Several economic myths have surrounded the legalization of marijuana. This has maintained the illusion that it would be a bonanza for federal and provincial treasuries because of the supposed enormous tax revenues that legalization would generate.
The Parliamentary Budget Office in Ottawa is to be congratulated for blowing up some of these myths in its report, published this month, on projected marijuana tax revenues following legalization in Canada in 2017 or 2018.
The first myth is that legalization would “create” a market value of over $20 billion. This number was offered in a recent Deloitte Canada report, taking into account investment, marijuana tourism, induced economic activity and so forth.
Well, in consumption terms, relatively little will be “created.” An illegal market may largely transform itself into a legal market, with only a small increment in total consumption. Legalization will bring about a displacement of production from private smaller-scale growers to licensed industrial-scale growers that operate in heavily regulated environments.
This may turn out to be market enhancing because of labelling requirements, and a greater degree of certainty regarding pharmaceutical content. It will also lead to the employment of PhD graduates in chemistry, students with MBAs, lawyers and marketing specialists. Some security jobs will migrate from enforcers to Garda employees. But if the legal market is to supplant the illegal market, it will at the same time create unemployment among the existing growers and distributors.
Indeed, the rents that will accrue will be spread differently — away from small operators and organized crime and toward individuals whose retirement portfolios hold MJ stocks.
The published report from the PBO finds that total tax revenues from legalized marijuana would actually come to only about $600 million, with large confidence boundaries around this central number, meaning that it’s just a ballpark estimate.
In contrast, some private-sector commentators have been projecting tax revenues in the billions. The key element in the PBO analysis is its recognition that buyers are price sensitive — a big price premium for the legal product could lead users to stick with their pre-existing dealer, whom they may have known for years.
This modelling assumption is based upon a careful scrutiny of existing economic research in the area. The Canadian illegal market is already very well developed: Product quality is high and it is readily available. The PBO simulates projected revenues on the basis of many different price differentials. It concludes that if the legal market price is only a dollar or two higher than the illegal price, most buyers will migrate to the legal market. But a larger price differential would lead to a low legal market share and therefore lower tax revenues.
Experience from two U.S. states highlights the importance of price. In Washington State, the initial tax was almost 100 per cent. As a result, shop prices were $28 per gram, compared to $9 on the street. Unsurprisingly, only 18 per cent of sales went through licensed vendors. In contrast, Colorado applied a lower rate— about 35 per cent. The shop price was $15 per gram, and about 65 per cent of the estimated market migrated to licensed sellers. Shop prices are slowly declining further, and the share of black market activity is now minimal. Almost all illicit cultivation is bound for out-of-state markets, where the street price is higher.
Notwithstanding the many challenges facing the legalization process, the PBO has gotten this process off to an excellent start, with its sober assessment, professional statistical modelling and avoidance of exuberance. The government should use this assessment to craft a law that takes its conclusions into account and avoid the pitfalls experienced in Washington State, and not fall for the illusion that billions of dollars will suddenly fall in its lap.
Ian Irvine est professeur d'économie à l'Université Concordia et chercheur associé à l'Institut économique de Montréal. Il signe ce texte à titre personnel.