The Quiet Substitution: Why the People Who Actually Switched From Beer to THC Are Switching All the Way

The Top-Line Number Everyone Is Quoting

Gallup's July 2025 poll dropped the same line into every food-and-beverage trade publication for six straight months: 54% of Americans aged 18 and older drink alcohol — the lowest figure Gallup has recorded in its 90-year tracking history. The reading was 62% in 2023 and 58% in 2024. The previous all-time low of 55%, set in 1958, has now been broken.

Layer in the trailing-twelve-month sales data through the 52 weeks ending January 3, 2026: beer down 3.3%, spirits down 4.1%, wine down 6.1%. Every legacy alcohol category is shrinking. For the first time in Gallup's trend, a majority of Americans — 53% — say drinking in moderation is bad for one's health. Six percent still say it's good for you. The rest are agnostic.

The instinctive read, and the one most cannabis-industry press has rushed to, is straightforward: alcohol is falling because Americans switched to weed.

It's not quite right.

What Gallup Actually Said About Substitution

Two-thirds of the way into the same Gallup analysis, the analysts pour cold water on their own clickbait: "Declines in alcohol consumption do not appear to be caused by people shifting to other mood-altering substances — in particular, recreational marijuana, which is now legal in about half of U.S. states. Although marijuana use is higher today than a decade ago, it has been fairly steady over the past four years and thus doesn't appear to be a factor in people choosing not to drink alcohol."

At a population level, the decline in drinking is being driven by health perception, not substitution. The decline among women (down 11 percentage points since 2023, to 51%) is more pronounced than among men (down 5 points, to 57%). That's a wellness-culture footprint, not a cannabis footprint.

So why does the cannabis-beverage market keep posting double-digit growth?

Because the population-level signal and the cohort-level signal are two different stories.

The Cohort-Level Story

The most-cited substitution data point in the industry comes from a January 2026 study published in the Journal of Psychoactive Drugs and amplified by NORML and ScienceDaily: people who started drinking cannabis beverages cut their alcohol consumption from 7.02 drinks per week to 3.35 drinks per week — a 52% drop.

Other findings from the same cohort:

• 58.6% of cannabis-beverage users reported substituting cannabis for alcohol (versus 47.2% of non-users)
• 62.6% either reduced (61.5%) or completely stopped (1.1%) drinking after starting cannabis beverages

This is not a population-wide trend. It is a self-selecting cohort effect, and it is enormous within that cohort.

The market data corroborates the cohort findings:

• 27% of Gen Z and 26% of Millennials consumed a THC beverage on-premise in the past six months (NIQ, fall 2025)
• 38% of Gen Z and 37% of Millennials plan to try a cannabis beverage
• U.S. cannabis-infused beverage sales are projected to reach $2.8 billion by 2028, a 16.9% CAGR
• A separate FMI forecast pegs the global cannabis drinks market at $1.37 billion in 2026, growing to $23.8 billion by 2036 at a 37.3% CAGR

The split, then: alcohol drinkers in their 40s, 50s, and 60s are quitting for health reasons. Twenty- and thirty-somethings deciding whether to start drinking at all are looking at a 2.5mg-THC seltzer and choosing that instead. "No, I'm fine — I'll have the THC one" at a bar is becoming culturally legible in a way it wasn't five years ago.

What the Bar Looks Like Now

The defining product feature of the cannabis-beverage wave is low-dose formulation: 2 to 10 milligrams of THC per serving, in seltzer or tonic format, often with botanical flavors and zero alcohol. This is not a substitute for getting high. It is a substitute for one or two glasses of wine.

The packaging is deliberate: 12-ounce can, often in flavors that mirror hard seltzer — lemon-lime, grapefruit, blackberry. The dosing curves are designed to peak in 15 to 30 minutes and clear in 60 to 90 — closer to a beer than to a 100mg edible.

Restaurant menus are catching up. New restaurant concepts in California and Colorado are offering "cannabis pairings" alongside wine pairings. Functional-beverage sections at higher-end grocery — Erewhon, Whole Foods, the new Wegmans wellness aisles — are devoting increasing shelf inches to THC seltzers next to kombucha and adaptogen drinks. The category code at retail is shifting from "cannabis product" to "functional beverage."

Cann, the most-cited celebrity-backed entry in the space, has more than 40 celebrities on its cap table (Paltrow, Rosario Dawson, others). That cap-table choice itself is the product story: this isn't a cannabis brand pretending to be a beverage. It's a beverage that happens to use cannabis as the active ingredient.

What's Hurting (and Who Loses)

Bars are the soft target. Technomic data shows roughly 57% of consumers swap alcohol for cannabis at least once a month — and that swap is happening at on-premise venues, not just at home. A bar that doesn't carry a THC option is losing a Friday-night order it used to get. By 2027, expect THC-seltzer SKUs on most mainstream on-premise menus in legal states the same way White Claw is on every menu now.

Wine and craft beer are more structurally exposed. Wine fell 6.1% YoY in the 52 weeks ending January 3, 2026 — the steepest decline of any major alcohol category. Wine's brand promise — relaxation, sophistication, low-key social — is exactly what a 5mg THC seltzer also delivers, at lower calorie count, with no hangover, and a cleaner ingredient label. There is no reason to assume wine is the floor of this trend.

Spirits brands have started preemptively launching low-/no-alcohol lines and hedging into functional categories. Diageo, Pernod Ricard, and AB InBev all have CBD or non-alcoholic functional pilots in market. The diversification is happening because the core thesis — that alcohol is a permanent, growing category — is no longer the consensus view.

What This Means for the Cannabis Sector

If you read the cannabis-beverage trajectory correctly, three things follow.

The substitution is asymmetric. Cannabis-beverage growth is being driven by people who weren't previously cannabis consumers. The format is the gateway. The flower-and-edibles core of the MSO economy is not where the new dollars are flowing — and the people new to cannabis are unlikely to graduate to a vape pen. They want a seltzer.

The category needs the rest of the regulatory machinery. On-premise sales of THC beverages still face state-by-state friction: most states either prohibit them, restrict them to dispensaries, or run them through experimental hemp-derived loopholes. The structural growth case requires the same federal-policy moves the MSO equity market needs — broader rescheduling, FDA pathways for low-dose THC beverages, ideally a beverage-specific category code.

Brand ownership will rewrite the cap table. Cann's celebrity-on-cap-table model has more institutional logic than the legacy MSO model. A cannabis-beverage brand with the right cap-table mix can plausibly list on a major U.S. exchange under existing FDA/FDCA pathways well before a plant-touching MSO ever can. The first plant-touching IPO on NASDAQ may not be a multi-state cultivator. It may be a seltzer brand.

The Gallup top-line read is real: Americans are drinking less because they think drinking is bad for them. But the people who are actively replacing the drink they used to order with something else — the cohort hidden inside the averages — are replacing it with cannabis. Half as many drinks per week. Same social occasion. A different glass.