394 ECONOMIC RECORD DECEMBER 394 THE ECONOMIC RECORD, VOL. 80, NO. 251, DECEMBER, 2004, 394–410 © 2004. The Economic Society of Australia. ISSN 0013–0249 Demand for Marijuana, Alcohol and Tobacco: Participation, Levels of Consumption and Cross-equation Correlations* XUEYAN ZHAO and MARK N. HARRIS Department of Econometrics and Business Statistics, Monash University, Melbourne, Victoria, Australia We investigated marijuana, alcohol and tobacco consumption using micro-unit data from the Australian National Drug Strategy Household Surveys. We estimated a multivariate probit (MVP) model to allow for correlations across participations of different drugs and a sequential model to study separately the determinants of participation and the levels of consumption. The MVP results indicate significant and posi- tive correlations across all three drugs through unobservable charac- teristics, with the correlation coefficient between marijuana and tobacco being the highest. The MVP approach allows for better prediction of conditional and joint probabilities, providing valuable information for policy makers in a multidrug framework. Issues relating to marijuana consumption are, to some extent, different from those of legal drugs. Although still an illicit drug in Australia, the con- sumption of marijuana among a significant pro- portion of Australians has persevered and the real price of marijuana has declined by almost 40 per cent over the past decade (Clements, 2004). The trend towards hydroponic cultivation has significantly increased productivity and product quality. Recent surveys indicate that some 34 per cent of the popu- lation in Australia aged 14 and over has used marijuana and 14 per cent have reported regular use of at least once a year (see Section II). According to an estimate by Clements and Daryal (2003), the expenditure on marijuana in Australia is almost twice of that on wine. During 2000/01, Australian law enforcement bodies seized almost 10 000 kilograms of marijuana, nearly twice that of the previous year (ABCI, 2002). Separately, there is an exhaustive literature on the consumption of tobacco, alcohol and, to a lesser extent, marijuana. However, few empirical studies acknowledge the potential intrinsic cross-commodity correlation across these three drugs. According to recent survey data in Australia (see Table 1), although I Introduction and Background The consumption of marijuana, alcohol and tobacco has long been identified as being associated with a range of adverse social, economic and health effects (see, for example, Cleeland Report, 1989; Model, 1993). According to a recent estimate (Collins and Lapsley, 2002), the total costs incurred from the use of alcohol, tobacco and illicit drugs in Australia was over $43 billion in the financial year 1998/99. A significant amount of public funds has been spent by governments to deal with the consequences and on educational programmes. Much debate has sur- rounded the use of governments’ coercive power in influencing drug consumption through public policies such as taxation and legislation (see, for example, Dave and Kaestner, 2002; Glied, 2002). * We thank Ben Machado and Jennine Boughton of ABCI and Steve Whennan from ABS for supplying the data, Ken Clements for helpful discussion and Preety Ramful for research assistance. Financial support from the ARC (large grant no. A00105073) is also kindly acknowledged. Correspondence: Xueyan Zhao, Department of Econometrics and Business Statistics, Monash University, Victoria 3800, Australia. Email: Xueyan.Zhao@BusEco.monash.edu.au