Community Attitudes to Cannabis use in Western Australia: Part I

Abstract

This paper presents results of a telephone survey of 400 Western Australians regarding attitudes to laws relating to possession of cannabis for personal use. Over a third of respondents believed cannabis should be made as legal as alcohol. Support for decriminalization increased from 64.0% to 71.5% when possible penalties associated with decriminalization were described. When penalties were described, more women than men favored decriminalization but age, political affiliation, and city or country residency no longer predicted attitudes to decriminalization. Almost two-thirds of respondents believed that many people used cannabis without experiencing serious problems and that the court system was overburdened by minor cannabis offenses. Half the sample believed it would not be a bad thing for the community if people were legally able to grow cannabis for their personal use. Results suggest there is considerable community support for removing criminal penalties for simple cannabis offenses.

a. Legislative Options
b. Public Opinion on the Legal Status of Cannabis
c. The Present Study

In a number of countries, including Australia, the issue of changing the laws regarding cannabis use has received increasing attention in the media and the scientific community over recent years (Makkai and McAllister 1993 Shapiro 1994- Single 1989). Typically the discussion has centered around the appropriateness of penalties for the possession of small amounts of cannabis for personal use. Perhaps the exception to this has been the United States where, fueled by fear of the high level of violent crime attributed to illicit drug use, notably cocaine and crack, the academic debate has in recent years appeared more radical and has canvassed the legalization of all illicit drugs (MacCoun et al. 1993; Trebach and Inclardi 1993). World-wide, the debate has intensified as more academics,members of the public and now politicians have begun to view the prohibitionist "war on drugs" as a failure.

Public policies on drug use often involve moral and religious arguments, and issues of individual versus collective rights. Makkai and McAllister (1993) have argued that in democracies there is little incentive for political parties to compete for votes on policies that could undermine their support among the electorate. Governments often follow public opinion. This has been relatively conservative on the issue of illicit drug use and consequently government policy has been extremely cautious. To put it crudely, politicians are "vote-junkies" and are unlikely to do anything that they perceive will comprormise their supply when their electoral fix Is due at each election.

Thus, the measurement of public opinion about legal options concerning cannabis use, and documenting and disseminating these findings to the general public and the politicians whom they vote for, is likely to be a powerful way of affecting the debate on this issue. If public opinion is strongly against liberalization of laws, then it is likely that the issue will be given low political priority. If, however, as opinion polls suggest, support is growing, then this may initiate a more rational consideration of the range-of policy options by politicians, who ultimately are responsible for framing and passing legislation.

a. Legislative Options

Conceptually, there are a number of ways of dealing with cannabis within the legal system. Six possibilities have been differentiated (McDonald et al. 1994). These are total prohibition without an expediency principle, total prohibition with an expediency principle, prohibition with civil penalties for minor offenses (decriminalization), partial prohibition, regulation, and free availability. In the present study the three most discussed ways of dealing with possession and use of small amounts of cannabis within the legal system were addressed. These were chosen as markers along the continuum of legal options and do not represent all possibilities.

The first of these, total prohibition, makes the possession and use of cannabis, even of small amounts used in private, a criminal offense. If an act (such as possession and use of small amounts of cannabis) is a criminal offense, conviction can result in penalties including a criminal record and a jail sentence. In Australia Custodial sentences are rarely given for simple cannabis offenses, although in Western Australia, until recently, small numbers of these offenders have received short prison terms for non-payment of fines (Lenton 1995).

The second possibility is decriminalization. It has been pointed out that although in the cannabis debate there is no consensus on the meaning of decriminalization, the term "usually refers to a reduction of penalties for possession of small amounts of cannabis to penalties other than imprisonment" (Single 1989-219). Perhaps a more accurate description of decriminalization is the removal of criminal penalties. Offenses that are not criminal are still unlawful. Convictions do not result in a criminal record but typically only a fine. In this scenario incarceration as a sentencing option would only be utilized rarely, for example for non-payment of fines (Sarre et al. 1989). Speeding in a motor vehicle is an example of such an offense. Possession and use of small amounts of cannabis could be decriminalized, whereas the possession of large amounts, use in a public place, selling, and supplying the drug could remain criminal offenses.

The third possibility is for cannabis legalization. If the possession and use of cannabis was legalized, it would not result in any legal sanctions, although otherspecific acts in relation to cannabis may be deemed unlawful by legislation. For example in Western Australia, as in many other places, the possession and use of alcohol and cigarettes is lawful, although the sale of cigarettes or alcohol to minors is not, nor is the promotion of such substances through certain forms of media. Thus legalization allows some degree of regulation, rather than the more extreme free availability option that is unlikely to be considered a viable alternative by the community.

In Australia, as in the United States, various states and territories have different laws and penalties relating to cannabis. Legislation relating to large scale cultivation and supply of cannabis is similar across the various jurisdictions, with provision for substantial fines and Jail terms for these criminal convictions. However, although in the majority of states and territories criminal penalties also apply to simple possession and use of small amounts of cannabis, in two jurisdictions such offenses have been decriminalized

The possession and use of small amounts of cannabis in private was decniminalized in South Australia in 1987 through the so called Cannabis Expiation Notice System (Morgan et al. 1993) and in the Australian Capital Territory in 1992 under the Drugs of Dependence Amendment Act (Advisory Committee on Illicit Drugs 1993). Similar measures have recently been considered in Queensland (Advisory Committee on Illicit Drugs 1993) and are being discussed in the media with respect to other state and federal jurisdictions. In 1993, the Ministerial Council on Drug Strategy (MCDS), a group comprising health and law enforcement ministers from each Australian state and territory, established a National Task Force on Cannabis under the auspices of the Commonwealth Department of Health, Housing, and Community Services to report to state and federal ministers on a number of aspects of the cannabis decriminalization debate. The National Cannabis Task Force reported to ministers in September 1994. Subsequently, the Federal Justice Minister announced that further research into the social impact of laws against cannabis would be conducted. One (Bowman and Sanson-Fisher 1994) of the four reports commissioned by the MCDS, and presented to them by the Task Force, was a survey of public opinion on cannabis and legislative optionsfor its control(1). This reflects legislators' interest in community attitudes regarding this issue.

b. Public Opinion on the Legal Status of Cannabis

Support for legalization of cannabis has fluctuated over time in both the United States and Australia. American Gallup surveys conducted from 1969 to 1985 asked- "Do you think the use of marijuana should be made legal or not?". According to these surveys, support for legalization peaked in 1977 (28% in favor), and subsequently declined to 23% In 1985 (Inciardi and McBride 1989). Makkai and McAllister (1993) attributed the reduction in support since the mid-1970s to the conservatism of youth in the 1980s, relative to their predecessors, and the harsher econon-dc circumstances in which they have grown up. However, national surveys conducted for the U.S. Department of Justice have recently shown an increase in support for cannabis legalization from approximately 16%, in surveys between 1987 to 1991, to 22% in 1993. Similarly, the 1993 survey of high school seniors in the United States found 22.9% supported legalization of cannabis use, 18.7% supported decriminalization, and 43.4% favored keeping it a crime. The level of support for legalization among high school seniors in 1993 was the highest it had been since 1991 (Maguire and Pastore 1994).

In Australia, Roy Morgan Research polls suggest that the proportion of the Australian community that believes smoking of cannabis should be made legal increased from 9% in 1970 to a peak of 3 1 % in 1984 and decreased to 25% in 1991. The household surveys conducted for the National Campaign Against Drug Abuse (NCADA) indicated that support for legalization was variable. In 1988, 26% of those surveyed agreed that "Smoking of marijuana should be legalized," in 1991, 38% agreed with the proposal, but in 1993 only 25% agreed with "the personal use of marijuana being legal" (Advisory Committee on Elicit Drugs 1993). In a national telephone survey conducted at the same time as the present study, Bowman and Sanson-Fisher (1994) found that between 52% and 55% of respondents believed that growing cannabis for personal use, possessing cannabis for personal use, using cannabis, and possessing implements for its use should be legal, which reflected greater support for this option than had previously been found in opinion surveys. Males were more likely than females to support legalization of activities associated with the use of small amounts, and support for cannabis legalization generally was strongest among younger respondents. However, the vast majority (between 75% and 85%) of those surveyed believed activities involving large amounts or supplies of cannabis should be illegal.

In an analysis of the 1993 NCADA household survey data, McAllister (1993) argued that the factors that predict those most likely to support legalization of the drug, in addition to whether or not respondents admit to ever having used cannabis, were being male, being aged between 20 and 39 years, and havingsecondary or post-secondary education. The weaker support for legalization among adolescents was again seen as indicative of youth conservatism, which was a more general finding. Significant differences in public opinion were also found between the various Australian states and territories. Support for legalization was highest in Western Australia (30%) but as low as 21 % in Queensland. Political affiliation has also been linked to attitudes to cannabis law. Makkai and McAllister (1993) analyzed the results of a 1990 electoral survey and reported 60% of LiberalNational (right wing) voters favored prohibition compared with only 41 % of Labor and Democrat (left wing) voters.

There is little published public opinion survey data from the United States that refer to attitudes to decriminalization as defined above, although Inclardi and McBride (1989) report the results of an ABC news poll conducted in 1986 in which 24% of respondents stated they favored "legalizing the possession of small amounts of marijuana for personal use." In Australia, Roy Morgan Research polls indicate that support for this option has remained at between 45% and 49% over the five surveys conducted from 1979 to 1987 (Advisory Committee on Illicit Drugs 1993). Bowman and Sanson-Fisher (1994) found that approximately 75% of the sample in their 1993 Australian survey believed that growing or possessing cannabis for personal use and using cannabis should not be criminal offenses.

c. The Present Study

The data presented in this paper were collected as part of a telephone survey of community attitudes toward the reduction of harm associated with illicit drug use including provision of needles to drug injectors and the possibility of changing the laws relating to cannabis (Lenton 1994).It was expected that the general public would have difficulty making distinctions between prohibition, decriminalization and legalization given the complexity of the issues. Anecdotal evidence suggested that many people believed that the alternative to criminalization was legalization, and that survey respondents would require an understanding of all three concepts if they were to meaningfully reflect on their attitudes to the possibilities of changing the laws relating to cannabis. As a result the items referring to cannabis legislation in this study were designed to distinguish the terms by using analogies that most respondents would understand with little explanation. In addition these items were ordered in such a way as to assess attitudes towards (and define), legalization first, then prohibition, indicating the status quo (keeping cannabis possession and use criminal), and finally decriminalization. It was hypothesized that responses to these items would be different and that more persons would be in favor of decriminalization when they understood all three terms rather than just one or two.

A total of 400 telephone interviews were conducted with members of the Western Australian general public 17 years old and older. Two hundred and fortyfour (61%) respondents were from the metropolitan area and 156 were from two regional centers, Bunbury (27%) and Geraldton (12%). The sampling and data collection were conducted by the marketing and social research company Reark Research Pty. Ltd. Interviewing was conducted between the November 26 and December 5, 1993.

Survey respondents were selected at random using CD ROM "Australia on Disk" to select the household. Within households, the closest birthday method was used to select the respondent. If the person selected by this method was not available at the time of calling, up to three call backs were made to contact them. No substitution within households was allowed. Most telephone calls were made after 5 p.m. on weekdays and during the day on weekends so as not to bias the sample against working persons. Toward the end of the data collection, telephone interviewers asked to speak to the man in the household whose birthday was closest to the date of call, to ensure approximately equal numbers of males and females in the final sample. Thirty-eight percent of contacts with eligible respondents resulted in a complete interview. Four (1%) of the 404 interviews commenced were terminated by the respondent prior to completion.

The questionnaire designed for this study took, on average, just under 25 minutes to administer by phone (range 13 to 55 minutes). It included both closed and open-ended items, questions pertaining to cannabis decriminalization as well as items covering knowledge and attitudes to injecting drug use, spread of HIV, and methods of reducing spread including needle and syringe provision. The HIV related data Will be reported elsewhere. Five Likert scale items related to cannabis were presented. In response to these statements respondents were asked to say whether they agreed or disagreed with the statement, and whether they believed this strongly or only somewhat. Respondents were subsequently asked for demographic information, including their political and religious affiliation. Respondents were not asked about their own drug use in this study.As the majority of variables used in the study were categorical, cross-tabulation was used for descriptive purposes for two-way comparisons. Association between variables in the study was tested using the Pearson x² test, unless otherwise stated. Comparisons between results of this study and those from other sources (e.g., population statistics) were conducted using the non-parametric x² one sample test, which allows specification of expected frequencies or proportions (SPSS, Inc. 1990). Unless otherwise stated, two-way comparisons for dichotomous variables were conducted using the non parametric McNemar Test (SPSS, Inc. 1990) which is not violated by cells of small, or zero, frequency.

Backward stepwise logistic regression analyses were carried out using the LOGISTIC REGRESSION command through SPSS (Release 4) (SPSS, Inc. 1990) to determine which of a number of other variables of interest predicted those scores on particular items. Unweighted data were used for these analyses. All other data were weighted for age (see below). All hypothesis testing was conducted allowing for a 5% type I error rate.

Results

a. Sample Characteristics
b. Knowledge, Attitudes, And Beliefs Regarding Cannabis Use and Legal Options

a. Sample Characteristics

The sample (n=400) comprised 45.3% men and 54.8% women. The age distribution was 5.8% were 17 to 19 years, 19.3% were 20 to 29 years, 28.5% were 30 to 39 years, 20.0% were 40 to 49 years, 11.5% were 50 to 59 years, and 14.5% were 60 years or over. There were no differences found between the metropolitan and the country samples with respect to gender (x²=0. 1072, df=1, and not significant [NS]) or age (x² =7.877, df=6, and NS).

Comparison of age and gender data for the sample as a whole, with estimates of population gender and age at June 30, 1991 (Australian Bureau of Statistics 1993), indicated no difference was found between the sample and the population with respect to gender (x²=3,460, df=1, and NS), but a difference was found with respect to age (x²=12.002, df=5, and p<.05). Persons in the 30 to 39 age group were overrepresented in the sample (x²=8.258, df=1 and p<.005), whereas those over 60 years of age were underrepresented (x²=3.901, df=1, and p<0.05). There were no other significant differences between the sample or population statistics for any of the other age groups. Previous research and preliminary analyses indicated age was likely to have a significant impact on responses. To control for the age bias in the sample, the data, unless otherwise stated, have been weighted.

Over half of the respondents (54.3%) were living in a two parent household with dependent children. The vast majority of respondents (71.0%) had children, about one in three (29.5%) of all respondents had children 22 years or over, and about one in five had children 6 years or under. Overall, 50.7% of respondents had children under the age of 22 years. Respondents were asked their religion and the importance of religion to their everyday lives. Just under one quarter (23.1 %) described their religion as very important and 34.6% as somewhat important whereas 41.6% rated religion as not very or not at all important. Respondents were also asked who they voted for in the last Western Australian state election (held on February 6, 1993). Responses to this question were recoded into night wing or left Wing where possible. (2) There were no differences found between these figures and voting patterns for the electorate as a whole supplied by the Western Australian Electoral Commission, which were similarly recoded (x²= 0.399, df=1, and NS).

b. Knowledge, Attitudes, And Beliefs Regarding Cannabis Use and Legal Options

Responses to knowledge-based and attitudinal Items regarding cannabis and legal options are presented in tables 1 through 3. Table 1. shows that almost two thirds of respondents remembered recent media reports concerning cannabis, and a similar proportion thought simple cannabis offenses should no longer be criminal offenses. Just over a third of the sample thought cannabis should be as legal as alcohol. Comparisons between country and metropolitan samples for these items revealed no significant differences, therefore, the separate results are not shown.

Respondents were asked to give their reasons for their response to Item 3. The main reasons given are presented in table 2. The most frequent reasons given for maintaining criminal sanctions related to the impact of sanctions on prevalence of use, whereas those reasons given for removing criminal sanctions related to the perception that the drug was not particularly harmful. Responses to the six Likert statements are presented in table 3. Respondents were equally divided as to whether cannabis use led to use of other drugs. Almost three-quarters of the sample agreed that penalties for personal use should be "like those for speeding in a motor vehicle, they should get a fine but not a criminal record." The majority agreed that cannabis could be used without serious problems, and the court system is overburdened by minor cannabis offenses." Just over half the respondents Supported allowing the cultivation of small amounts of cannabis for personal use.

It was hypothesized that respondents would be more in favor of cannabis decriminalization when the penalties were explained. To test this hypothesis a comparison was made of responses to Items 3 and 5. In Item 3 only criminal penalties were described but when Item 5 was presented likely non-criminal penalties were given by way of explanation of the meaning of the term decniminalization. Data for the Item 5 were transformed so as to collapse the "strongly" and "somewhat" in favor categories and "strongly" and "somewhat" against categories to produce an "agree with decriminalization' category and a "disagree with decriminalization' category, respectively. Responses in the "neither" and "don't know" categories were collapsed into "unsure/don't know." This comparison is shown in figure 1. There was a significant difference between the proportion of respondents 'in favor of decriminalization when the term was explained (Item 5), compared to when it was not explained (Item 3). (x² McNemar = 18.933, df=4, and p<.001). Furthermore, most of this difference was due to 24 (68.8%) of the 34 persons in the "unsure" or "don't know" group and 45 (40.9%) of those in the "not in favor" group in Item 3 moving in favor of decriminalization in Item 5 when the likely penalties associated with the decriminalization of cannabis were described. Only 36 (14%) persons who stated they were in favor of decriminalization in Item3 stated they were against it in Item 5. The majority (70.9%) of people in this group supported cannabis being made as legal as alcohol. Table 4 presents a comparison of those who changed their support for and against decriminalization from Item 3 to Item 5.

A factor analysis using the principal components extraction method was employed to convert responses to Items 4, 6, 7, and 8 to a reduced number of scales. To do this, "don't know" and "neither" responses were combined to produce 5-point scales. The analysis produced a single factor with an Eigenvalue of 2.02292, which accounted for 50.6% of the variance. The factor loadings for each of the variables are presented in table 5. Note that high scores represent disagreement with the item. Conceptually the factor seems to represent a position that is favorable to cannabis law reform. On the basis of this analysis a new variable pro-cannabis (PROCAN) was created using the formula- Item 4 score + Item 8 score -Item 6 score - Item 7 score = PROCAN. This new variable had a range from -8 to 8 with positive scores indicating a pro-cannabis law reform position.

Three logistic regression analyses were carried out to determine which factors predicted those persons who were in favor of the various legal options for cannabis. The first looked at predictors of attitudes to cannabis legalization. Given the findings shown in figure 1, two analyses regarding support for decriminalization were conducted. One looked at predictors of attitudes to cannabis decriminalization in Item 3 when the term was not explained, the second looked at predictors of attitudes to cannabis decriminalization in Item 5 when the term was explained. The predictor (independent) variables entered into the regression equation for all three analyses were: age, gender, political affiliation, religious affiliation, whether respondents worked in the medical or allied health area, whether respondents were from the city or the country, whether they had seen anything in the media regarding cannabis, whether they had children under the age of 22 (thought to be the most worrying age group for parents in terms of their children's' future or current substance use), and their score on the PROCAN variable.

To examine predictors of attitudes to cannabis legalization, Item 2 was recoded and responses were dichotomized by excluding respondents who answered don't know. Responses that cannabis should be made as legal as alcohol were coded as 1 and responses that it should not were coded as 0. To examine predictors of attitudes to cannabis dechirruinalization, Item 3 was recoded and responses were dichotomized by excluding respondents who answered don't know. Responses that cannabis should remain criminal were coded as 0 and responses that it should not remain criminal were coded as 1. To exanuine the predictors of responses to Item 5, the Likert scale item was transformed into a dichotomous variable: scores of strongly or somewhat agree with decriinninalization were collapsed into agree, with a value of 1, and strongly and somewhat disagree were collapsed into disagree with a value of 0.

Table 6 shows that support for making cannabis as legal as alcohol was predicted by age, political affiliation, whether respondents were from the city or country, and scores on the PROCAN law reform variable. Respondents in the 17

to 19 year-old group were 10.3 times more likely, and those in the 50 to 59 year old group were 3.3 times more likely, to be in favor of making cannabis legal than those in the over 60 year-old group. Those who voted for a left wing party were

1.6 times more likely than those who voted for a right wing party to be in favor of making cannabis legal, and those from the city were 1.6 times more likely than those from the country to support the same position. Each I point increase in score on the PROCAN variable resulted in a 0.7-fold increased likelihood of support for legalization. PROCAN was the strongest predictor of attitudes to legalization, the full odds ratio for this variable being 12.9. Attitudes toward making cannabis as legal as alcohol were not affected by the other variables entered into the regression.

When decriminalization was not explained, scores were also predicted by age, political affiliation, whether respondents were from the city or country, and scores on the PROCAN law reform variable. These results are presented in table 7. Respondents in the 17 to 19 year-old age group were 7.3 times more likely, those in the 30 to 39 year-old group were 2.2 times more likely, and those in the 50 to 59 year-old group were 2.5 times more likely to be in favor of removing criminal penalties applying to cannabis, than those in the over 60 year-old group. Those who voted for a left wing party were 1.4 times more likely than those who voted for a right wing party to be in favor of changing the laws that apply to cannabis, and those from the city were 1.5 times more likely than those from the country to support the same position. Each one point increase in score on the PROCAN variable resulted in a 0.7-fold increased likelihood of support for removing criminal penalties. PROCAN was the strongest predictor of attitudes to decriminalization when the penalties were not described, the full odds ratio for this van* able in this case were also 12.9. Attitudes toward cannabis decriminalization when the term was not explained were unaffected by the other variables entered into the analysis.

Table 8 shows that when decriminalization was explained, scores on this variable were predicted by only PROCAN and gender. That is, women were 1.7 times more likely to be in favor of cannabis decriminalization than men. Each I point increase in score on the PROCAN variable resulted in a 0.2-fold increased likelihood of support for removing criminal penalties. PROCAN was the strongest predictor of attitudes to decriminalization when the penalties were described, the full odds ratio for this variable was 4.4. Attitudes toward cannabis decriminalization when the term was explained were not affected by the other variables entered into the analysis.

To examine which variables predicted favorable attitudes toward cannabis law reform as measured by PROCAN this variable was dichotomised and included as the dependent variable in another logistic regression. Scores under 0 were given a value of 0 and those above were given a value of 1. Table 9 shows that age and religious affiliation predicted whether people were favorable to cannabis law reform as measured by PROCAN. Those in the 20 to 29 year-old group were 7.5 times more likely, and those in the 50 to 59 year-old group were 2.5 times more likely to be in favor of law reform, than those in the over 60 year old group. Those who described their religion as none or unimportant were 1.5 times more likely than those who described their religion as important or very important to be in favor of law reform as measured by PROCAN.