuana itself does not have the capacity to produce criminogenic effects. About one-third of the respondents, however, believed that these crimes occur when the user is attempting to obtain the drug, suggesting an addiction model; that is, that users commit crimes to support a "habit."`

This interpretation is corroborated by the significant findings that 65% of the respondents either did not know (23%) or thought that a few (21%), some (17%) or most (4.4%) of the regular users were physically addicted to marihuana


(Figures in Percentages)

Most non-drug crimes committed by persons who use marihuana occur when the offender is-

Under the in- Attempting to

fluence of mari- obtain marihuana


Prob- Prob- Prob- Prob-

ably ably ably ably

true not true true not true

Judges 17.1 % 46.8 % 34.2 %, 30.4 %,

Probation officers 21.8 47.3 30.0 37.3

Court clinicians.. .. . 10.4 65.2 16.5 60.0

Total 16.8 49.6 31.0 35.7

The National Survey of the general public (Abelson, et al., 1972) likewise showed that large segments of the population (65% of all adults and 48% of all youth) regard marihuana as addictive. Even among the users, 40% of adults and 21% of the youth believed marihuana to be addicting (p. 22)

These findings, taken together, suggest that much of the confusion regarding the relationship between marihuana and crime may be predicated on public and professional misconceptions about the drug's addiction potential. The Commission's National Survey showed that, among those persons who believed marihuana to be addicting, 67.5% of the adults and 52.3% of the youth believed that many crimes are committed by persons under the influence of marihuana. Among those who said that marihuana was not addicting, 34.6% of the adults and 30.2% of the youth agreed that many crimes are committed by marihuana users. Table 8 below shows this relationship (unpublished data, Abelson, et al., 1972).


(Figures in Percentages)

Persons who think marihuana is

Addictive Not addictive

Youth Adults Youth Adults

Many crimes are committed under the influence of


Mostly agree.. . . 52.3 67.5 30.2 34.6

Mostly disagree 20.8 14.5 39.9 41.1

Other 2.3 3.1 4.4 2.0

Not sure/no answer 24.6 14.9 25.6 27.3


(Figures in Percentages)

Judges Probation Court Total

(N=556) officers clinicians (N=781)

(N=110) (N 1.15)

Use of marihuana causes or leads to antisocial behavior in the sense

that it leads one to commit other criminal or delinquent acts 33.8 18.2 1.7 26.9

Involvement in a criminal or delinquent subculture causes or leads

to the use of marihuana 18.3 20.0 14.8 18.1

There is a statistical relationship or association between marihuana

use and other criminal or delinquent behavior, but it is not a

cause-effect relationship 31.1 40.0 60.0 36.6

There is absolutely no relationship between marihuana use and

other criminal or delinquent behavior 6.7 14.5 15.7 9.1

(X2=76.895, p<.001).

To summarize, these opinion surveys demonstrate that there is considerable uncertainty about the existence and the nature of a relationship between marihuana use and crime and that this uncertainty exists among youth and adults, practicing professionals in the criminal justice community and the lay public, marihuana users and non-users alike. The data suggest, however, that the persistent and fairly widespread belief in a cause-effect relationship between marihuana and crime may, at least in part, be predicated on the erroneous but equally prevalent belief that marihuana is physically addicting. To the extent that the general public operates under this misconception, they may be more likely to believe that, like the heroin user, the marihuana "addict" commits crime in order to support his "habit" and obtain the drug.

The logical extension of this belief is that marihuana users, like heroin user-,, will be considered more likely to commit their crimes in the attempt to obtain the drug than while under the influence of the drug. Stated another way, persons who believe that marihuana is physically addicting would be more likely to base the purported causal relationship between marihuana and crime not on the chemical effects of the drug per se but on the user's physical need for the drug. As such, they would be more likely to postulate that most crimes committed by users occur not when the offender is under the influence of the drug (which, like heroin, makes one passive, lethargic and stuporous) but when the, "addict" needs a "fix" and is desperately trying to obtain the drug or the money to buy it. This interpretation is, indeed, supported by the findings of the Commission-sponsored survey of judges, probation officers and court clinicians and is suggested from the findings of the National Survey.


Over the years, numerous studies have relied on offender populations and zero-order statistical correlations to "demonstrate" a cause-effect relationship between marihuana and crime. This method of " proof by enumeration" has become probably the most common approach to "demonstrating" the purported causal relationship.

As Goode (1970) has noted, however:

Not even marihuana's staunchest supporter would argue that a crime has never been committed by a user while high. Yet, incredible as it seems, the burden of many proofs of marihuana's criminal effects has been precisely the simple fact that it is possible to locate crimes committed in conjunction with smoking marihuana. "Proof" by enumeration is no proof at all. By examining an enumeration of crimes which were committed under the influence of marihuana (even were this definitely known), it is impossible to determine the " cause" of the event taking place, in this ease the crime-or, indeed, that marihuana has anything whatsoever to do with its commission (pp. 215-216).

Even if these studies are taken at face value, their findings do not generally support the thesis of a positive and significant statistical association. In most instances, marihuana has not been found to predispose one to commit crime or to serve as an initiator of criminal careers; nor does the use of the drug appear to alter the progression to other non-drug offenses among those without prior criminal records or histories of psychological maladjustment. Predictably, however

the data do indicate a greater rate, of progression among marihuana-using offenders (both those arrested specifically for their marihuana offenses and those charged with non-drug crimes) than that which might be expected from similar, non-using offenders or from the general population.

Because these studies do provide at least interesting insights into the possible nature and direction of a relationship between marihuana and crime or, conversely, a non-existent relationship, several of these studies are summarized below.

Bromberg, 1939. Between 1932 and 1937, Bromberg and his team of researchers reviewed all cases eventuating in conviction by New York's Courts of General and Special Sessions; conducted interviews with about 17,000 drug and non-drug-using offenders; and analyzed the statistics from both courts. The report of this work, published in 1939, yielded the following information.

Among the 16,854 offenders convicted of felonies in the court of General Sessions, 67 (.005%) were identified as marihuana users. Of these 67, 46 (697c) had been convicted on charges of possession with intent to sell; 16 (24%) were charged with burglary, robbery and grand larceny; two persons were charged with assault and one each was charged with petty larceny, forgery and murder.

The researcher points out that in only nine cases did the, offenders' criminal records commence with a drug charge, "indicating that there was not in those cases a close relationship between drugs and the beginning of a career in crime" (p. 10).

In a 25% random sample of offenders convicted in the Court of Special Sessions (misdemeanors), 135 or 9% were charged with possession of marihuana. Of these 135, 93 or 69% had no previous record, 8 had been charged previously with drug violations only, 5 had mixed drug and non-drug charges and 29 (21%) had only non-drug arrests. In summary, the researcher stated that:

As measured by the succession of arrests and convictions in the Court of General Sessions (felonies) . . . ,

it can be said that drugs generally do not initiate criminal careers. Similarly, in the Court of Special Sessions (misdemeanors), only 8% of the offenders had previous charges of using drugs and 3.7% had previous charges of drugs and other petty crimes. In the vast majority of cases . . . then, the earlier use of marihuana apparently did not predispose to crime, even that of using other drugs. . . . The expectancy of major crimes following the use of cannabis in New York County is small, according to these experiences (p. 10).

Bromberg and Rodgers, 1946. A study of 8,280 naval and marine prisoners at the U.S. Naval Prison in Portsmouth, New Hampshire (1946) revealed that 40 or .0048% of the offenders were marihuana users. The offenses they committed while in the service were: AWOL or AOL (32, or 80%), assault or striking an officer (3), theft (3), and narcotics violations (2). Their previous civilian offenses included three each of assault, theft and traffic violations and one each of gambling and narcotics violations, drunkenness, draft dodging, and violation of the Mann Act. Twenty-eight or 70% of the marihuana users had no previous civilian criminal records (p. 825).

The preponderance of psychiatric disorders in the user group over the non-users (40 randomly selected non-using naval prisoners) led the researchers to conclude that "marihuana usage is but an aspect of some type of mental disorder or personality abnormality" (p. 826), a conclusion also reached by Charen. and Perchan 1946), Ausubel (1958), Andrade (1964), Lambo (1965), Bloomquist (1968), Simmons (1969) and Grinspoon (1971).

Gardikas, 1950. The researcher reviewed the criminal records of "379 individuals either sentenced or arrested flagrante delicto for using hashish publically" between 1919 and 1950. More than half (55%) of these individuals were already known to law enforcement authorities prior to their use of hashish and an additional 14% had no subsequent difficulties with the law except for hashish and vagrancy offenses. The researcher notes, however, that the remaining 117 offenders (31% of the original sample) went on to become " confirmed criminals after their first hashish arrest."

Dividing this last subgroup of 117 into three approximately equal parts, Gardikas. then described the subsequent criminality of each group. The first group was given 332 more sentences following their first sentence for using hashish; 142 (43%) were additional hashish offenses, a similar number were for violent crimes or crimes involving weapons, 18 were for "insults" and two were for "high treason."

In a critique of the Gardikas study, Kaplan were accumulated; mostly for hashish offenses (42%) and thefts (20%). Although there were a small number of violent crimes subsequently committed by this group of offenders, most sentences subsequent to their initial hashish sentence were for such relatively minor offenses as illegal gambling, living on immoral earnings, "false statements of identity," and "fishing with dynamite,."

The, third group discussed by Gardikas, comprised of those "who after having, made use of hashish became criminals," accumulated 332 additional sentences. Hashish offenses were again responsible for 30% of these offenses and most of the rest were for theft. Gardikas notes with respect to this group that although not all of these individuals were criminal before they used hashish, their use of hashish "turned [them] into habitual hashish smokers and habitual criminals with a, strong propensity leading toward crime of dishonesty and particularly theft and f rand. At least one-half of them surely and undoubtedly are even to be characterized as dangerous idle, vagrants" (p. 203).

In a critique of the Gardikas study, Kaplan (1971) has stated that:

The basic problem underlying Gardikas' paper is his conclusion that hashish use caused his subsample of 117 arrests to become criminals. This assumes, first. that they had not engaged in crime before beginning to use hashish and, second, that it was the hashish use and not something else that turned -them toward crime. Neither of these assumptions is justified by the data. . . . As for the first assumption . . . all we know is that they had not been arrested or convicted for other crimes first. . . . The second assumption . . . is also impossible to justify. Even if the members of the subsample had not previously been criminals, it may very well be difficult to distinguish the criminogenic effects of hashish from those of conviction and sentence to Greek jails.

Even aside from any criminogenic effect of the Greek penal system, the figures may indicate only that once somebody has been arrested for hashish-and probably served a term in jail as well-lie is more likely to be picked up by the police when further crimes are committed. Finally, the years from 1919 through 1950 . . . were a period of enormous social dislocation [in Greece] . . . . It is obviously an almost impossible task to sort out how much of the subsequent criminality in the subsample was due to hashish use and how much to the social chaos that prevailed in Greece (pp. 103-104).

Andrade, 1964. The researcher retrospectively reviewed the examinations of 120 patients sentenced between 1951 and 1960 by a Brazilian Court to Heitor Carrilho insane asylum. Sentencing in all cases was based on the allegation that cannabis was directly responsible for their crimes. The researcher concluded that:

In the 120 patients examined, we did; not find any criminogenic action that could be attributed to cannabis; ... In the majority of cases examined (83), we found patients with psychotic disturbances. . . . In the cases in which we did not find mental disturbances, a total of 37, the crime attributed to them was that of carrying or selling marihuana.... In the study of the dynamics of the crimes of all the others, we saw that there was a link between the crime and the illness, independent of the use of cannabis (Tinklenberg, 1971: 12).

Lambo, 1965. The researcher compiled lists of crimes occurring over a, two year period in three West African nations and recorded the number of offenders perpetrating these crimes. He found that approximately one-fourth of the offenders had previous convictions. Forty-seven percent of the 863 offenders convicted for false pretenses, 61% of the 2,880 convicted burglars and 54% of those convicted on charges of "culpable driving" were said to have histories of cannabis use. Many of these users, however, were also shown to have long histories of psychological maladjustment (p. 10). The researcher suggests that "the use of cannabis enhances suggestibility in certain individuals, and this may be a factor in the commission of crime, by these chronic abusers" (p. 11).



As indicated at the outset, enumerative studies of the crimes committed by apprehended marihuana, law violators and non-drug offenders identified as marihuana users cannot, by their very nature, either prove or disprove a causal connection between the use of cannabis and the commission of crimes; nor can their rates of crime be projected onto the larger universe of marihuana users.

As several researchers have pointed out, those individuals apprehended for marihuana offenses represent only a. tiny fraction of all marihuana users (Kaplan, 1971; Weitzner, et al., 1971) ; likewise, those in other offender samples who are identified as marihuana users represent only a small proportion of the offender populations of which they are a part (Bromberg, 1939; Bromberg and Rodgers, 1946; Kaplan, 1971).

Although the data do indicate that some individuals identified as marihuana users are subsequently involved in crime, both drug-specific and non-drug offenses, they do not provide support for the thesis that cannabis was the determining factor in their criminal careers. Likewise, they fail to indicate that the rate of progression to other crimes manifested by these offenders was significantly greater than that which might be expected in either a non-using criminal subgroup or the general population.

As Tinklenberg (1971) has stated:

The important issue is that unless one has systematic data on the proclivities toward crime of the various subgroups of marihuana users, one cannot answer the crucial question of whether the use of marihuana alters the actual rate of crime among these various subgroups over the criminality which would be expected. At this time, it is unknown whether individuals with these characteristics seek the use of marihuana or whether the use of marihuana in any way contributes to the development of these traits (P. 24).

By reading between the lines of these enumerative studies, however, one would be tempted to postulate that individuals with certain psychological, social and cultural characteristics are more likely to seek the use of marihuana, than is marihuana likely to contribute to the development of delinquent or criminal behavior patterns.

The number of researchers pointing to individuals with long histories of psychological maladjustment or disturbances has already been noted. Others have alluded to persons involved, prior to cannabis use, in criminal or delinquent subcultures (Blumer, 1967; Robins, et al., 1970; Kaplan, 1971; Weitzner, et al., 1971). Still others suggest that marihuana use is more likely to develop among persons living in underprivileged communities or within a social structure that limits achievement and advancement (Asuni, 1964; Blumer, 1967; Goode, 1972).

In recent years, considerable evidence has been gathered to suggest that the use of other drugs and association with drug-using friends are also likely to enhance the probability of marihuana use (Hochman and Brill, 1971; Goode, 1972; Abelson, et al., 1972).

Because many of these characteristics are likely' to be associated with both criminal or delinquent behavior and marihuana, use, some individuals have mistakenly concluded a. cause-effect relationship rather than a statistical correlation.

The studies reviewed in the following pages axe probably the most methodologically adequate assessments of the. purported relationship between marihuana and crime undertaken to date. The data show that the seemingly significant statistical correlation often observed between marihuana, use and crime is spurious; it is dependent not on the chemical effects of the drug but upon the operation of several extra-pharmacological variables which have little or nothing to do with the use of marihuana. per se.

The first of these studies (Goode, 1970) is based on a sample of marihuana users. Because it, like the study by Robins and his colleagues (1967, 1970) referred to earlier, involves a selected population, the extent to which its findings can be applied to the general population remains speculative.

The second study presented, a Commission sponsored survey of young male urban dwellers, is probably the most adequate assessment to date of the relationship between marihuana use and crime. The results are based on a representative sample of the general population of males, 15 to 34 years of age, the self reports of criminal behavior have been cross-checked with Philadelphia Police Department files; and numerous statistical controls have been applied to the data. The study therefore provides at least tentative answers to the following questions:

-Do young, male marihuana users and nonusers in the general population differ significantly in the nature and extent of their criminal activities?

-Is marihuana use, in and of itself, the principal determinant of any observed differences? -If not, what are the variables which explain a statistical correlation between marihuana use and crime?

Goode, 1970. Between February and September of 1967, face-to-face interviews were held with 204 marihuana users residing in the New York City Metropolitan Area. The sample, admittedly nonrepresentative, was drawn front membership lists of drug-related organizations and also included friends and acquaintances of the researcher.

The sample, suspected of differing to some unknown extent from a random sample of marihuana users, can be characterized -as primarily male (53%), young (median age 22 years), white (89.5%), single (78%), middle class urban dwellers, including students, dropouts, business people, housewives and the unemployed (p. 316).

All of the respondents, by virtue of their marihuana use, possession or sale, had engaged in law violative behavior, but, only nine marihuana arrests were reported. This finding indicates that persons who are arrested for marihuana related activities constitute only a small fraction of total marihuana users.

Because of their marihuana-related criminal activity, one would expect that persons in this sample would have a higher probability of being arrested than would the "average" person from the general population. To determine if, in fact, this might be the case, the researcher computed the rate of arrest for his sample and compared it with the national rate of arrest provided by the Federal Bureau of Investigation for the year 1965. Despite the fact that the rates are not strictly comparable (the sample rate is based on arrests ever experienced; the national rate is based on the number of arrests recorded for one year), the similarity is striking. The national rate was given as 3.7 per 100; the sample rate was computed to be 3.9 per 100 per year, 4.5 if one adds the marihuana arrests (p. 236).

Although the arrest, rates of the users and the general population were found to be roughly similar, the types of crimes committed by the users differed significantly from those crimes recorded for the population as a whole. In contrast to the majority of drunkenness and disorderly conduct charges noted for the general population, the offenses of the users most often resulting in arrest involved participation in political demonstrations (19 out of the 55 arrests). No other single offense attracted more, than a few arrests (p. 2335).

In addition to examining arrest rates and the types of offenses committed, the researcher also investigated the relationship between the amount of marihuana smoked and the criminal behavior of the user. If it were true that marihuana did produce a dangerous and criminogenic state in the user, the greater would be his likelihood of committing crimes and of being arrested.

The data, however, show no relationship; excluding the "political crimes" (which were most common among the least frequent smokers and least common among the most frequent smokers), the "serious" crimes committed by 15 respondents (non-marihuana narcotics possession, disorderly conduct, drunkenness, burglary, theft, assault, auto theft, serving liquor to a minor and larceny) resulted in a total of 21 arrests. The heavy smokers did not commit these crimes significantly more frequently than did the light smokers.

Similarly, there was no relationship found between the frequency of use and the likelihood of arrest; three of the daily smokers, three of those who smoked three to six times per week, three who smoked one or two times weekly, one who smoked four times monthly and three who were less than monthly smokers were arrested for committing these "serious" crimes (pp. 237-38).

The researcher concludes that:

Although these numbers are extremely small, the fact of their perfect dispersal is perhaps indicative of the lack of a crime-inducing effect of the drug. It is, at any rate, a proposition which ought to be tested more systematically in the future with more complete data. For the moment, there are indications that point to the fact that the marijuana smoker is no more criminal than the rest of the population (p. 238).

Goode, 1972. The Commission-sponsored Philadelphia survey set out to resolve these issues. Goode (1972) analyzed the data collected during August of 1971 from an interview survey of 559 15-to-34-year-old residents of West Philadelphia.

That part of the report devoted specifically to the analysis of the data is presented in its entirety below. The only deletions made were the author' references to previous studies bearing on the issues addressed, the findings of which have been presented earlier in this chapter.

Excerpts From



*Prepared by Professor Erich Goode for the National Commission on Marihuana and Drug Abuse, January 15, 1972. The data on which this report is based were obtained from a survey conducted by the Institute of Survey Research of Temple University. All tables referred to in Professor Goode's text are found at the end of these excerpts. 'The interview schedule was developed and executed and the data processed for the Commission by Temple University's Institute for Survey Research, in Philadelphia.

The data analyzed in this report were gathered in August 1971 by the Institute for Survey Research of Temple University In Philadelphia; the author of this report did not have a hand in any phase of the data gathering, including constructing the interview schedule, although he is solely responsible for the data analysis. The sample consists in toto of males age 15 through 34 living in households located in a section of West Philadelphia. Each respondent was interviewed; a total sample size of 559 was collected. (The refusal rate was 10%, which is typical for house-to-house surveys of this type.) The area in which the sample was drawn is composed mainly of two heterogeneous populations: lower-middle and working-class, high school educated blacks, and college-associated whites-students and professionals associated with [Drexel University and] the University of Pennsylvania. About four respondents in ten in this survey are black, and just over half are white. (The number of Orientals and Puerto Ricans or Mexicans is too small to permit a meaningful statistical analysis, so that when race is discussed in this report, only the black-white comparison will be made.) The racial composition is roughly equivalent to the census figures for the sample area although no attempt was made, through a weighting of cases, to "correct" any discrepancies which might have prevailed in population composition. More detailed information is available from the Institute for Survey Research.

A quarter of the sample (23%) had never smoked marijuana; in 17 cases (3%), no information on marijuana use was given by the respondent. Of the marijuana users the question asked about frequency of use during the respondent's most recent year of use-about 1 in 10 smoked marijuana daily or more, about 1 in 5 smoked approximately once or twice a week, 1 in 4 smoked once or twice per month, on the average, and about 4 in 10 smoked marijuana several times a year. (The study also asked how often the respondent smoked marijuana in his first year of use a variable which I have not made use of in this report.) Questions were also asked concerning the respondent's use of other drugs-"stimulants or uppers, such as bennies, speed, or dex," "sedatives or downers, such as sleeping pills, amytal or blues, or nembutal, or yellow jackets," "hallucinatory drugs, such as LSD, mescaline, or STP," and "hard drugs, such as heroin, morphine, demerol, cocaine, or codeine." In addition, questions on one's friends' drug use-both marijuana and these other drugs--were asked.

The commission of various offenses formed another segnient of the questionnaire. There were 16 of them, and ranged from those which would not be considered crimes in most instances-such as "disturbing people"-to those which would be considered felonies, drawing long prison sentences such :as rape, armed robbery, and assault. As would be expected], the serious offenses were committed (or admitted to) by very, very few respondents-and thus, any comparison between users and non-users has to be made mainly on the basis of committing trivial offenses....


At this point, I will introduce the data from the Philadelphia marijuana use and crime study, specifically mandated by the National Commission on Marihuana and Drug Abuse. This survey's data will form the bulk of the factual evidence on the marijuana-crime connection in the report which follows.

The question of marijuana's relationship to crime and aggressive behavior is obviously closely related to the drug's effects. One of the most fundamental generalizations in psychopharmacology to be obtained from thousands of recent research efforts is that the same drug does not have standard and invariant "effects," but that effects are sensitive to a number of extra-pharmacological variables, including the personality of the user, the social setting in which drug, use takes place, the user's past experience with the drug, his expectations, and so on. Thus, it is elliptical and somewhat artificial, and incomplete, to speak of any drug's "effects," as if they occurred under any and all conditions, in all users. We should bear this qualification in mind when looking at the relationship between the ingestion of a drug and any subsequent behavior-with the latter supposedly ",Caused" by the effects of the drug. Drug effects vary, and, in addition, even standard effects do not automatically translate into specific forms of human behavior. Even opposite forms of behavior could follow the same effects, given different individuals taking the drug, different settings in which the drug is used, different definitions of the drug and its use and so on....

The data collected by the Philadelphia survey on the subjective effects of marijuana dovetail precisely with those of earlier studies. The effects agreed to by the users in the sample are clearly inconsistent with aggressive behavior. This does not mean that marijuana cannot be related to the commission of criminal or aggressive acts-but it does suggest that the effects of marijuana, per se, may have nothing to do with the commission of crimes, especially violent crimes. Table 1 summarizes the results from this survey's question on marijuana's effects, which was: "I am going to read to you some of the ways using marijuana may affect people. For each could you tell me whether marijuana had this effect on you almost every time, more than half the time, less than half the time, or never or almost never, when using marijuana." The effects asked about were related to feelings generally thought of as criminogenic either in a positive or a negative direction. Those that are generally considered to be related to crimes deal with anger and frustration, or with derangement, or with suggestibility. Those inversely related to the commission of crimes and aggressive acts dealt with relaxation and sleepiness.

As can be readily seen from Table 1, users overwhelmingly deny that marijuana has effects on them which could be interpreted as criminogenic or violent in nature. Nearly all respondents specifically denied that aggressive feelings came over them during the marijuana intoxication. When asked whether, under the influence of marijuana, they had a feeling of wanting to hurt someone, 96% said that this occurred never or almost never. When asked whether they had feelings of wanting to do something violent, 95% said never or almost never. Almost nine in 10 (88%) said that they never or almost never felt more angry when high. And almost eight in 10 (78%) said that they never or almost never felt frustrated when tinder the influence. Thus, the frustration-anger-aggressive impulse syndrome seems to be an extremely rare phenomenon associated with the marijuana intoxication. Likewise, the suggestibility syndrome does not appear to be characteristic, although it seems to be somewhat more common than aggressive feelings. About six in 10 (or 59%) of the respondents said that they had a feeling of "being more willing to follow other people's suggestions," never or almost never. And about three-quarters (77%) said that they had a "feeling of being able to do anything," never or almost never. Those sensations, however, which would be seen as inhibiting criminal and aggressive im pulses and acts were decidedly more common, Exactly half of the respondents said that they felt "a feeling of relaxation" almost all the time. About a third (31%) said that they felt less angry than usual almost all the time. About a fifth (22%) felt drowsy and sleepy almost all the time-and another quarter (25%) felt this more than half the time.

These data suggest-but do not demonstrate-that the effects of marijuana per se are probably not related in any meaningful or causal manner to aggressive or criminal acts. They give us a clue, but do not prove, that in and of itself, marijuana use does not "cause" the commission of criminal acts. In fact, if we were to look at the subjective effects of marijuana themselves, they appear to point in exactly the opposite direction-they would seem to inhibit crime, indeed, activity of all kinds. The effect of marijuana would be more in the direction of reducing than stimulating aggressive, criminal or violent activity.

Two final qualifications would appear to be in order before I explore the marijuana-crime link more systematically and fully. One has to do with the length of time that the marijuana Intoxication lasts. Two facts bear on this issue. First, each episode of use generally produces an intoxication which lasts roughly three or four hours, at the most. Marijuana's effects wear off about three hours or so after the user smokes the drug. Secondly, from previous studies ... as well as the survey whose data I am analyzing In this report, it is clear that the average, or median, level of marijuana use Is roughly once a week. This means that the typical marijuana user is under the influence about three or four hours per week, or roughly 3% or so of his waking hours, and under a "normal" state of mind the remaining 97%. Now, it is possible that this tiny segment of time would influence some users in some significant way, but the fact of its briefness of duration ought to be kept in mind when reasoning about the impact of the drug on the lives of users. The daily user comprises roughly one marijuana user out of 10, and the individual who is high all, or nearly all, of his waking hours, probably constitutes about I or 2% of all marijuana smokers. These facts cannot be ignored in our exploration of the causal connection between the effects of marijuana and criminal activities.

And the last issue I will raise in the marijuana effects topic has to do with the truthfulness of the answers given by users. A plausible objection to taking the word of users concerning the effects of the drug on them would be that they are untrustworthy-that they have a motive for lyIng, for portraying the drug and their experiences in a positive light. Actually, what nearly all researchers have found in an interview situation with drug users is that they very rarely lie; in fact, their honesty about themselves and their activities and experiences appears to be the rule, overwhelmingly. . . .

The overall picture that we receive from . . . various studies is that, like most interview subjects, marijuana users may occasionally lie or hide the truth, but their answers will, in general, be truthful. This does not mean that everything they say must be taken at face value-as with any other group of interviewees, anything which the researcher is capable of checking independently, and which we have data on, we should corroborate with what we learn in the interview situation. At the same time, we would be as fallacious in assuming that everything that marijuana smoker-, said to be true is suspect as if we accepted everything they said to be true in all respects. What all of this means is that we must reason with caution from self-reported data, use them whenever we must, and check them whenever we can.


The Philadelphia survey on marijuana use and crime asked several questions concerning 16 different offenses. The questions began as follows: "'Many people in every community commit acts which others consider offenses, delinquent acts, and violations. Here is a list of these acts." The specific questions dealt with whether the respondent had ever committed each one, how often, how old lie was when the offense occurred, whether lie was caught by the police, and whether the respondent had been drinking, or was under the influence of marijuana, 24 hours or less before committing the act-as well as whether the respondent thought that using marijuana influenced him to commit the act; these questions were asked about the first as well as the most recent such of. fense, if they occurred. The offenses were: hurting someone in a minor way, hurting someone badly, carrying a knife. stealing a car, disturbing people, threatening to hurt someone, taking money, stealing from a store, making an obscene telephone call, forcing sexual intercourse with a woman, breaking into a house or store, damaging property, buying stolen property, setting off an alarm, carrying a gun, and using a weapon to steal. From these facts, it might be interesting to find out what proportion of acts involved some drug use occurring within 24 hours-and whether that drug is alcohol or marijuana. I will explain after the data are presented what these facts can and cannot tell us. We must not be too hasty in reading too much into any given set of data.

Table 2 presents summary information from this set of questions. I have presented figures for only six offenses: stealing from a store, damaging property, hurting someone in a minor way, breaking into a house or store, stealing a ear, and hurting someone badly. Some of the crimes asked about were committed by so few respondents (such as forcing sexual intercourse) that any statistical analysis would be completely meaningless. Other offenses seem to be unrelated to the aggressive syndrome associated with marijuana use-such as buying stolen property. The six I have chosen are, in any case, representative. Tile first three are relatively minor, and would usually be classified by the law as misdemeanors; the second two are considerably more serious, and would often be classified as felonies. Table 2 presents the proportion who drank alcohol, and smoked marijuana 24 hours or fewer before the crime was committed; the figures in the first two columns are for the first time the respondent committed the offense, and the second two columns are for the most recent time, if it occurred more than once.

What generalizations may we make front Table 2?

First of all, committing these crimes is atypical. Not one these six offenses was committed by a majority of the sample. Secondly the more serious the offense, the less likely it was to be committed. Minor offenses were committed at least once by between four and five out of every 10 respondents, but major offenses had been committed at least once by something like one respondent in 25. (And two of the offenses not in the table were rarer still; forcing sexual intercourse was admitted to by six respondents, or about one in 100, and using a weapon to steal was reported by only three respondents, or one out of 200.)

What about drugs and crime? Compared with not being under the influence, drugs tend to be atypical in the commission of crimes. For none of these categories, whether first offense or most recent, was being under the influence of either alcohol or marijuana characteristic of over a quarter of all offenses, and most are considerably below this. Most crimes here were committed when the offender was not under the influence of any drug, marijuana, or alcohol, when lie was in a "normal" state of mind pharmacologically.* Secondly, in terms of absolute incidence, alcohol is involved in the commission of crime considerably more often than marijuana. Adding together all of the offenses committed 24 hours before each drug was used, alcohol had been used in conjunction with first offenses a total of 47 times, and marijuana only five times-a ratio of not quite ten to one. For the most recent offenses, alcohol was involved a total of 66 times, and marihuana 19--a ratio of about three to one.

In conjunction with this point, it should be stressed that the classic, aggressive, violent crimes traditionally and historically associated with the marijuana intoxication are very, very rarely committed-by anyone, high or normal, user or non-user alike. Six men in this survey admitted to forcing a woman to have intercourse with him as a first offense, and three admitted to rape more than the first time. In two out of the six first-time cases, the offender said that he had been drinking 24 hours before, and in one out of the three most recent-time cases, drinking was involved as well. But in none of the six first time cases, and in none of the three recent-time cases, had anyone smoked marijuana 24 hours before. (None of the men had been apprehended for this offense, incidentally.) Only three of the men in the survey said that they had used a weapon to steal as a first offense, and two admitted this offense more than the one time-and none said that lie had either drunk or smoked 24 hours before the offense.

*Several facts should he born in mind when considering this generalization. First of all, the survey asked about the use of each drug 24 hours before the offense was committed or less-and hence, the respondent was not necessarily under the influence when the offense was committed, although certainly many respondents were under the influence at that time. Secondly, anyone who uses marijuana is far more likely to become intoxicated with each episode of use than is true of each episode of alcohol use.

Considering the fact that the typical marijuana smoker in this study had been intoxicated several hundred times for a total of well over a thousand hours, and therefore in a state of mind which, if the "causal" theory is correct, is aggressive and criminogenic, then this almost total absence of violent and aggressive crimes committed by users should be puzzling. This fact indicates that the "causal" theory may be inadequate and erroneous.

However, it is not possible to tell from Table 2 whether the commission of crimes is more frequent for alcohol or marijuana on a relative basis during the period of intoxication. That is, if anyone wanted to know whether a thousand hours of a marijuana or an alcohol intoxication was more like to result in the commission of various crimes, this table would not convey this information. Unfortunately, this survey did not ask a question on the frequency with which the respondents drank alcohol or, indeed, whether they drank at all so that an alcohol -marijuana comparison cannot be more systematically made. It is entirely possible that alcohol-related crimes are far more frequent than marijuana-related crimes. simply because alcohol is still, even in this sample, probably the drug of choice among young people, and is more often used than marijuana (just as not being under tile influence of any drug is more frequently involved with crimes than being under the influence, simply because most people, most of the time, are not intoxicated). As to which drug is more "criminogenic" in the sense of what proportion of the time under the influence of marijuana vs. alcohol one commits various crimes cannot be determined from the data from this study. However, in absolute terms, alcohol is considerably more often used before crimes than marijuana.

Another generalization that may be made from Table 2 is that the contribution of marijuana from first offense to most recent offense grows somewhat relative to alcohol. That is, for the first offense, alcohol is used before it about 10 times as frequently as is marijuana, but for most recent offense, alcohol is only three times as frequent. This probably is a reflection of the following two trends: (1) Marijuana smoking is considerably more frequent today (i.e., closest in time when the most recent offense occurred) titan a few years ago (i.e., closest in time when the first offense occurred), and hence, its frequency relative to alcohol would be greater in conjunction with any activity-criminal or non-criminal as well-simply because its use is more common ; (2) Marijuana smoking is more common among young adults (the age group most closely represented by the age when last offense was committed) than among adolescents and preadolescents (the age group most closely represented by the age when first offense was committed) and, moreover, as age rises, marijuana use rises faster than alcohol use rises, at least up, until early adulthood. . . .

It is possible that, as marijuana usage increases, the number of crimes committed under the influence of this drug may increase as well. However, the question as to marijuana's direct contribution to the commission of crimes, especially aggressive crimes, is ail independent issue, and one in need of exploration.


In this section, it is our job to explore two basic issues:

  1. Do marijuana users commit crimes and offenses any more frequently titan non-users do?
  2. Are any variables with which marijuana use is strongly related also correlated with criminality?

In the next section I will deal with a third question, and one which is probably the most crucial one in this report: Can the marijuana-crime connection be explained mainly by the use of marijuana in and of itself, or is use itself dependent on third variables which themselves explain the commission of offenses?

As the principal measure of committing crimes, this study has employed the number of different types of offenses admitted to in the 16 categories asked about. Naturally, there are many other crimes not asked about, and, in addition, some of these crimes would not correspond to the image most people would have of "classic" aggressive offenses-such as "disturbing people," or receiving stolen goods. However, as a general index or overall measure of criminality of different groups or categories of individuals, this one is as good as any others that have been employed, and is probably adequate for our purposes. (The survey, it must be noted, did not ask questions about any white collar crimes, which must be reckoned into an adequate measure of output of criminal behavior-but these crimes do not correspond with the public's image of aggressive crimes committed under the influence of marijuana, and so their absence is less relevant than would appear at first glance.) Throughout the remainder of this report, the number of crimes and offenses will be employed as the dependent variable-the outcome to be explained. And throughout, I will be examining the marijuana-crime connection-as well as other relationships-insofar as it bears on file basic issue.

Before the analysis proceeds, it should be pointed out that of the 16 offenses, not one of the violent crimes correlated with marijuana use in any meaningful way at all, and a very weak relationship was evidenced with only five of the offenses-stealing from a store, buying stolen property, disturbing people, damaging property, and hurting someone in a minor way (the last of which showing the weakest association of all). The statistical differences in rates of offenses between users and non-users rest on adding together a small number of weakly correlated offenses. When offense differences are discussed, the reader should not hold the mistaken impression that they indicate massive differences, or differences indicating a higher rate of classic, violent crimes among users. With that warning in mind, we may now proceed to user/non-user differences in offenses.

The first relationship to be presented, then, is whether marijuana users as a whole commit offenses any more frequently than non-users as a whole. The answer is yes. I have employed two indicators of marijuana use; one is frequency of marijuana use during the period of most recent use (which may have been in the past, or may be at present), and whether or not the respondent has ever, or has never, used marijuana. Both indicators of marijuana use correlate very powerfully with committing offenses. For both tables, the differences are significant beyond the .001 level, employing Chi-square as a test of significance, which means that the differences observed could occur at random only one chance in a thousand. This is considered extremely significant. There is a regular and step-wise relationship between frequency of use and committing offenses; the more that a given respondent smokes marijuana, the greater is the likelihood that lie will have committed four or more offenses, and the lower is his likelihood of committing no offenses at all. Over a third (38%) of those who have never smoked marijuana said that they committed none of the 16 offenses asked about-but this was true of only one respondent in eight (or 12%) of the regular smokers. At the other end, there was a 17% difference between non-users and regular users (24% vs. 41%) in admitting to four or more of these offenses. Tables 3 and 4 present these data in detail.

We would be remiss in our duties as sensitive and acute social analysts if we ended the analysis there. The simple correlation between marijuana use and offenses may very well mask important and even more basic relationships buried beneath it. Marijuana is correlated in a simple manner with the commission of crimes,, but does it remain correlated when controls are applied? In other words, is it a spurious relationship, or one which will remain when crucial variables are held constant' Which-factors are also related to the commission of offenses-which might actually themselves explain the simple marijuana-crime connection? There are several such variables-race, age, education, the use of other drugs, and the respondent's involvement in the drug subculture. Blacks are significantly less likely to use marijuana (64% of the blacks in the survey said that they had tried marijuana, but 86% of the whites said that they had done so), but slightly more likely to admit to the commission of offenses--41% of all blacks said that they committed four or more offenses but this was true of only 25 ) % of all whites (see Tables 23 and 25). Does the marijuana-crime correlation hold up for blacks and whites separately? Age was also related to both offenses and marijuana. The oldest respondents (age 29-34) were least likely to have tried marijuana (61%), the youngest (15-20) were next least (73%), and the intermediate age groups (21-23 and 24-28) were most likely to have tried marijuana (82% and 80%). There was a linear relationship between age and offenses, however. The youngest group was least likely to have committed no offenses (20%), and most likely to have committed four or more (42%) ; the oldest group was most likely to, have committed no offenses (35%), and least likely to have committed four or more (22%) (see Tables 29-31). Thus, we would want to know whether marijuana and crime still correlate in each age group separately. The same thing can be said for education (Tables 26-28), the use of drugs other than marijuana (Tables 11-16), and having friends who use drugs (Tables 17-22)-the relationship between marijuana use and crime could be mitigated or even transformed altogether if these third variables are considered. One possibility is that the marijuana-crime relationship is wiped out altogether with the application of these controls-that is, that the relationship is completely spurious. A second possibility is that the marijuana-crime connection may be specific to some groups or categories, but not others. And the third possibility is that the same original relationship remains basically unaltered, or even is strengthened, by the application of these controls.

Special attention ought to be paid to the drug-related variables. There is a powerful and significant association between the use of marijuana and the use of other drugs. Although recent studies have presented evidence that this basic relationship is probably not due to the effects of marijuana per se, but to friendships and associations made in conjunction with marihuana use . . . , the simple relationship between marihuana and other drugs is a statistical fact. There is, in addition, a strong and marked association with using drugs other than marijuana and the commission of crimes. Thus, a legitimate question to be explored is whether "marijuana only" users have a greater crime rate than non-users, whether the higher rate of the commission of offenses cannot be attributed largely or solely to using drugs other than marijuana.


Our first clue as to the nature of the underlying relationship between marijuana use and crime is obtained by examining Table 5, which documents the association between number of offenses and marijuana use, taking into account the use of other drugs as well. Table 5 shows that "marijuana only" marijuana users are only very slightly more likely to commit crimes than non-users. Tire original nine percentage point difference between users and non-users in committing four or more offenses has shrunk to only two percentage points (in fact, only one percentage point--we are comparing 24.4% with 25.5%, and the original 19 percentage point difference between users and non-users in committing no offenses has been reduced to 13 percentage points. In fact, the "marijuana only" user is far more similar to the non-user in number of offenses committed than he is to the user of marijuana plus two or more other drugs. The application of the control involving other drugs clearly attenuates the marijuana-crime relationship, and much of the user-non-user differences in offenses committed can be traced to the fact that marijuana users are significantly more likely to use other drugs, rather than the use of marijuana per se.*

*The exploration of the "escalation" hypothesis-the question of whether or not marijuana "leads to" the use of other more dangerous drugs-would take us far afield in this report. The issue is dealt with in a separate report by the author prepared for the National Commission on Marihuana and Drug Abuse, however.

Our confidence in the lack of basic association between marijuana use itself and offenses--and our ability to attribute the simple original relationship to the use of drugs other than marijuana-is strengthened when we examine the relationship between frequency of use and offenses, holding the use of other drugs constant. Table 5A presents these data. Among "marijuana only" users, there is no statistically significant correlation between frequency of use and the commission of crimes--the differences are small, insignificant, and in no consistent direction. Among users of one other drug, likewise, frequency of marijuana use is extremely loosely associated with committing offenses-the differences which are observed are small, statistically insignificant, and point in no particular direction. And lastly, among users of two or more other drugs aside from marijuana, the association between frequency of marijuana use and the commission of the offenses described earlier is loose and not at all significant. Table 5A very strongly indicates the validity of the "spurious" model.

Closely related to the use of other drugs is a control which should be explored: the respondent's involvement with a drug-using subculture-with others who also use drugs. Naturally, as we saw in the last section, individuals who use marijuana tend to have friends who also use marijuana. What we would like to know is whether the social patterns they exhibit can be traced to their use of the drug. or to the fact that they have friends who use drugs. Thus, we should be very interested in looking at the individual who uses marijuana, but who does riot have drug-using friends, as well as the individual who does riot rise marijuana, but has drug-using friends. If interesting changes take place in the original marijuana-crime correlation, then we have indications that it is the individual's social relations, and not his rise of marijuana, per se, that determine marijuana use's correlation with crime. Tables 6 and 7 deal with this issue; they should be examined in conjunction with one another.

Among respondents with no marijuana-using friends (out of their 10 closest friends). as well as those with no drug-using friends at all. almost an identical relationship between marijuana use and crime obtains as in the original Table 3. Marijuana users are, significantly (but not strikingly) more likely to commit four or more offenses, and significantly less likely to commit none. The percentage differences in Tables 6 and 7 are almost the same (even slightly more) as those in Table. 3-about 25% at the "none" end vs. 19% for the original table, and about 18% at the "four or more" end vs. 9% for the original table (Table 3). If we were to rely only on these segments of the table, we would be led to the, inference that marijuana is meaningfully related to the commission of crimes.

However, an extraordinary thing happens to the other segments of the table: among those with drug-using friends, there is no relationship at all between marijuana use and committing offenses. The differences are so small as to be statistically insignificant. (And they are actually in the opposite direction-non-users are slightly more likely to commit four or more offenses than users are.) In other words the marijuana-crime relationship is completely wiped out by the application of controls-in this case, integration into the drug-using subculture. We are led overwhelmingly to the conclusion that marijuana users tend to be somewhat more likely to commit crimes solely because they are part of a drug-using subculture; the actual properties of marijuana appear to be completely unassociated with criminal behavior. Anyone (whether he uses marijuana or not) who makes friends and becomes involved with others who use drugs-especially others who use drugs in addition to and aside from marijuana stands a higher likelihood of committing offenses, simply because this segment of the population tends to be more lax about obeying the law. It is merely because marijuana users tend to associate with others who are part of this subculture that their crime rate is somewhat higher. In other words, the marijuana-crime relationship-in terms of the causal or effects model is completely spurious.

What of the marijuana-using isolate? What of those who use marijuana but who have no friends who use drugs-either marijuana or other drugs? How can we explain the fact that the original difference in offense rate between users and non-users were not wiped out among those without drug-using friends-in contrast to being wiped out among those who had drug-using friends? Looking back at Tables 6 and 7, we are struck by the fact that the offense rate of the marijuana-using group without drug-using friends is as high as the groups with drug using friends. What does this indicate? First of all, involvement with a drug-rising subculture is clearly not the only determinant of a high rate of committing offenses. Probably something else is at work. The fact that the offense differences wash out among those involved with the drug subculture indicates that rise by itself (i.e., being high) cannot be a factor in the original marijuana-crime relationships difference cannot explain two things that are similar. The using isolate, however, is probably a (Qualitatively different social being from those who either do not rise, or those who use and have friends who also use. The isolate is probably deviant in a variety of ways crime being one of them. The fact that he uses in spite of having no friends who use (although lie may have friends who do not use) means that he uses most of the time alone. has not integrated his activities into a social life or a subculture, and probably pursues some activities which the dominant society would judge to be eccentric and unusual. In other words, it is possible that his involvement with crime is related to his social isolation from a group which pursues drug activities similar to his own, rather than his use of marijuana. However, these remarks must be regarded as speculation, since more complete information is not available.

A related process occurs when race is employed as the control variable. Among blacks, the original relationship remains intact; marijuana users are significantly (although not markedly) more likely to commit four or more offenses, and less likely to commit none, than is true of non-users. (The relationship is significant at the .004 level, using Chi-square.) However, among whites, the original relationship is completely washed out; users are not at all more likely to commit offenses. Table 8 presents these data. The minor differences can be completely accounted for by random fluctuations; the relationship is not significant at any level.

When education is treated as the control variable in this relationship, the marijuana-crime connection follows a pattern parallel to the two previous explorations-that is, it is completely washed out for some groups, but remains for others. In Table 9, we see that marijuana users who have at least attended college, but not graduate school, are not any more likely to commit offenses than college non-users. The percentages committing four or more crimes, and those committing none, are almost identical for users and non-users alike. (The differences are insignificant, and due to random fluctuation.) However, among respondents with only a high school education, the same basic relationship produced in Table 3 is upheld: marijuana users are significantly more likely to have committed offenses than non-users. However, at least two variables are compounded here, since most respondents with a college education are white. Thus, the retainment of the original marijuana-crime correlation among high school education respondents, and its wash-out among college educated respondents, should not come as any surprise. However, among respondents with at least some graduate school education, the original relationship asserts itself; users are more likely to commit offenses than non-users. (The differences observed are significant at the .05 level; in other words, the differences have a one in 20 chance of occurring purely at random.) Table 9 presents the marijuana-crime relationship, holding education constant.

Using age as a control variable also produces mixed results. Among the two youngest age groups (15--20 and 21-23), the differences obtaining between users and nonusers are statistically insignificant, and could have occurred by chance alone; in the youngest of these two groups, the non-user is slightly more likely to have committed four or more offenses, and in the next to youngest of these groups, the user is slightly more so--but in both, the differences are too small to be meaningful. In the 24-to-28-year-old group, the differences approach statistical significance, but they are not substantial (.07 level of significance, using Chi-square). However, among the oldest group (age 29 to 34), the same basic difference in offenses as obtained in the original relationship holds up here; users are about twice as likely to have committed four or more offenses as non-users (28% vs. 13%), and less than half as likely to have committed no offenses (21% vs. 58&). These data appear in Table 10.


The central effort in this report has been to determine whether marijuana use, in and of itself, is meaningfully and causally related to the commission of crime. Two models have been used in the past to answer this issue: the causal model, which holds that using marijuana, being under the influence of the drug, actually does stimulate the will to commit antisocial acts, and the spurious model, which holds that marijuana use is merely a reflection of independent and more powerful forces, and that in itself, marijuana use is unrelated to criminal and aggressive behavior. Some small amount of research has been conducted in the past on this issue, but no consistent findings have been turned up; . . . It is because of these factual lacunae that the National Commission on Marihuana and Drug Abuse mandated a study of marijuana use and crime, which culminated in the Philadelphia survey, as well as this report.

The findings from this study strongly support the view that marijuana, use by itself is not related in any meaningful or systematic fashion to criminal behavior, that marijuana use probably does not "cause criminiality. The "spurious" model appears to be a far more accurate description of the marijuana-crime connection than does the "causal" model. The use of marijuana per se is probably completely unrelated to criminal and aggressive behavior. The drug does not "cause" any significant number of users to commit crimes, or aggressive or violent behavior. The effects of the drug seem to be, from what can be gathered from the available data on the question, without criminogenic causality. This does not mean that it is not possible to commit crimes, including aggressive crimes, under the influence of marijuana-but that being high does not increase one's probability of doing so, that on an hour-for-hour, crime-for-crime basis, there are probably no significant differences between being intoxicated on marijuana and being "normal," not under the influence of any drug.

A wide variety of data have been brought to bear on the marijuana-crime issue.

First of all, regarding the subjective self-reported effects of marijuana, all indications are that the effects classically described as being related to aggressive behavior and the commissions crimes-feeling angry, frustrated, wanting to -hurt someone, being willing to follow any and all suggestions of others, being deranged, wanting to do something violent-have no empirical support whatsoever; users consistently describe these "effects" as non-existent or as extremely rare and atypical, no different from normally. In fact, precisely the opposite is the case: if anything, the effects of marijuana would have far more to do with reducing the criminal "impulse," whatever that might be, due to the fact that they tend to be in the direction of relaxation, feeling calm, of not wanting to move about, feeling somewhat drowsy, sleepy. Activity of any sort tends to be inhibited by the marijuana intoxication.

Secondly, the Philadelphia survey asked the respondents about using marijuana and/or alcohol 24 hours or less before various crimes were committed. Marijuana very rarely figured into the commission of crimes in any way. For only a tiny minority of all crimes committed was the respondent under the influence. This was especially the case for serious and aggressive crimes. Alcohol was far more likely to be used soon before criminal activity than marijuana. -

The third type of information brought to bear on the marijuana-crime question was the self-admitted offense rate of users versus non-users. The total number of different types of crimes which respondents said that they had committed was compared. The simple relationship between using marijuana and committing offenses was positive and statistically significant, and there was also a high correlation between frequency of smoking marijuana and committing offenses. However, a wide range of other variables, themselves related to both crime and to marijuana use, were also correlated-race, education, age, the use of other drugs, and having drug-using friends. Thus, the issue became: is it the causal connection with these third variables which produces the marijuana crime simple correlation, or does the correlation hold up even when these factors are controlled? In other words, which is right, the causal or tile, spurious model? Is marijuana use merely dependent itself on larger, broader, more potent factors-or does it exert an independent power? Do users commit crimes more frequently than nonusers because they use marijuana or because they happen to be the kinds of people who, would have a higher crime rate, marijuana or no marijuana.

The evidence from these three-variable tests support the "spurious" model. The control tables show that the differences in crime rate between users and non-users is, in most cases, dependent not on marijuana use per se, but on these larger factors. The fact that the relationship disappears, or is wiped out, when some controls are applied shows that the marijuana-crime connection is dependent on sociological variables, and not chemical effects. When the use of other drugs was held constant, the marijuana-crime correlation was severely attenuated, and in some cases, washed out entirely. When the variable of having friends who use other drugs-both marijuana as well as stimulants, sedatives, hallucinogens and "hard drugs" was applied, the relationship disappeared among those with drug-using friends, but not those who had no drug-using friends. When race was held constant, the marijuana-crime connection was wiped out completely among whites, but not blacks; among white marijuana users, their crime rate was not statistically different from non-users, but black users did have a higher crime rate than non-using blacks-similar in magnitude to the original relationship. When education was used as a control variable, we saw that among college-educated respondents, the marijuana-crime correlation was nonexistent, but it remained among respondents with a high school education, and for those who had attended graduate school. And lastly, age was applied as a control. Among the youngest groups, the marijuana-crime relationship was attenuated, but it remained somewhat strong among the two oldest age categories.

If there were truly a causal relationship between marijuana use and crime, these disappearing correlations would not occur. The application of sociological variables cannot wash out a chemical reaction in the user. The enormous variability dependent on social factors indicates that marijuana effects do not inherently produce behavior of a criminal nature. The kinds of people who use the drug tend to follow Patterns of behavior regardless of whether or not they use marijuana. By itself, marijuana, use is not a potent producer of behavior, and certainly not criminal behavior. Individuals who commit criminal acts are those who would do so with or without the use of marijuana.

The data from this study support-the conclusion, almost without qualification, that marijuana use does not cause criminal behavior.*

*A subsequent regression analysis with the six variables used independently as controls showed that when all six variables are applied simultaneously, the use of marijuana in and of itself accounts for less than 1% of the variance in committing offenses (personal communication by the author, 24 January 1972).


Contents | Feedback | Search | DRCNet Home Page | Join DRCNet

DRCNet Library | Schaffer Library | Major Studies

Marihuana, A Signal of Misunderstanding