Ontario Premier Kathleen Wynne is the latest leader to advocate distributing marijuana through provincial liquor control boards, should Ottawa legalize the drug. It is not an entirely stupid idea. Yet it will encounter three intractable problems.

First, there is a fundamental conflict between harm prevention and revenue maximization. The LCBO may talk about the millions of times each year underage would-be drinkers, or the obviously inebriated, are denied service. Yet the ugly truth is that the alcoholic who stops by his local Liquor Control Board of Ontario to buy a bottle of whisky every day or two is more likely to be greeted with a friendly “Good day, sir” than a health warning.

Heavy consumers of alcohol are the mainstay of the LCBO’s revenue base; the agency has little incentive to persuade them to stop drinking. As my Carleton colleague Jean Daudelin, who has studied drug markets in Canada and elsewhere, argues, “any kind of business model which is geared mainly to generating revenue creates incentives that are not compatible with any rational public health policy.” There is no reason to believe that liquor control boards would curtail heavy cannabis use any more (or less) effectively than they do alcoholism.

Second, it is not obvious how recreational marijuana sales will co-exist with the already existing market for medical marijuana. Canada’s tax system divides (legal) drugs into two distinct categories. There are medicines, which are taxed lightly, if at all, and there are recreational drugs, such as alcohol and tobacco, which are subject to heavy excise or “sin” taxes. How many serious users will want to buy heavily taxed recreational marijuana from the local “Liquor and Cannabis Control Board,” when they could use medical marijuana instead and claim it as a medical expense on their income-tax return?

The third stubborn problem the legal pot industry must face is competing with home and black-market production. Marijuana producers and distributors who wish to participate in the legal market will have to pay a whole range of taxes, from HST and excise taxes to the income, payroll and other taxes every Canadian business is expected to pay. Small-scale home production and tax-evading black-market production faces none of these costs.

To compete with untaxed home- and black-market operations, legal producers must be able to grow marijuana so efficiently that they cover the cost of taxation and legal distribution, yet are still competitive with home- and black-market production in terms of price and quality. The markets for alcohol and tobacco show how this is done: globalization plus large-scale capital intensive production. Half of the tobacco cigarettes sold (legally) in Canada are produced by just one company, Imperial Tobacco Canada Ltd., and manufactured in Mexico. The LCBO actively attempts to promote Ontario wines and beer, but makes its money elsewhere. Ontario wines account for only 7 per cent of the LCBO’s sales, and Ontario craft beers account for less than 1 per cent. If cannabis were legal, what reason is there to think that the market for marijuana would not show the same tendencies toward globalized, capital-intensive production that tobacco, alcohol and so many other products exhibit right now?

There are alternatives. There are models out there that might – just might – allow the Canadian cannabis industry to flourish, mitigate harm to users and, at the same time, generate tax revenue.

One option is to stick with the current model of medicinal use only, forgo the potential tax revenue that could be raised from recreational marijuana and focus only on harm reduction. This model assumes that doctors are effective gatekeepers and marijuana is sufficiently harmful that recreational use should be banned. Both assumptions are dubious.

The Colorado model is another alternative worth studying. Marijuana sellers in Colorado operate under a similar regime to Ontario craft breweries: They sell only (or mostly) cannabis that they produce themselves. Retail licences are available only to marijuana cultivators and there are limits to the number of plants each producer can cultivate (the full set of rules runs to209 frequently impenetrable pages).

The Colorado model appears to have achieved two key policy objectives – protecting existing producers and raising some amount of tax revenue. But it is not without problems. As cannabis is still illegal in the United States, entrepreneurs are reluctant to invest capital into marijuana production. Yet without capital investment, legal producers are left using the same growing and harvesting technologies as illegal ones. With a similar cost structure, and higher regulatory and tax burdens, it is a challenge for legal producers to compete.

A final option would be to use marijuana legalization as an opportunity to rethink existing liquor distribution structures, especially in Ontario. It is sometimes not understood (in Ontario, anyway) that it is possible for a province to retain a monopoly on liquor (or cannabis) distribution at the wholesale level, and at the same time have competition at the retail level. Alberta and British Columbia show how it can be done. If Ontario were to follow the example of these provinces, all marijuana sales would have to go through a Cannabis Control Board of Ontario. But instead of the board having a monopoly at the retail level, it could contract out sales of marijuana to independent sales agents.

Even if Ontario decides to have some element of government monopoly over cannabis distribution, there is no reason why that government monopoly needs to extend to the retail level. The experience of other provinces suggests that allowing more competition at the retail level could lead to lower costs and promote retail service that is more responsive to customer needs.

Follow on Twitter: @franceswoolley