What would legal cannabis really earn Victoria?

Victoria's Parliamentary Budget Office (PBO) says a legal, state-run cannabis market would leave the budget $1.37 billion better off over ten years. That is a large headline figure. But the figure counts only part of the ledger, and some of what it leaves out is large.

The costing, published on 21 October, was requested by Libertarian Party MP David Limbrick ahead of the 28 November state election. It finds the policy would improve the state's net position by $113.4 million over the four years to 2029–30, and by $1,367.8 million over the costing period to 2036–37 (ten years of operation, as nothing happens in 2026–27). By 2036–37 the annual gain reaches $246.4 million.

The PBO itself calls its estimates "subject to considerable uncertainty". Read closely, the document also shows how much it was never asked, or never chose, to count.

What the policy is, and how the PBO modelled it

The policy would legalise the production and sale of cannabis to adults from 1 July 2027. It has three parts: a regulator that licenses producers and retailers, funded entirely by licence fees; a government-owned company with a statutory monopoly on wholesale; and a right for adults to grow a limited amount at home. Sales to under-18s would stay illegal, with penalties modelled on alcohol law.

The wholesaler is where the money is. Victoria cannot levy its own cannabis excise, because the Constitution gives excise to the Commonwealth alone. So the policy captures value the way Ontario does: the state buys from growers, sells to shops, and keeps the margin. The PBO assumed the wholesaler would behave like a profit-maximising monopoly, pricing to stay competitive with the black market while squeezing producer margins over time, and paying dividends at rates similar to other state-owned entities.

The regulator was assumed to cost exactly what licence fees bring in, so it nets to zero. Licence numbers were scaled to Victoria's packaged liquor licences, with fees similar to alcohol licences. The state would also put $9.4 million into the wholesaler's fit-out and equipment, and pay its first year of rent.

Legal sales would not start until July 2028, a year after commencement, to allow crops to be grown and supply chains built. That is why 2027–28 shows a $10.9 million loss before the dividends arrive.

Over the full period, revenue totals $1,430.2 million against $53.0 million in operating expenses and the $9.4 million capital outlay.

What the costing leaves out

The $1.37 billion is, in effect, the wholesaler's dividends minus set-up costs. Several money flows that legalisation would change are not in it.

Policing, courts and prisons. The method lists eight steps; none touches law enforcement. Yet cannabis policing is not small: in 2022 the PBO itself estimated the Victorian Government spends over $70 million a year on cannabis-related law enforcement. Victoria Police recorded 8,078 alleged offenders for cannabis possession in 2022, and 11,493 in 2020, according to Crime Statistics Agency figures given to Parliament (Questions on Notice).

The same office has counted these effects before. Its 2022 costing of the Reason Party's legalisation policy included about $60 million in policing and prosecution savings (ABC report, via Fiona Patten). Its 2022 costing of Mr Limbrick's own "End the nanny state" policy accounted for lost cannabis fine revenue and enforcement savings (PBO). This costing does neither, and does not say why. The likely answer is that this policy did not ask for savings to be banked or redirected, and freed-up police time is usually redeployed rather than returned to Treasury. Readers should still know the figure excludes it.

What happened in Canada. Cannabis offences fell by about two-thirds as soon as legal sales began. Statistics Canada recorded 136 offences per 100,000 people in 2017 and 44 in 2019 (Statistics Canada). Statistics Canada notes part of that drop reflects prosecutors deprioritising simple possession, not only the law change. Legalisation also brought new enforcement spending. Ottawa set aside up to C$274 million over five years for enforcement tied to legalisation, including up to C$161 million for drug-impaired driving (Public Safety Canada). City forces asked for more too: Toronto police sought C$1.5 million (Stratcann) and Regina police estimated over C$1 million a year (CBC). We found no evidence of police budgets being cut as a result.

Legal does not mean unpoliced. This policy still leaves plenty of ways to break the law: growing more than the "prescribed limit", holding more than any possession cap, selling without a licence, supplying under-18s. Each still needs police, courts and prosecutors. The tighter the limits, the more people stay on the wrong side of them, and the policy has not said what the home-grow limit is. A low cap would keep many of today's home growers criminal. The ACT, which legalised small-scale possession and home growing in 2020, shows both sides. A 2024 government review found cannabis charges fell from 200 a year to fewer than 20, but police still described the clash with Commonwealth law as an ongoing problem, and one man who sold cannabis to children argued successfully that he held less than the possession limit (Canberra Times). Six years after legalisation, Canadian police still recorded 8,879 cannabis offences in 2024, and the offence rate rose 30% on the year before.

A legal state surrounded by illegal ones. Cannabis bought legally in Melbourne becomes illegal the moment it crosses into New South Wales or South Australia, where larger amounts can be treated as trafficking, and it stays illegal under Commonwealth law throughout. Interstate cannabis runs already happen: in June 2026, NSW highway patrol found more than 50 kg in a Victorian-registered car on the Sturt Highway at Hay (Region Riverina). A legal Victorian market would turn much of that traffic around, with Victorian product heading north and west, and bring pressure on Victoria to help police it. The US shows how this plays out: in 2014 Nebraska and Oklahoma sued Colorado, arguing its legal cannabis was spilling across their borders and straining their police. The Supreme Court declined to hear the case in 2016 (NPR). The interstate buyers who would lift the wholesaler's revenue are the same people creating that enforcement work.

Drug-driving could rise, not fall. Victoria's roadside tests detect the presence of THC, not impairment. More legal users would mean more drivers testing positive after any impairment has passed, unless that law changes. The policy does not address it, and the costing does not count it.

So while legalisation would cut arrests for simple possession, the realistic budget saving is small. That is why our calculator's central case banks only a quarter of the PBO's $70 million.

Tax-equivalent payments. The PBO counts dividends from the wholesaler's post-tax profits. Victorian government businesses generally also pay income tax equivalents to the state. The costing does not say whether those payments are included. If they are not, state revenue is understated.

Taxes on a new industry. Licensed growers, processors and shops would pay payroll tax, land tax and stamp duty like any other business. None of this is counted. Nor is Victoria's share of GST on legal sales, which flows through the Commonwealth's distribution formula. The federal PBO's 2025 costing of a national scheme, by contrast, counted GST and a 15% excise (PBO), though excise is a lever Victoria does not have.

Interstate buyers. Victoria would be the first state with legal retail sales. The PBO flags cross-border purchasing and interstate visitors as a sensitivity but says the scale is hard to estimate. Its method mentions overseas visitors only, so interstate demand appears to be excluded. That is more likely to push revenue up than down.

Where the assumptions are thin

The key numbers are not published. The result rests on a handful of inputs: price per gram, the legal market's share, total consumption, the wholesaler's markup and its dividend payout. The costing names these as the main sensitivities but discloses none of them. The federal PBO's 2023 national costing did publish its core inputs, including a $13.40 pre-tax price per gram and a legal share of 52% in year one (Honi Soit). Without them, nobody outside the PBO can test the Victorian figure.

One number, no range. The PBO says its estimate is subject to "considerable uncertainty", then offers a single figure. There is no low or high scenario, even for the three sensitivities it lists.

Survey data with known gaps. Consumption appears to draw on the National Drug Strategy Household Survey, which relies on people self-reporting illegal drug use. Street prices appear to draw partly on the Ecstasy and Related Drugs Reporting System, which surveys people who regularly use ecstasy and other stimulants. Neither is a cannabis market study. The PBO also cites Canada's 2024 cannabis survey, taken six years into a legal market, when people report use more readily than under prohibition. It does not say how that survey feeds into its Victorian estimates.

An Ontario template. Thirteen of the cited sources are Ontario Cannabis Store reports, plus three more from other Ontario agencies. Ontario operates where cannabis is legal nationally, with federal excise and a mature licensed industry. The PBO concedes that differences in pricing, tax and market structure limit how far overseas outcomes carry over.

A regulator sized to its fees. The regulator's budget is assumed to equal licence revenue, with fees set like alcohol licences. That makes it budget-neutral by construction. If testing, compliance and action against unlicensed sellers cost more than liquor-style fees raise, either fees rise or the net gain shrinks.

A home-grow limit nobody has named. The policy lets adults grow cannabis at home only "up to a prescribed limit". Neither the policy summary nor the costing says what that limit is, how many plants it allows, or whether there is a cap on how much people may possess. The PBO says it accounted for self-grown consumption, so it must have assumed some limit, but it does not disclose it. That matters twice over. For growers, the limit could be as tight as the ACT's, where adults may possess no more than 50 grams of dried cannabis and grow two plants each, capped at four per household. For the budget, the limit cuts the other way: the tighter it is, the more people must buy from the state's wholesaler, and the bigger the dividends. The costing shows the revenue without the rule that helps produce it.

Squeezing growers. The wholesaler is assumed to raise its selling price and cut what it pays producers over time, while staying cheaper than the black market. With bulk Australian medicinal flower already trading at a few dollars a gram (see the price section below), the costing does not test whether growers could supply at the prices assumed.

Medicinal patients. Australia already has a large prescribed cannabis market. The costing does not say how many patients it expects to move to the recreational channel. Canada's experience, set out below, suggests it could be around half.

An unreconciled capital figure. The costing reports a $9.4 million increase in assets but a total estimated investment of $6.3 million. It does not explain the $3.1 million gap.

The price problem the costing skips

In the legal markets we looked at, prices fell hard once supply caught up. The PBO's revenue rests on a monopoly wholesaler holding a wide margin between what it pays growers and what it charges shops. It publishes no price path and shows no scenario in which prices fall.

*Ontario legal retail sales divided by grams sold, all product types, so not strictly like-for-like with the 2019 survey. Sources: Statistics Canada, OCS By the Numbers 2025, Colorado Department of Revenue via Westword, Cannabis Business Times, Stratcann, PBO.

The black-market price is a moving target. The PBO assumes the wholesaler prices to stay competitive with the illegal market. But illegal prices fall too once legal supply arrives. In Canada's first year, Statistics Canada recorded the average illicit price dropping from C$6.51 to C$5.59 a gram while legal stores were still charging over C$10.

Ontario's monopoly had to cut its own margin. The PBO assumes Victoria's wholesaler would use its market power to push up the price it charges shops. The OCS, the PBO's main template, went the other way. In September 2023 it moved to a fixed, lower markup to help legal stores compete with illegal sellers. Its gross margin fell from 22.8% in 2021–22 to 19.6% in 2024–25, and its profit has sat at about C$234–246 million for three years despite rising volumes (OCS 2024–25 Annual Report).

Australia already has cheap legal cannabis. Medicinal cannabis has been legal federally since 2016, and prices have kept falling. Bulk medicinal flower traded at about $1.20–4.80 a gram in late 2025, with prices weakening through the year as imports competed with local supply, according to Global Cannabis Exchange data reported by Stratcann. Even after processing, packaging and dispensing, that cost base sits well below the $12–15 street prices the PBO itself used. A Victorian recreational wholesaler would have to price against cheap prescriptions as well as dealers, and the costing never discusses that.

The PBO cited the evidence but does not show how it used it. The costing's own source list includes an industry report on record-low Australian flower prices and Stratcann's analysis of global price compression. Yet the document lists price only as a sensitivity, with no scenario showing what happens if prices fall.

Lower prices are not all bad for the budget. They pull buyers from the black market and lift volumes, as Ontario's grams sold kept rising. But Ontario shows the end result: a profit that stops growing. The PBO's Victorian profit, by contrast, climbs every year to 2036–37.

What it means for medical cannabis companies

Canada's experience suggests adult-use legalisation roughly halves the legal medical market. The costing does not discuss what happens to the medical market.

Canada's medical market halved. Active registrations with licensed medical sellers fell from 345,520 in October 2018, when adult-use sales began, to 161,267 in March 2025. The fall was not instant: registrations held up for about two years, then slid steadily as legal stores spread. Monthly shipments to patients fell from 143,752 to 71,411 over the same period. In Ontario, registrations dropped from 160,374 to 82,483 (Health Canada).

Researchers link the fall to legalisation. A peer-reviewed study compared medical shipments with where pre-legalisation trends pointed. By early 2020, shipments were 48% lower nationally and 74% lower in Alberta (Nguyen et al., Cannabis and Cannabinoid Research, 2023). The authors note Canada also applied its new excise tax to medical cannabis, which may have pushed some patients away as well.

Australia's medical market has more to lose. It has grown quickly since 2016: in 2024 alone, the TGA approved at least 979,000 prescription applications for medicinal cannabis, according to an analysis in MJA InSight+. A Victorian who can walk into a licensed shop no longer needs a script or a consultation. If Victorian patients behave as Canadians did, a large share of prescribed sales in the state would move across, hitting producers, telehealth clinics and pharmacies.

Victorian growers would face one buyer. Under the policy, every gram sold in the recreational market must pass through the state wholesaler, which the PBO assumes will push down what it pays producers. Even if federal law allowed it, a medical company losing patients could only follow them into the recreational channel on the wholesaler's terms. The PBO lists an investor presentation from Melbourne-based medical producer Cann Group among its sources, but the costing says nothing about what happens to the existing industry.

Not every patient would leave. Canada still has over 160,000 medical registrations, for reasons that likely include clinical oversight, insurance cover and products shops don't stock. And patients who switch add to the wholesaler's sales, which may already be in the PBO's volumes. The costing doesn't say either way.

Who would grow it?

The PBO assumes a supply chain is up and running 12 months after the law starts. Australia's licensed cannabis industry is small, relies on imports, and under current Commonwealth law cannot supply a recreational market.

The medical market runs on imports. The TGA counts more than 1,000 medicinal cannabis products being supplied in Australia, and only two of them are approved medicines (TGA consultation paper, August 2025). In 2024 Australia imported more than 77,000 kg of medicinal cannabis, about 80% of it from Canada, while licensed local growers produced 41,328 kg, according to Office of Drug Control figures reported by Stratcann. Roughly two-thirds of supply came from overseas, and cheap Canadian product is a large part of why prices have fallen so far.

Supply depends on what Canberra does. As the law stands, cannabis can be imported only for medical or scientific purposes under Office of Drug Control permits, and Commonwealth cultivation licences cover medicinal and research production. On those settings, a Victorian recreational market could not legally be supplied from imports or from today's licensed medical growers. That could change. The Commonwealth could amend its drug laws, create a licence for adult-use supply, or choose not to act against a state scheme, and a federal bill to legalise cannabis was introduced in 2023 (Australian Greens). Each path would change who could supply the market, how quickly and at what price. The costing does not say which it assumes.

The volumes would not appear in a year. The PBO's 2023 estimate of Victorian consumption is 84,700 kg a year. That is more than double what every licensed grower in Australia produced in 2024 for medical and research use. Even our central scenario has legal shops selling about 34,000 kg in the first year of sales.

There is a case for local cultivation, on the wholesaler's terms. If imports stayed barred, a recreational market would be a captive market for Victorian growers, a real opening for a local industry losing ground to Canadian imports. But those growers would sell to one state buyer directed to maximise its profit by limiting their margins, while, unless federal law changed, operating outside Commonwealth law. Investors and lenders would weigh both. The costing addresses none of this.

Testing, and who pays for it

Every legal market requires cannabis to be tested for potency and contaminants before sale. The policy summary and the costing do not mention testing.

It is a real cost to growers. University of California, Davis researchers estimated that mandatory testing cost California growers about 10% of the average wholesale price of legal cannabis. Much of that came from destroying batches that failed; the failure rate was 5.6% in 2018 and 4% in 2019. California tests for more than 100 contaminants, including 66 pesticides (Valdes-Donoso, Sumner and Goldstein, PLOS ONE, 2020, via ScienceDaily). Under a wholesaler that the PBO assumes will squeeze producer margins, a cost of that size comes out of growers' pockets or the wholesaler's margin.

Private labs need policing. Where producers choose and pay their own labs, results can drift. California brought in a standard potency method in January 2024 to tackle inflated THC results, and at first only 12 of its 37 licensed labs could test flower (MJBizDaily). In Colorado, a state audit released on 19 October 2026 found most products sold from 2023 to 2025 were not consistently tested for contaminants or potency, and that the state lacks a fully functional safety reference lab (Axios).

Australia has a standard, but only for medicine. Medicinal cannabis must meet the TGA's TGO 93 standard. When the TGA tested 22 widely supplied products in 2020, three fell short of their stated cannabinoid content (TGA). A recreational scheme would need its own standard, accredited labs and audit testing.

The costing does not price it. The regulator is assumed to cost exactly what licence fees raise, about $5–8 million a year from 2030–31. Overseeing labs, running random audits and maintaining a reference lab would sit on top of licensing work. For scale, a cannabis reference lab proposed for Colorado was estimated at about US$413,000 to set up and US$800,000 a year to run (MJBizDaily). That is modest, but it is not zero, and the costing does not show it.

Drinks, and what it means for alcohol

Cannabis drinks are a small part of legal markets so far, but legal cannabis does appear to take some sales from beer. The costing considers neither.

Cannabis drinks are a niche. In Ontario, cannabis beverages made up about 1.6% of legal retail sales in 2025, roughly C$37 million of C$2.28 billion (OCS By the Numbers 2025). At the wholesaler, beverage revenue fell from C$38.8 million in 2023–24 to C$33.0 million in 2024–25 (OCS Annual Report 2024–25). Canada caps edible and drinkable products at 10 mg of THC per package, which limits how far drinks can compete with a beer or a glass of wine.

What shops can sell matters more than drinks. Dried flower was only 39% of the Ontario wholesaler's revenue in 2024–25. Vapes, pre-rolls, infused pre-rolls and edibles made up most of the rest. The Victorian policy summary refers only to "cannabis", and the costing does not say which products it assumes can be sold. A flower-only market would earn well below an Ontario-style one.

Legal cannabis takes some sales from beer. A 2024 study in Drug and Alcohol Dependence found Canadian legalisation was associated with a modest but economically significant fall in beer sales in most provinces, and no change in spirits (Mital et al., via Cannabis Science and Technology). Canada's alcohol sales volume fell 3.8% in 2023–24, the biggest drop on record, with beer down for the eighth year in a row (Statistics Canada, via Retail Insider). That decline began before legalisation, so cannabis is at most one of several causes.

In Australia, the effect would land mostly on Canberra and the alcohol industry. Alcohol excise goes to the Commonwealth, so any fall in drinking would cut federal revenue rather than Victoria's. Victorian brewers, wineries and venues would feel any shift in spending. Less drinking could also mean less alcohol-related harm. The costing counts none of this, in either direction.

Checking the numbers against Ontario

The PBO leans on the Ontario Cannabis Store (OCS), so its real accounts are the obvious test. They show the Victorian figure is reachable only on generous assumptions, and much faster than Ontario managed.

Sources: OCS 2024–25 Annual Report, OCS 2019–20 Annual Report, ABS, Ontario Ministry of Finance.

Per head, Victoria is assumed to earn more than twice what Ontario's wholesaler does. By 2036–37 the PBO has Victoria's wholesaler matching the OCS's profit almost dollar for dollar, with less than half the population. That gap narrows only slightly if you allow for the Canadian dollar usually being worth a little more than the Australian.

There is a fair answer to that, and it has a cost. Ontario's government also collects its share of federal cannabis excise, about C$346 million in 2023–24 (Stratcann). Add that to the OCS's profit and Ontario takes roughly C$37 per resident. Victoria cannot levy excise, so its wholesaler would have to take an excise-sized markup on top of a normal wholesale margin. That only works if legal prices still beat the black market. It is the same tension the PBO identifies but does not test.

The market has to be big, and mostly legal. The OCS's profit is about 11% of Ontario's legal retail sales. At that margin, Victoria would need about $2.2 billion a year in legal sales to deliver the PBO's 2036–37 figure. Even at Ontario's combined government take of about 28% of sales, it would need around $880 million. The PBO's own published advice put the value of Victoria's illicit cannabis market at $1.02–1.27 billion in 2022, based on consumption of 84,700 kilograms (PBO, Cannabis consumption in Victoria). So on the generous reading, the policy needs the legal market to win most of today's spending, in a decade. Ontario has reached 72% so far.

Ontario took five years to pay its first dividend. The OCS lost money in its first year, borrowed from the province to start up, and sat on a growing cash pile until its first dividend in September 2023. It has since paid C$943 million, including a C$355 million one-off in June 2025. The PBO's Victorian wholesaler pays out from its first year of sales on a $9.4 million investment.

These are rough checks, not a rival costing: Victoria's figures are a decade out and in nominal dollars, and some of the PBO's inputs are unknown. But they show how much rests on assumptions the costing does not publish.

Jobs, shops, licences and tax: Ontario scaled to Victoria. The costing says nothing about the industry a legal market would create. Ontario gives a guide, but Victoria cannot simply be scaled by population, because reported use differs. In 2022–23, 11.5% of Australians aged 14 and over said they had used cannabis in the past year (AIHW). Comparing that with Canada today would be unfair: people under prohibition are likely to under-report, and Canada's figure rose from 22% in 2018, just before legalisation, to 26% in 2024 (Health Canada). So the table adjusts using Canada's 2018 figure, when both countries were reporting under prohibition. The two surveys still use different methods, so the true Victorian figures probably sit between the last two columns.

*Jobs are a Canada-wide figure scaled to Victoria's population, because the Ontario figure is reported inconsistently. Sources: OCS By the Numbers 2025, OCS and Deloitte report via Stratcann.

The adjusted retail figure, about C$530 million, is in line with our central case of $520–560 million a year, which suggests that case is not unduly pessimistic.

Most of the tax would go to Canberra. Ontario's C$5.3 billion includes income tax on wages, company tax on producers and retailers, and sales tax. In Australia those go mainly to the Commonwealth. GST at 10% on our central case's legal sales would raise about $40–55 million a year, but it goes into the national pool and comes back to Victoria through the Commonwealth Grants Commission formula, not dollar for dollar. Income and company tax from roughly 9,000 jobs and several hundred businesses would go to the Commonwealth outright. What Victoria would collect directly is payroll tax, land tax and stamp duty, though many small shops would sit under the payroll tax threshold.

Licence fees are modest everywhere. Ontario charges C$6,000 for a retail operator licence and C$4,000 for each store, with renewals of C$2,000 and C$3,500 every two years (AGCO). The PBO sets Victoria's fees at liquor-licence levels and assumes they exactly cover the regulator. The PBO's own figures put the regulator's cost, and so licence revenue, at $5–8 million a year from 2030–31: enough for a lean regulator, but not obviously for compliance work against unlicensed sellers.

Running the numbers differently

Swap the PBO's unstated assumptions for ones grounded in what has happened elsewhere, and the policy is worth about $545 million over the decade to 2036–37, not $1.37 billion. The annual gain levels off at around $75 million a year from the early 2030s instead of climbing to $246 million.

We built a simple model starting from the PBO's own 2022 estimate of Victorian consumption, 84,700 kg a year, and ran it three ways.

*At Canada's ratio of about 17 jobs supported per $1 million of legal retail sales, including indirect jobs (OCS and Deloitte report via Stratcann, OCS Annual Report 2024–25).

Our totals include enforcement savings, industry taxes and unrecovered running costs, which the PBO leaves out. Counting only what the PBO counts (wholesaler dividends and set-up costs), the 11-year figure is about $54 million pessimistic, $391 million central and $1,408 million optimistic. In other words, the PBO's number holds only if nearly every input lands at the optimistic end.

Where our assumptions differ. The PBO does not publish its inputs, so this compares ours with what it states or implies. Figures are our central case, with pessimistic and optimistic in brackets.

What drives the gap. Mostly price and margin. In our central case, legal retail sales flatten at around $520–560 million a year because the price per gram keeps falling, even as volumes grow. At a 16% profit margin and a 65% state take, that supports about $55–60 million a year in returns. To reach the PBO's $246 million on a market that size, the wholesaler would need a margin of more than 60% of retail, which is hard to square with beating the black market on price.

This is a rough model, not a costing, and our assumptions are judgement calls. That is exactly the criticism we make of the PBO, which is why ours are set out in full.

What it also doesn't count on the cost side

The omissions do not all point the same way. A fair reading notes these too.

• Health. There is no estimate of changes in demand for drug treatment, mental health or emergency services, in either direction.
• Prevention and education. The policy funds no public health campaign, and the costing includes none. Canada and many US states fund public education from cannabis revenue.
• Roads. Drug-driving enforcement and any change to roadside testing are not costed. Canada set aside up to C$161 million over five years for drug-impaired driving alone.
• The remaining black market. Policing unlicensed growers and sellers would continue, and that cost is not estimated either.
• Federal law. Cultivation and supply remain offences under Commonwealth law. The PBO notes that the ACT legalised use and home growing but not sale. The costing assumes the scheme operates as designed and does not consider the risk of a legal challenge or Commonwealth intervention.

The PBO's job is to cost the policy as described, not to rule on those questions. But each of them could change the final number.

The bottom line

The PBO's $1.37 billion is a single, unexplained estimate of what a state cannabis wholesaler could pay into the budget. Rebuilt with assumptions drawn from markets that have legalised, it looks closer to $545 million over the decade. The result swings from a small loss to about $1.9 billion depending mainly on price, margin and market share, and the PBO publishes none of those three inputs.

What the evidence points to. Each of these can be checked against the sources linked in this article.

• Legal prices fall. Ontario's legal price per gram roughly halved in six years, on our comparison of 2019 survey prices with 2025 sales data. Colorado's wholesale flower price fell 62% between 2021 and 2023.
• State wholesalers earn modest margins. Ontario's made about 11% of legal retail sales in 2024–25, cut its markup in 2023 to help legal shops compete, and paid its first dividend five years after legalisation.
• Medical demand shrinks. Canada's medical cannabis registrations fell 53% between October 2018 and March 2025.
• Supply would have to be built. Licensed Australian growers produced 41,328 kg in 2024, most of the medical market is imported, and under current Commonwealth law imports for recreational use are barred. The PBO's 2023 estimate of Victorian consumption was 84,700 kg a year.
• Arrests fall, enforcement spending doesn't disappear. Canadian cannabis offences fell by about two-thirds, while governments added money for drug-driving and illegal-market enforcement.

Where the PBO may have undercounted. Tax-equivalent payments from the wholesaler, payroll and land tax, and interstate buyers could all add to Victoria's take. A legal market the size of our central case would support roughly 9,500 jobs, though most of the income, company and GST revenue would go to Canberra.

What would settle it. The PBO and the Libertarian Party could publish the price, market share, margin and volume inputs behind the figure; show a falling-price scenario and a low-to-high range; state the home-grow and possession limits assumed; and explain who would grow the cannabis, under Commonwealth law as it stands or as it might change. The costing expires at the next state budget or budget update.

What a budget number can't capture. Fewer people with criminal records, tested products and a local industry are real outcomes of legalisation. So are new compliance, health and road-safety costs, and any shift in spending away from alcohol. None of them shows up in a dividend forecast, and the case for or against legalisation should not rest on one.

Our own model and every assumption in it are set out above, and the calculator lets anyone test them.

About this article

Independent analysis, no affiliation. Marijuana.com.au has no association or affiliation with the Libertarian Party, the Parliamentary Budget Office or any other political party, candidate or elected member, and was not paid or asked to publish this article. It reports on and analyses a published PBO costing ahead of the 28 November 2026 Victorian election. It does not support or oppose any party, candidate or policy, and it is not electoral advertising. Facts are stated as we understood them on 22 October 2026 from the sources listed below, and may change as new information is released.

Not financial advice. Nothing in this article, including our scenarios and the calculator, is financial, investment or business advice, or a recommendation about any company, security or market. Get licensed advice before making financial decisions.

Not medical advice, and no product promotion. This article discusses medicinal cannabis only as a market and policy issue. It does not promote or endorse any therapeutic good, product, brand or supplier, and nothing in it is a claim about the safety or effectiveness of medicinal cannabis. In Australia, medicinal cannabis products are prescription medicines regulated by the Therapeutic Goods Administration, and almost all of them are not approved (registered) medicines. Speak to your doctor or pharmacist about any treatment.

Our estimates. Our model and calculator are simplified estimates, not a costing, and are not produced or endorsed by the PBO or any government. Use of this site is subject to our legal terms.

Sources

• Victorian PBO, Election policy costing: Legalised cannabis (released 21 October 2026; PDF response dated 8 October 2026)
• Victorian PBO, Cannabis consumption in Victoria: value of the illicit cannabis market (2023)
• Victorian PBO, End the nanny state costing (2022)
• Federal PBO, A legal cannabis scheme for Australia (2025)
• Ontario Cannabis Store, Annual Report 2024–25
• Ontario Cannabis Store, Ontario Cannabis Marketplace: By the Numbers 2025
• Ontario Cannabis Store, Annual Report 2019–20
• Alcohol and Gaming Commission of Ontario, cannabis licence fees
• Health Canada, Data on cannabis for medical purposes
• Health Canada, Canadian Cannabis Survey 2024: Summary
• Statistics Canada, police-reported drug crime (2025)
• Public Safety Canada, law enforcement funding for cannabis legalisation (2017)
• Nguyen et al., Changes in Medical Cannabis Use After Recreational Cannabis Legalization in Canada (2023)
• TGA, Reviewing the safety and regulatory oversight of unapproved medicinal cannabis products (August 2025)
• Office of Drug Control import and production data, reported by Stratcann
• AIHW, National Drug Strategy Household Survey 2022–23 release
• ABS, National, state and territory population, March 2026
• Victorian Parliament, Question on Notice 1047: arrests for possession of cannabis
• Canberra Times, ACT review of personal cannabis laws (2024)
• Mital et al., Association between non-medical cannabis legalization and alcohol sales (Drug and Alcohol Dependence, 2024), summarised by Cannabis Science and Technology
• Statistics Canada alcohol and cannabis sales, fiscal 2023/2024, reported by Retail Insider
• Valdes-Donoso, Sumner and Goldstein, Costs of mandatory cannabis testing in California (PLOS ONE, 2020), via ScienceDaily
• TGA, Testing of medicinal cannabis products being supplied in Australia (2020)
• Axios Denver, State audit of Colorado's marijuana testing framework (October 2026)