Over the years a number of approaches have been utilized in an effort to assess the relationship between marihuana use and violent crime. Perhaps the simplest approach is to compile a. list of violent offenses allegedly committed by marihuana users and to establish, retrospectively, the role of marihuana in the commission of these offenses.

The Indian Hemp Drugs Commission (1894), for example, investigated 81 cases of violence allegedly caused by hemp drugs in in effort to determine whether or not a causal relationship existed. Of these 81 cases, 11 were too old to permit adequate investigation. In 23 of the cases examined, however, 18 showed no evidence of a connection between the crimes and the use of hemp drugs. The Commission concluded that:

In respect to his relations with society, however, even the excessive consumer of hemp drugs is ordinarily inoffensive,. His excesses may indeed bring him to degraded poverty which may lead him to dishonest practices; and occasionally, but apparently very rarely indeed, excessive indulgence in hemp drugs may lead to violent crime. But for all practical purposes it may be laid down that there is little or no connection between the use of hemp drugs and crime (Indian Hemp Drugs Commission, 1893-1894, reprinted 1969: 204).

Similarly, in 1938 the Foreign Policy Association published the accounts of 10 marihuana crimes, including murder and assault, "culled at random from the files of the U.S. Bureau of Narcotics" (Merrill, 1938: 28). These cases were presented in such a way as to imply that marihuana caused the offenses. According to Grinspoon (1971: 302), they "gave the reader the distinct impression that the -user of marihuana was a violent criminal who was given to rape, homicide, and mayhem." Bromberg (1939), however, questions the validity of the causal assumption.

It is difficult to evaluate these statements, because of their uncritical nature.... Among the ten patients, the second, J. O., was described as having confessed how lie murdered a friend and put his body in a trunk while under the influence of marihuana. J. 0. was examined in this clinic (Bellevue Hospital) ; although lie was a psychopathic liar and possibly homosexual, there was no indication in the examination or history of the use of any drug. The investigation by the probation department failed to indicate use of the drug marihuana. The deceased, however, was addicted to heroin (p. 9).

Based on retrospective case analyses, some observers have attempted to specify more precisely. the nature of the purported relationship or the situations in which aggressive behavior may result from marihuana use. Bromberg (1939), for example, suggested that aggressive or violent behavior may arise when a naive subject develops a panic state in response to marihuana-induced hallucinations. Allentuck and Bowman (1942) believed that aggressive or antisocial behavior following use may occur as a reaction to some unpleasant external stimulus during the phase, of hypersensitivity and heightened psychomotor activity. Others have suggested that antisocial conduct of an aggressive or violent nature may occur when marihuana is used, as alcohol often is, to release repressed feelings of hostility (Siler, et al., 1933; Chopra and Chopra, 1939; Allentuck and Bowman, 1942; Freedman and Rockmore, 1946; Murphy, 1963), and to serve as a fortifier for aggressive or violent crimes (Ewens, 1904; South Africa Interdepartmental Committee on Abuse of Dagga, 1952; Ames, 1958; Watt, 1961; Blumer, et al., 1967; Miller, 1968).

The available evidence bearing on these issues, however, suggests that panic reactions rarely occur; that psychomotor activity is more often reduced than enhanced following use; that aggression rarely follows use, but when it does, it generally occurs among individuals with histories of maladjustment, emotional instability or impulse disorders (Bromberg, 1934, 1939; Charen and Perelman, 1946; Ausubel, 1958; Bloomquist, 1968; Grinspoon, 1971; Kaplan, 1971; National Institute of Mental Health, 1972).

After a series of studies of marihuana and crime, Chopra and Chopra (1939) concluded that if any relationship existed between marihuana use and violent crime, it was an indirect one. They stated that:

So far as premeditated crime is concerned, especially that of a violent nature, hemp drugs . . . may not only not lead to it, but they actually act as deterrents.... One of the, important actions of these drugs is to quieten and stupify the individual so there is no tendency to violence. . . . The result of continued and excessive use of these drugs in our opinion is to make the individual timid rather than lead him to commit a crime of a violent nature (p. 92).

Over the years, the conclusion of the Chopras has received increasing support from many quarters of the research community (Mayor's Committee on Marihuana, 1944; Maurer and Vogel, 1962; White House Conference on Narcotic Drug Abuse, 1962; Murphy, 1963; President's Commission on Law Enforcement and Administration of Justice, 1967; National Institute of Mental Health, 1970, 1972).

In the absence of possibilities for addressing the issue more directly, several researchers have relied on statistical studies and have sought to establish the overall and comparative incidence of detected violent crimes among cannabis users. One method has been to compile lists of violent crimes committed during specific periods of time and to determine the proportion of these offenses committed by cannabis users.

Lambo (1965) compiled a list of crimes occurring in three West African countries during a recent two-year period. He claimed that users of cannabis had committed 51% of the 73 murders, 31% of the 263 cases of assault and battery and 26% of the 472 cases against women.

Some have preferred to base their statistical studies on samples of offenders (rather than lists of offenses) drawn from the arrest or conviction files of law enforcement agencies. Several researchers adopting this method are content to identify the marihuana users in their samples and then simply report the number of users charged with violent crimes or the proportions of the total number of violent crimes perpetrated by the users.

The District Attorney of New Orleans, for example, testified before the House Ways and Means Committee that of the 450 men convicted of major crimes in 1930, 125 were identified as regular marihuana users. Approximately one-half of the murderers and one-fifth of those charged with assault, robbery or larceny were said to be regular marihuana users (U.S. House of Representatives, 1937: 23-24).

Bromberg (1939) reviewed the records of 16,854 offenders in the psychiatric clinic of New York County's Court of General Sessions during the period 1932 to 1937. Of the 67 marihuana users identified, only six had been charged with violent crimes. He concluded that there was no causal relationship between marihuana use and aggressive crime.

Others go one step further and attempt to compare the users' rates of violence with those of other selected populations such as non-marihuana using offenders, offenders using other drugs, or all of fenders in a given file.

Bromberg and Rodgers (1946) studied the civilian and military criminal records of 8,280 convicted offenders at the United States Naval Prison in Portsmouth, Now Hampshire between January 1, 1943 and July 1, 1945. Of the total number of offenders investigated, 40 or .0048% were identified as marihuana users (23 used to excess, 10 were moderate users and seven were described as light users). Of these 40, only two reported being more aggressive while under the influence of marihuana than they would be under normal conditions and three had been charged with violent crimes (assault or striking an officer) while in the military. Comparison of the users' criminal records with those of 40 randomly selected non-using prisoners revealed that the non-user group had committed more aggressive crimes than the users.

In conclusion, the researchers stated that:

There is no positive relationship between aggressive crime and marihuana usage in the Naval service; . . .
. . . there is no significant causal relationship between aggressive crime in civilian life (of the naval offenders studied) and the use of marihuana . . . .
Marihuana usage is but an aspect of some type of mental disorder or personality abnormality (p. 826).

Maurer and Vogel (1962) have stated a similar conclusion.

"It has not been our impression from contact with many hundreds of marijuana users that these people are violent criminals; ... While there may be occasional violent psychopaths who have used marijuana, have committed crimes of violence, and who have, in court, explained their actions as uncontrolled violence resulting from the use of the drug, these are exceptions to the general run of marijuana users.... Marijuana is not possessed of any mysterious power to force people to commit acts which they would not otherwise perform (p. 281).

Blum (1969) reviewed the data provided in 1966 by the Federal Bureau of Investigation as part of its Careers in Crime Project. The data showed that marihuana users did not differ significantly from either heroin users or from all offenders (drug and non-drug users) in their rate of violent crime relative to their total non-drug offenses (28%, 26% and 26%, respectively).

There have been several statistical studies, also using offender populations, designed to assess the degree to which persons arrested for cannabis use have previous or subsequent arrests for violent crimes.

Bromberg and Rodgers (1946) found that of the 40 offenders in their sample identified as marihuana users, 12 had been charged with previous offenses, three of them for assault.

Gardikas (1950) reviewed the criminal records of 379 persons arrested between 1919 and 1950 for publicly using hashish. Of these, 117 (31%) reportedly progressed from hashish use to other crimes, about one-third of these subsequent crimes involving violence or weapons. Gardikas stated that one subgroup of these, offenders accumulated 420 offenses of assault, woundings, threats, robberies, and manslaughter (p. 5).

In sum, these statistical studies based on samples of violent offenses, violent offenders or arrested marihuana law violators indicate that some individuals identified as marihuana users do commit violent crimes, have committed them in the past and go on to commit them in the future.

Their numbers, however, are generally small, both absolutely and relatively. These studies therefore suggest a very weak and insignificant statistical association between marihuana use and violent crime which may itself be completely attenuated when the proper statistical controls are applied.

These studies do not establish a causal relationship between cannabis use and violent crime; nor do they permit an affirmative response to the crucial question of whether the use of marihuana alters the progression to violent crime at a significantly greater rate than that which might be expected from some other criminal subgroups, more representative samples of cannabis users, or samples drawn from the general population. They also fail to address themselves to the external conditions and circumstances which might serve to mitigate the observed relationships.

Several recent and more sophisticated empirical investigations have addressed some of the critical questions left unanswered by statistical studies of offender populations. Namely, they set about to determine whether marihuana users in the general population commit acts of aggression or violence significantly more frequently than do nonusers; and whether any observed differences between users and non-users may be more directly attributable to extra-pharmacological (social, cultural, psychological) variables than to the use of marihuana itself or to the pharmacological action of the drug per se.

In 1965, Robins and his associates (1970) conducted a survey of 20 black men born in St. Louis between 1930 and 1934. The marihuana users in the sample were then compared with the nonusers relative to their ratings on a "violence syndrome" constructed by the researchers.

Respondents were rated high on the violence syndrome if they reported having participated or felt like participating in three or more of the following items, one of which was a judgment by the interviewer that the respondent demonstrated hostility during the interview:

1. Getting hurt in a fight.

2. Ever feeling like killing someone.

3. Ever hitting people when angry.

4. Being quick to lose one's temper.

5. Throwing or breaking things when angry.

6. Ever hurting someone in a fight.

7. Ever fighting with a weapon.

8. Getting mean when drinking.

9. Interviewer's observation of respondent's hostility.

The researchers found that those who had used marihuana during adolescence were more likely to score high on the violence syndrome than were those who did not use the drug for the first time until adulthood; 31% of the respondents who had used marihuana and no other drug during adolescence exhibited three or more of these measures, 24% of those who used marihuana only but started as adults scored high, 16% of the non-users were classified as high on the violence syndrome, and 45% of those who used marihuana and other drugs were so classified.

The researchers also found that users were significantly more likely to report the, commission of " adult person or property offenses" than were nonusers; 32 % of the nonusers, 48 % of the marihuana only adult starters, 56% of the marihuana only adolescent starters, and 77% of the multiple drug users reported committing crimes against persons or property as adults.

Based on these data, marihuana users were said to be significantly more likely to have exhibited violent behavior than were nonusers. This conclusion, however, seems somewhat premature.

First, upon the application of three "pathological" controls (dropping out of school, alcoholism and involvement in juvenile delinquency), the original relationship was reduced to some unspecified extent; the researchers reported only that these controls "failed to completely wipe out" the statistical correlation. Secondly, several of the items would not appear to be sufficiently discriminating in that large numbers of people, whether drug users or not, have probably been quick to lose their temper, hit people when angry (most parents, for example), or felt like killing someone in a moment of anger. The subjective nature of the last item the interviewer's observation of hostility has already been noted.

Thirdly, neither the drinkers nor those who were multiple drug users were isolated in the analysis, making it impossible to determine whether or not the apparent relationship between marihuana use and violence may have been a. function of these other drugs rather than the marihuana.

Lastly, but perhaps most importantly, the researchers found several other "pathological" variables correlated with the use of marihuana in their sample: (a) low income, (b) low status jobs, (c) unemployment, (d) receiving financial aid, (e) failing to graduate from high school, (f) fathering illegitimate children, (g) marrying women who had been married previously or who had children, and (h) drinking "heavily enough to create, social or medical problems."

Because none of these variables was controlled in the data analysis (despite the fact that most of them have been found to be significantly related to delinquency and crime and characteristic of persons involved in delinquent, criminal and drug subcultures), it is impossible to determine whether or not any or all of these variables played a mediating role in the observed relationship between marihuana use and antisocial behavior in the, sample. As Goode (1972) has noted, the fact that the three controls which were applied did reduce the relationship, makes it likely that "additional controls would reduce the relationship even more, indeed, reduce it to zero, if applied simultaneously" (p. 13).

As such, although this study represented a significant step forward in investigating the relationship between marihuana and crime or violent behavior, its methodological limitations preclude generalizations of the findings to the larger universe.

In a study of drug use among lower class minority group youth, Blumer and his associates (1967) found that marihuana users were much less likely to commit aggressive or violent acts than were those who used amphetamines or preferred alcohol, and that most of the marihuana using youths deliberately shunned aggressive behavior and adopted, instead, a "cool," non-violent style. The researchers did find a small group of youths (termed "the rowdy") who were oriented toward aggressiveness. Generally, these youths preferred alcohol over other drugs and were found, for the most part, to have been raised in an aggressive and combative social milieu. The researchers point out, however, that most marihuana users in the sample were not of the rowdy type, even though it is this small group which often forms the basis of the public and police image of the youthful marihuana user. In commenting on the role of marihuana in the passage of youth from rowdy to a cool style, the authors note, that the passage from the rowdy type to a cool and mellow youngster, as it relates to the use of drugs, involves chiefly a shift to the smoking of marijuana. . . . [The youngsters'] accounts and discussions also stress that the use of marijuana both produces and symbolizes a "mellow" mode of conduct that is opposed to that associated with rowdy behavior. They place great weight on the "socializing" effects of marijuana use, declaring that its use not only leads youngsters away from violence but has the effect of changing them into social human beings (p. 30).

In a large scale, systematic survey (questionnaire and interview) of more than 1300 students at five West Coast colleges and universities, Blum and his associates (1969) found that 19% of the total sample had used marihuana but that 94% had used alcohol. One, percent of the marihuana users reported fights or other criminal behavior which they attributed to the drug. Of those who used alcohol, 8% reported fights and 2% reported offenses while under the influence of this drug. The researchers make particular note of the fact that despite the increase of marihuana use on these campuses since the middle and late sixties, there has been no comparable increase in assaultive crimes.

In a Commission-sponsored household survey of 15 to 34 year-old male residents of West Philadelphia, Goode (1972) found that not one of the violent crimes, including "forcing sexual intercourse," was significantly correlated with marihuana use. Among the five offense types showing a very weak relationship to the use of marihuana, the only so-called violent offense, "hurting someone in a minor way," showed the weakest association. The author notes that "the statistical differences in rates of offenses between users and nonusers rest on adding together a small number of weakly correlated offenses. . . . [They do not indicate] massive differences, or differences indicating higher rates of classic, violent crimes among users" (p. 32a).

These more rigorous studies of the relationship between marihuana use and violent crime suggest that marihuana users in the general population do not commit acts of aggression or violence significantly more frequently than do nonusers; that marihuana does not heighten aggressive tendencies in most users and may, in some cases, serve to reduce aggressiveness; and that much of the observed relationship between marihuana and violence is probably a function of social, cultural or psychological variables such as multiple drug use, set and setting and involvement in a criminal or drug subculture.

There is no reason to believe that marihuana use will cause or lead to the commission of aggressive or violent acts by the large majority of psychologically and socially mature individuals in the general population.