How Cannabis Markets Actually Work: The Dan K Reports Framework

The cannabis industry has a forecasting problem. Published market projections range from $76 billion to $1.4 trillion. States that modeled hundreds of millions in tax revenue are collecting tens of millions. Operators who raised capital on demand creation assumptions are competing in commodity markets. The gap between what the industry projected and what it delivered isn't a series of unfortunate surprises. It's the predictable result of a framework that never understood what cannabis markets actually are.

This document explains what they are.

The US Market: What It Is and What It Isn't

Applied to the United States — 240 million adults 21+, 18% participation, 1.0 gram per day, 365 days — the demand baseline is 15.8 billion grams annually. At a blended national average of $5.50 per gram, that produces a Total Addressable Market of approximately $87 billion.

The calculation:

• 240 million US adults 21+
• 43.2 million estimated regular consumers (18%)
• 15.8 billion grams annual demand (43.2M × 365)
• $87 billion TAM (15.8B grams × $5.50)

Against actual 2025 US legal sales of $30.1 billion, the implied national legal capture rate is approximately 35% — meaning roughly two-thirds of total cannabis demand in the United States still flows through the illicit channel. That gap is the market. Closing it is the entire story of US cannabis legalization.

What $5.50 per gram represents. Mature legal markets have already demonstrated the floor: Colorado averages $3.18 per gram; Oregon $3.33. These are not loss leaders — they are the natural outcome of a competitive, well-supplied legal market where illicit operators have been neutralized by price parity. Those prices will not return to $6 or above in any mature market. The forces that drove them there — oversupply, cultivator competition, efficiency gains, and the structural ceiling imposed by illicit pricing — are permanent.

The $5.50 figure is a blended national average accounting for the realistic spread across a future fully-legal US market. High-tax states like Illinois may never reach Oregon-level pricing. States entering legalization later will start elevated and compress over time.

Price sensitivity — US TAM at the same demand baseline:

The $87 billion figure is not a forecast for 2030. It is an estimate of total cannabis demand in the United States right now — the majority of it currently served by an illicit market that has had decades to establish itself.

Cannabis Is an Agricultural Commodity. It Always Was.

The most important thing to understand about cannabis markets is the simplest: cannabis is a plant. It requires no exotic chemistry, no proprietary process, no specialized manufacturing infrastructure to produce at scale. The $400/ounce illicit market price was entirely a prohibition premium — compensation for legal risk across the supply chain from cultivator to consumer. Remove the risk and you remove the premium.

Canada proved this is a one-way door. Prices declined every single year after federal legalization without exception. Alberta reached $3.80 CAD per gram — roughly $2.80 USD — within five years. There was no floor set by brand premiums, no consumer segment willing to sustain a premium for legal status once legal status became normal, no supply constraint that persisted once licensing opened. The market found its commodity floor and stayed there.

The outdoor California and Oregon grow cost is the gravitational center of the entire US market. Indoor premium product will always command some differential — as premium wine commands more than table wine — but the floor, where the vast majority of volume trades, trends toward the cost of efficient cultivation plus competitive margins. That number is $2-4/gram retail depending on regulatory overhead. Oregon and Colorado are already there. Every other US legal market is moving toward it, and nothing — not brand investment, not product innovation, not licensing restrictions — has demonstrated any durable ability to hold prices above that floor once supply normalizes.

Regulators cannot limit supply to sustain $6+/gram without handing the illicit market a permanent price advantage. Every state that has tried — Illinois with its limited licensing, New York with its slow rollout, California with its local approval requirements — has simply preserved illicit market share. The illicit grower benefits from the same advances in cultivation technology and genetics as the licensed operator. Their costs are falling too, just without the regulatory overhead. Any attempt to hold legal prices above the competitive level widens the gap the illicit channel exploits.

The long-run revenue TAM is $50-60 billion — not as a pessimistic scenario but as the logical endpoint of a commodity price discovery process already well underway and never reversed anywhere it has run its course.

Revenue Must Fall Before It Grows

This is the thesis the industry doesn't want to publish. The path to the real TAM runs through a compression phase that looks like failure by the metrics everyone watches.

A market that doubles volume while revenue stays flat or declines is succeeding at its actual mission — displacing the illicit channel — while appearing to stagnate. Volume growing, revenue shrinking: that is the mechanism working as designed. Colorado peaked at $2.2 billion in 2021 and has since declined toward $1.4 billion while volume has grown. California peaked and declined three consecutive years. Oregon. Washington. Nevada. Arizona. Every mature market without exception. Revenue peaks within 3-5 years of adult-use launch, then price compression outpaces volume growth and revenue contracts.

Whitney Economics, the most grounded analyst in the space, reported that 2025 was the first year-over-year revenue decline in US legal cannabis history. That's not a blip. That's the mechanism.

The realistic trajectory:

• 2025-2028: Flat to negative revenue CAGR as price compression in existing markets outpaces new state additions. Volume growing, revenue stagnant or declining.
• 2028-2032: Inflection as prices stabilize near competitive floor in most legal states and volume capture accelerates. Revenue CAGR turns positive at 5-8%.
• 2032+: Mature market growth — population growth, modest participation expansion, physician-channel additions. Low single-digit CAGR indefinitely.

The published forecasts of $76-120 billion by 2030 require both higher prices AND higher capture — which is precisely backwards. Higher capture requires lower prices. The path to full capture runs through price compression that temporarily destroys revenue before volume catches up.

The Demand That Already Exists

The $87 billion TAM is not a future aspiration. It is current consumption — the vast majority of it served by an illicit market that has never lacked for customers because the product has been effectively available for decades.

Cannabis has been accessible everywhere for a long time. This is not a drug requiring specialized chemistry or foreign supply chains. In virtually any American community, a person with a handful of contacts and the basic intent to find it could locate a source within days. Supply was never the constraint. The illicit market solved distribution long before any state legislature acted.

The legal risk was always manageable. Prosecution for simple possession was historically a secondary phenomenon — you got caught during a hand-off, during transportation, or because you were already being arrested for something else. The legal exposure was real but modest, and most consumers understood this. Meaningful deterrence requires meaningful risk, and for the average consumer in most jurisdictions, that risk was never particularly meaningful.

The people who wanted cannabis were already using it. The heaviest consumers — those accounting for 50-70% of total volume — are not waiting for a dispensary. They have had supply relationships for years, often decades. Legalization makes their lives more convenient. It does not change their consumption behavior in any material way.

Many people simply don't want it. Consider alcohol: fully legal since 1933, deeply embedded in social culture, sold in every grocery store, carrying zero legal risk. Even so, 54% of US adults currently report drinking — nearly half the adult population abstains. When non-drinkers are asked why, the most common answer is simply that they have no desire to. Cannabis will always face a larger non-participating population than alcohol. These are not barriers legalization removes.

Zero-tolerance employment is a structural ceiling. Transportation, aviation, healthcare, heavy equipment, law enforcement, military, federal contracting — tens of millions of workers for whom cannabis consumption carries ongoing professional consequences no state law can eliminate.

Why Legal Displaces Illicit: The Five Levers

The Black Market Death Equation scores five variables on a -1 to +1 scale, each weighted by its empirical contribution to legal capture rates. Variables don't simply help or not help — they can actively damage a market's score.

1. Price competitiveness (4× weight). The dominant variable by a wide margin. Markets with retail flower under $5/gram score 0.85 or higher and achieve near-complete displacement. Markets where legal prices significantly exceed illicit prices score negative — and no combination of the other four variables compensates for a deeply negative price score.

2. Retail density (1× weight). Stores per 100,000 adults. Below roughly 3-4 stores per 100,000, the legal channel isn't physically accessible to most consumers regardless of price or quality. Virginia at 0.3 and Rhode Island at 0.96 are effectively negative on this variable.

3. Product quality and variety (1.2× weight). Markets with full product lines — flower, concentrates, vapes, edibles at 30%+ of sales — score 0.85 or above. The legal channel's structural advantage here is permanent: lab testing, consistent potency, and real selection are things the illicit market structurally cannot replicate.

4. Transaction convenience (1× weight). Posted hours, online menus, loyalty programs, and payment options versus a phone that may or may not be answered. Most mature legal markets score positively here by default. Excessive regulatory friction — cash-only, restrictive hours, no delivery — pushes the score negative.

5. Enforcement intensity (0.6× weight). The mechanism is simple: close the unlicensed storefronts. No unlicensed retail operating openly scores +1. Any meaningful unlicensed storefront presence scores -1. New York and California aren't stuck at 16% and 63% because of supply-chain failures — they're stuck because unlicensed stores compete directly with licensed operators at lower prices and the state hasn't closed them.

The state grid below is the output of this equation across every market studied. Markets that score positively across all five variables achieve near-complete or above-100% capture. Markets that score negatively on price and enforcement simultaneously get stuck — regardless of how long they've been legal.

For the full BMDE equation, variable scoring, and state-by-state scores, see Why Legal Cannabis Markets Fail.

Why New Product Categories Don't Expand the Pie

New consumption formats — beverages, concentrates, edibles — appear in nearly every bullish cannabis forecast as drivers of TAM expansion. The data suggests otherwise.

Product diversification is revenue mix shift, not market growth. When a daily flower consumer switches to vape cartridges or edibles, total cannabis consumption doesn't increase. The same underlying demand gets routed through a different delivery mechanism. Alberta's closed-system data shows this precisely: dried flower is only 34.6% of volume in a mature market, with extracts at 21%, pre-rolls at 17.6%, and vapes at 15%. Demand diversified across delivery formats. Total consumption per participant didn't change.

Beverages illustrate the pattern clearly. Cannabis beverages have been legal in Canada since October 2019. After six years of legal availability and significant operator investment, beverages represent 0.8% of volume in Alberta. The observed behavior is consistent with a straightforward substitution problem: the edible already addressed the primary consumer case for beverages — discrete, dose-controlled, no device — and did it at lower cost with better dose clarity. A gummy is a defined unit that a consumer can reason about. A beverage gets consumed at a socially determined pace with highly variable onset depending on metabolism, food consumption, and tolerance. The cost structure compounds the issue — water weight, refrigeration, liquid distribution, and packaging all layer onto the cannabis input cost, producing a price per milligram of THC well above gummies or flower. The category will find its niche, much like hard kombucha found a shelf. The data doesn't support it as a TAM driver.

Consumption lounges face both a consumer preference problem and a structural economics problem. On the consumer side, the vape pen effectively solved public discretion before the first lounge opened — a vape pen in a pocket is invisible, vapor dissipates quickly, and anyone who wants to consume in a public setting already does so without a dedicated venue. Amsterdam's coffee shops emerged specifically because home consumption was legally ambiguous; the café was the only reliable sanctioned space. That structural rationale disappears once home consumption is normalized.

On the economics side, a bar generates revenue by continuously serving the same seat for hours. Observed behavior at cannabis lounges shows a different pattern: a consumer arrives, purchases one preroll or shares one among a group, and is comfortably set for two hours without further purchases. Unlike alcohol, additional cannabis consumption beyond a personal threshold produces diminishing returns that most experienced consumers recognize quickly. The table turns once. No amount of ancillary offerings has yet demonstrated the ability to compensate for a product that drives a single purchase per visit.

New product categories follow demand — they don't create it. Every successful format expansion has been a better delivery mechanism for existing consumers, not a mechanism for generating new demand from non-consumers. The observed abstention data is consistent with this: the population not currently using cannabis is primarily non-users by preference, not by format availability.

Legalization Is a Logistics Problem, Not a Demand Problem

Only one country has ever designed its legalization framework around this principle. That country is Canada.

Canada's Cannabis Act of 2018 stated its objective explicitly: displace the illicit market. It wrote that goal directly into legislation alongside protecting youth and public health. That framing had regulatory consequences. It forced policymakers to ask "are we actually winning against the illicit channel?" as an ongoing performance metric. When the answer came back slower than expected, it produced real policy responses — price adjustments, licensing acceleration, product category expansion. The goal was stated, so failure to meet it was measurable and politically accountable.

US legalization campaigns were built around a different coalition. The arguments that won ballot initiatives were personal freedom, racial justice in enforcement, and tax revenue for schools and infrastructure. Those are winning political messages. Black market displacement is not a bumper sticker — and critically, it never got encoded into the regulatory architecture afterward either.

The tax revenue framing is almost perfectly opposed to displacement as a goal. If your objective is maximizing tax revenue, you set rates high. If your objective is displacing the illicit market, you set rates low enough to compete. These are in direct tension. Virtually every US state struggling with illicit market persistence — Illinois, California, New York — chose tax revenue optimization and expressed surprise when illicit operators retained their customer base. They optimized for the wrong variable because the wrong goal was on the scoreboard.

No US state systematically publishes a displacement metric. Canada tracks illicit market share as an ongoing policy output. In the US, states report legal sales revenue, tax receipts, and licensee counts. Almost none publish a formal estimate of what percentage of total consumption their legal market is capturing. The metric doesn't exist because the goal was never stated.

What legalization does is transfer demand from an illicit channel to a legal one. It does not conjure demand that did not previously exist. Canada understood this from the start. US cannabis policy, with few exceptions, still hasn't fully reckoned with it.

The Illicit Market Has Fundamentally Changed

The illicit cannabis market that existed before legalization — criminal supply chains, dedicated distribution networks, wholesale relationships between growers and dealers — has been largely destroyed by commodity pricing in legal states. At $3/gram in Oregon and Michigan, the economics that sustained a professional illicit supply chain no longer exist. You cannot build a viable criminal enterprise around a product that retails legally for less than a cup of coffee.

What replaced it is structurally different and far harder to suppress. The modern illicit supply in prohibition states is primarily sourced from legal markets — not grown illicitly, not imported from foreign criminal networks, but purchased legally at $62/oz in Michigan or $80/oz in Oregon and redistributed informally into states where the same product commands $200-250/oz. The margin is 3-4x on a single legal purchase. The legal risk is concentrated entirely in the transportation or shipping step — not in growing, not in processing, not in any step that requires criminal infrastructure.

Oregon's own data confirms this dynamic explicitly. The state has spent $46 million fighting its illicit cannabis market and openly admits it cannot determine whether any of it is working — because the enforcement apparatus is searching for supply in Oregon that is already gone. The Oregon Criminal Justice Commission stated in both its 2024 and January 2025 annual reports: "It is not possible to draw conclusions about whether the grant has reduced Oregon's illegal marijuana market at this time." Between 2015 and 2018 alone, authorities seized $48 million of Oregon cannabis out of state, found en route to 37 other states — primarily Texas, Florida, Wisconsin, Missouri, Virginia, Illinois, Arkansas, Iowa, and Maryland. That was before Oregon prices collapsed to $3/gram. The economics today are dramatically more compelling.

Oregon's own policy researchers acknowledge that stagnant prices and massive oversupply may be contributing to increased participation in the illicit market, as licensed growers look for other ways to recover losses. This is the piece most policy analysis misses entirely. The modern illicit supply chain isn't a competing criminal enterprise — it is Oregon and Michigan's legal supply chain operating under different economics. A licensed cultivator sitting on product that wholesales at $1/gram in an oversupplied legal market has every rational incentive to develop relationships in prohibition states where the same product is worth $8-10/gram. No criminal infrastructure required. Industry events, trade forums, existing contacts provide the relationship layer.

Michigan's Cannabis Regulatory Agency director has explicitly stated the state has become a central location for illicit operations because penalties for illicit activity are so low — while the legal industry privately attributes depressed prices partly to illicit product entering the licensed supply chain. Ohio passed legislation specifically making it illegal to bring cannabis from another state into Ohio, naming Michigan directly in public commentary. The informal personal reshipping network — someone buying legally in Michigan and redistributing to a small social network in a prohibition state at a meaningful markup — requires no criminal organization, no dedicated infrastructure, and produces legal exposure only at the point of interstate transport.

The publicly traded MSOs understand this compliance risk intuitively. It explains one of the most conspicuous patterns in US cannabis: zero major MSO presence in Oregon, Washington, or Oklahoma — the three most oversupplied, lowest-price legal markets in the country. This is not a margin calculation. It is a compliance calculation. A Nasdaq-listed company with SEC reporting obligations cannot participate in supply chains where chain of title is ambiguous — and in these markets it demonstrably is. The small independent operators filling those markets don't face the same obligations and have accepted that reality.

The implication for TAM modeling is significant. Oregon's 100% legal capture and Michigan's 165% are accurate representations of legal channel sales in those states. They are not accurate representations of where all that product is consumed. A meaningful share of those figures represents product that exits informally and is consumed in Indiana, Tennessee, Florida, or Texas — appearing as Oregon or Michigan legal revenue while actually serving demand in prohibition markets. The national 35% legal capture figure likely overstates true displacement of illicit demand in prohibition states, because some fraction of that "captured" legal revenue is simply the supply side of an informal arbitrage network serving the same consumers the legal market hasn't yet reached.

The clearest evidence that the legacy illicit supply chain has collapsed is at the US-Mexico border. In fiscal year 2013, US Customs and Border Protection agents seized 2.4 million pounds of cannabis at the southwest border. By 2023 that figure had fallen to 61,000 pounds — a 97.5% decline in a decade. The DEA's own National Drug Threat Assessment states explicitly: "In US markets, Mexican marijuana has largely been supplanted by domestic-produced marijuana." Cannabis is now being smuggled in the other direction — US-produced legal product entering Mexico, not Mexican product entering the US. The cartel supply chain that served the American market for fifty years was not suppressed by enforcement. It was economically destroyed by domestic legal production at higher quality and lower cost.

At the state level, a March 2026 peer-reviewed study published in the International Journal of Drug Policy by researchers at Columbia University's Mailman School of Public Health found adult-use legalization associated with a 45% relative reduction in state law enforcement cannabis seizures. Within legal states, only 6% of consumers report primarily sourcing from a dealer. In Oregon's own 2024 enforcement data, only 15% of investigated incidents involved larger criminal organizations. The organized crime model is functionally gone even in the most oversupplied markets.

This border seizure data carries a direct TAM implication that almost no analyst has made explicit. The TAM figures produced between 2015 and 2022 were built on a demand baseline calibrated to a market predominantly served by Mexican-produced brick weed at 3-5% THC. The illicit consumer at the center of those models was consuming a specific volume of a specific product at specific prices. Legal markets replaced that supply with domestically-produced product at 20-25% THC, lab-tested, and priced at commodity floor. The transition involves two compounding corrections that cut the same direction simultaneously.

First, volume. Higher-potency domestically-produced cannabis requires materially fewer grams per session to achieve equivalent effect compared to the legacy product those demand baselines described. A model that captures the illicit market and prices it at legal market rates is inflating the consumption baseline — the volume embedded in pre-legalization illicit market estimates reflects a lower-potency product that no longer dominates the market.

Second, price. The $400/oz illicit price was a prohibition premium, not a reflection of production cost. Legal markets don't capture that premium — they eliminate it. The revenue that emerges from capturing illicit demand is revenue at commodity pricing on legitimate consumption volumes, not revenue at prohibition pricing on inflated volumes. Both inputs compress simultaneously.

The honest read of the 97.5% border seizure collapse is not "legalization worked" — it is "the market most TAM models were calibrated to no longer exists." The $87B TAM in this framework is built on actual legal-market consumption data from closed systems at commodity prices. The $50-120B figures in commercial market research reports are largely built on a market dynamic that disappeared when the first Oregon dispensary opened.

What Would Actually Move TAM

Less than most people assume, and through specific pathways rather than broad normalization.

Perception improvement is real but already largely priced in. A 2014 CNN poll found 18% of Americans believe alcohol should be illegal — a substance legal and culturally embedded for nearly a century. Cannabis legalization support sits around 68%, implying roughly 32% still oppose it. The people most likely to convert as stigma fades have already largely done so. What remains is a population of committed opponents. Perception improvement is a modest TAM driver already incorporated into the 18% assumption.

Medical research and physician adoption is the only pathway to material TAM expansion. The single most powerful thing that could meaningfully grow total cannabis demand is physicians recommending it. Not dispensary budtenders. Not social normalization. Doctors.

The populations currently outside the 18% participation figure are disproportionately older, more conservative, more deferential to medical authority. These are not people who will walk into a dispensary because it got less stigmatized. But a meaningful fraction will follow a physician's recommendation for chronic pain, anxiety, sleep disorders, or inflammation. An older adult managing chronic pain who adds cannabis to their regimen is genuinely new demand — they were not in the illicit market. They needed a doctor to tell them it was appropriate.

This is why rescheduling matters beyond the 280E implications that dominate industry coverage. Schedule III or below unlocks the research pipeline that produces the clinical evidence that moves physicians. The chain runs: Rescheduling → Trials → Evidence → Guidelines → Prescriptions → New Participants. That chain is long and slow, but it is the only identifiable pathway to TAM expansion that doesn't rely on capturing demand that already exists.

Employer testing policy erosion is gradual but real. The zero-tolerance employment ceiling is already cracking. As more states legalize and labor markets tighten, employers are quietly dropping random testing programs that are increasingly difficult to justify. This is genuine TAM expansion — people currently excluded by professional risk who would participate given the opportunity, concentrated in higher-income demographics that would skew toward the legal channel.

The long-run ceiling is higher than most analysts model. The 60-65% participation rate alcohol sustained historically is not an unreasonable long-run analog — particularly if clinical evidence eventually reframes cannabis as a broadly useful therapeutic tool. That outcome takes decades and requires the research infrastructure that rescheduling could unlock. The 18% figure used in this framework is a near-to-medium term convergence benchmark, not a claim about the ultimate ceiling.

The Framework: Why 18% and 1.0 Gram

The BMDE framework uses two anchor figures: 18% adult participation and 1.0 gram per day average consumption. Neither is a precise empirical truth. Both are used deliberately.

The 18% figure is drawn from national consumption surveys measuring past-year use — the broadest standard definition. If anything, it understates true participation. Survey respondents systematically underreport cannabis use due to lingering stigma and legal exposure concerns. The real past-year participation rate in a fully normalized environment likely exceeds 20%.

But past-year use is a wide net. It includes someone who consumed at a party in February and hasn't touched it since. These occasional consumers technically qualify as participants — but they contribute almost nothing to total demand. A consumer using five or six times a year at 1-2 grams per occasion produces maybe 10-15 grams of annual consumption. Against a daily user producing 300-500 grams per year, they are a rounding error.

The population that actually drives cannabis demand is more precisely the past-week user — probably 10-14% of adults. That population captures every daily and near-daily consumer, every committed weekly user, and essentially all the volume consumed in the United States. These are the consumers who were already in the illicit market before the first dispensary opened.

So why use 18% at 1.0 gram rather than 12% at 1.5 grams?

Because 1.0 gram per day is not the heavy user consumption figure — it is the blended average across the entire participating population, including the occasional tail that past-year surveys capture. The daily core consuming 1.5-2 grams gets averaged with the monthly consumer contributing 0.05 grams per calendar day, and the result approximates 1.0 gram per day across the full 18%. The two assumptions are internally consistent. Model it with segments and you arrive at essentially the same TAM. What no one in the cannabis market research industry has ever published is a rigorous segmented demand model that shows the daily user contribution, the weekly contribution, and the occasional contribution separately. That analysis does not exist in the public literature.

What "1.0 gram per day" means in practice. Consider what one gram actually represents. Consumed efficiently — via one-hitters, pipes, or low-waste delivery methods — a single gram can sustain a full day of use. A 6:30am wake-and-bake, followed by measured titration every few hours through a 10pm endpoint, requires surprisingly little material when delivery is efficient.

Several dynamics keep the upper tail of consumption compressed in ways that don't apply to alcohol:

Over-titration has diminishing returns. Cannabis exhibits a ceiling effect for most experienced users. Beyond a personal threshold, additional consumption produces minimal additional benefit. This isn't moralizing — it's pharmacology, and consumers understand it intuitively.

T-break culture is real and pervasive. Tolerance management through deliberate breaks is a documented, widespread practice. Users actively manage their tolerance to preserve cost-effectiveness and effect quality. This behavioral pattern has no meaningful analog in alcohol consumption.

Users optimize for cost-effectiveness. Cannabis consumers seek to maximize effect-per-dollar. This creates natural downward pressure on average consumption that doesn't exist for less expensive, more socially normalized substances.

Increasing potency tempers volume. As average THC concentrations have risen, the volume of material required to achieve a given effect has declined.

The empirical validation doesn't come from surveys. It comes from closed legal systems where illicit competition has been eliminated, making legal sales data a direct proxy for actual consumption. Florida's vertically integrated medical system — mandatory seed-to-sale tracking, no unlicensed storefronts, meaningful enforcement — shows 0.89 grams per day across 930,000 active patients. Alberta's provincial wholesale monopoly (AGLC) shows 182,442 kg sold annually against a BMDE-derived TAM of ~242,000 kg at 18% participation — producing 75% legal capture at $3.80 CAD/gram, consistent with near-complete displacement when homegrow and gifting (both federally legal in Canada) are accounted for. Neither survey. Both closed systems. Both consistent with 1.0 gram per day.

The 1.5 g/day industry-standard figure fails this test at every data point and in every closed system studied. It is the product of early industry-funded research during the 2015-2019 investment boom when everyone needed the market to be as large as possible. It got cited in one report, then cited by the next report citing the first, and became institutionally embedded before the legal markets produced transaction data that contradicted it. That data now exists. The figure hasn't moved.

The full empirical case for 1.0 g/day — including multi-jurisdiction closed-system analysis and the Florida dispensation data — is published at Cannabis Consumption and the Phantom Demand Problem. The peer-reviewed paper is available on SSRN and the underlying dataset on Harvard Dataverse.

Why the Published Forecasts Are Wrong

Published US cannabis forecasts range from $39 billion by 2029 (BDSA) to $428 billion by 2032 (Fortune Business Insights). The lower end reflects actual market data. The upper end reflects something else entirely.

The $400+ billion figures combine four methodological failures:

They are global, not US. The Fortune Business Insights $444 billion figure covers every country on earth, including projections for China and India that require assuming national legalization programs that don't exist and have no near-term legislative path.

Hemp and CBD are bundled in. "Cannabis market" in most reports includes hemp fiber for construction, hemp seed oil in food, CBD in cosmetics, and CBD as a pharmaceutical ingredient. These are not the same market as someone buying flower at a dispensary. Once you include industrial hemp as an agricultural commodity, the numbers explode without representing anything meaningful about the intoxicant market.

Pharmaceutical assumptions are layered on. Most bullish reports assume cannabis-derived compounds eventually get approved as pharmaceutical treatments at pharmaceutical pricing — $50-100/gram equivalent rather than $5/gram dispensary pricing. If even a fraction of the global pain management pharmaceutical market converts to cannabis-derived treatments at pharmaceutical margins, you can justify almost any number.

CAGR does the compounding work. At 34% annually a market doubles every two years. Apply that to any starting point and you get a very large number regardless of whether the growth rate is physically, economically, or logistically possible.

The historical track record is brutal. Grand View Research projected global legal marijuana at $146 billion by 2025 in a 2018 report at a 34.6% CAGR. Actual global cannabis revenue in 2025: roughly $68 billion — less than half of what was projected seven years out. Nobody went back to audit this. The same firms published new reports with new horizon dates and new large numbers. The product isn't accurate forecasting. The product is a large, official-looking number someone can put in a pitch deck.

The firms with actual point-of-sale data — Whitney Economics, BDSA, Statista — project $39-55 billion by 2029-2030 at 3-6% CAGR. These are the numbers grounded in what the market actually is. They are also materially lower than what the industry has spent fifteen years telling investors to expect.

Hemp-Derived THC: Prohibition Demand Made Visible

BDSA reported $21.8 billion in intoxicating hemp-derived THC product sales in 2025 — delta-8, THCA flower, hemp-derived gummies and beverages sold in gas stations, convenience stores, and smoke shops across America. At first glance this looks like an additive market: cannabis plus hemp equals a larger total opportunity.

It is not. It is the same demand.

The $21.8 billion in hemp-derived THC sales represents consumption that was already in the illicit market, temporarily routed through a federal loophole created by the 2018 Farm Bill's hemp definition. Texas and Florida — both prohibition states — are among the top hemp-derived THC markets in the country. These are not new cannabis consumers. They are existing consumers who found a legal-ish retail channel for an inferior substitute because their state doesn't offer a licensed dispensary option.

The product is structurally inferior: delta-8 and other synthetically converted cannabinoids have inconsistent potency, uneven quality control, and limited lab testing compared to licensed cannabis. Consumers in prohibition states buy it because it is available, not because it is preferred. This is the same dynamic as the informal reshipping network from Michigan and Oregon — existing demand finding the best available channel, not new demand being created by a new format.

The TAM implication is direct: the $21.8 billion is not additive to the $87 billion framework figure. It is a subset of it — prohibition-state demand that has self-organized around the best available option. When Florida or Texas eventually legalizes, hemp-derived THC revenue does not stack on top of new dispensary revenue. It converts. Licensed cannabis captures those consumers with a better product at competitive prices, and the hemp channel largely collapses. The hemp figure is a preview of what large prohibition-state markets are worth, not an increment to the addressable market.

The November 2025 Farm Bill amendment substantially closed the federal loophole by redefining hemp to exclude intoxicating cannabinoids. What portion of that $21.8 billion survives in state-level programs, migrates to licensed dispensaries, or returns to informal channels remains to be determined. But the honest read is straightforward: hemp-derived THC proved that demand already exists at scale in prohibition states. Florida and Texas are not waiting for legalization to create cannabis consumption. They are already consuming at a rate that supports a multi-billion dollar market through an inferior, legally fragile channel. The TAM is already there. The legal market just hasn't been allowed in yet.

280E Removal: Survival Filter, Not TAM Driver

President Trump's December 2025 executive order expediting cannabis rescheduling from Schedule I to Schedule III removes Section 280E of the Internal Revenue Code as a burden on licensed operators. 280E disallows standard business deductions — cost of goods sold, rent, labor, operating expenses — for any enterprise trafficking a Schedule I or II controlled substance. For profitable cannabis operators this functions as a confiscatory tax, in some cases pushing effective federal tax rates above 70%.

Rescheduling to Schedule III eliminates this entirely. Industry estimates put the unlocked after-tax cash flow at $1.6-2.2 billion annually across the licensed market. This is a federal benefit — uniform across all operators regardless of state, and distinct from the state-level 280E decoupling that California, Colorado, Illinois, New Jersey, New York, Oregon, Vermont, and Connecticut had already implemented. For operators concentrated in those states, the incremental state-level relief is marginal. The federal removal is the real event.

The operators who benefit most are those with the highest current federal taxable income being taxed without deductions. Trulieve — with $427 million in adjusted EBITDA and management explicitly stating that 2025 GAAP net income would have been positive without 280E — is likely the largest absolute beneficiary. GTI, which posted $114 million in GAAP net income despite 280E, also has substantial exposure. Curaleaf's persistent GAAP losses despite strong adjusted EBITDA makes its relative benefit more modest.

280E removal does not move TAM. It does not create new consumers or expand the addressable market. Its function is more specific: it determines which operators are still solvent when the next major market openings — Florida, Pennsylvania, Texas — eventually occur. The compression cycle currently underway is eliminating undercapitalized operators across every legal state. 280E removal meaningfully improves the odds that the current MSO leaders are still standing when those markets open, rather than being acquired at distressed valuations or simply failing in the interim. It is a survival filter for the next phase of the industry, not a growth catalyst for this one.

The Framework in Context

These assumptions — 18% participation, 1.0 g/day consumption — are tools for constructing a consistent, comparable baseline across jurisdictions. They are not claims about current behavior or predictions about any specific state's trajectory.

They answer a specific question: what does a fully normalized cannabis market look like at minimum, and what share of it is the legal channel actually capturing?

The answer, consistently, across every state and every legal market studied on this platform, is that the full market is substantially larger than current legal sales suggest — and that the gap between what the legal channel captures and what the full market supports is almost entirely explained by price competitiveness and retail density. Not enforcement. Not product selection. Not marketing. Price and access.

The states that got both right — Colorado, Oregon, Massachusetts, Montana, Michigan — captured their markets completely. The states that got either wrong still have most of their market in the illicit channel. The math is not subtle and it is not complicated. It has just never been applied honestly.

State-by-State Analysis

The framework above produces testable, falsifiable predictions for every legal market. Below is every US state and Canadian province analyzed on this platform, with the key metrics that determine legal capture outcomes.

Click any state for the full analysis.

Price reflects retail flower per gram from official state/provincial regulatory data. Legal capture calculated against BMDE framework TAM (18% participation, 1.0g/day). Density reflects licensed retail stores per 100,000 adults.

Consumption baseline derived from the multi-jurisdiction consumption paper with published datasets on Harvard Dataverse and available on SSRN. For the Black Market Death Equation methodology, see Why Legal Cannabis Markets Fail. For real-time state pricing, see Cannabis Prices by State. For legislation tracking across all 50 states, see Cannabis Legislation Tracker.