The Addiction Economy

Five industries now generate over $700 billion a year from products deliberately designed to be hard to put down. They share a business model, a design logic, and a playbook. This is not five separate problems. It is one system.

The Business Model Nobody Names

Here is something the five industries that sell you gambling, alcohol, cannabis, nicotine, and social media would prefer you not think about as a single sentence: they all depend, primarily, on their most damaged customers.

Philip Cook's landmark research on alcohol showed that the top 10% of American drinkers — people averaging 74 drinks per week, roughly four and a half bottles of whiskey — account for more than half of all alcohol sold in the United States. In sports betting, the numbers are more extreme: somewhere between 2% and 3% of sportsbook users generate 60% to 70% of gross gaming revenue, and research now shows that roughly half of all online sports-betting revenue comes from people meeting clinical criteria for gambling disorder. In nicotine, the dependent user is not the edge case — they are the market. On social platforms, Pew found the top 10% of users produce 80% of content and log a disproportionate share of time. In cannabis, daily users — who now outnumber daily drinkers for the first time in recorded history — dominate revenue despite being a fraction of the user base.

The industry term for this customer is the "whale." The whale is the business model. And the whale is usually someone in trouble.

This matters because it reframes every conversation about personal responsibility that these industries work extremely hard to keep front and center. The system is not accidentally reliant on its most vulnerable users. It is structurally dependent on them. Reducing harm to the whale doesn't trim the margins — it threatens the income statement. Which is why, across all five sectors, you find the same pattern: sophisticated harm-reduction marketing aimed at casual users who don't actually need it, and sophisticated engagement mechanics aimed at heavy users who desperately do.

Gambling's Seven-Year Experiment

In May 2018, the Supreme Court's decision in Murphy v. NCAA struck down the federal ban on sports betting, handing the question to states. Thirty-eight states and D.C. have now legalized it. In 2024, Americans wagered $149.9 billion on sports — up 23.6% year-over-year — generating $13.7 billion in gross gaming revenue for operators. Total commercial gaming revenue hit a record $71.9 billion, its fourth consecutive all-time high.

The industry is a functional duopoly. FanDuel holds roughly 41% of the gross gaming revenue market; DraftKings holds about 33%. Together they spend more on marketing than most pharmaceutical companies spend on drug development. DraftKings alone spent over $500 million on promotions and advertising in 2024, including free-bet offers designed to convert casual football fans into active accounts during the critical first months of legalization in each new state.

Seven years in, the harm data has arrived and it is not ambiguous. A 2024 working paper from UCLA and USC found that every dollar wagered on sports betting was associated with a roughly two-dollar reduction in net household investment — meaning people aren't funding their bets from their entertainment budget. They're liquidating savings, retirement accounts, and brokerage positions. A separate NBER paper found that household bankruptcies rose approximately 28% in states with legal online wagering, with the increase concentrated about two years after launch — long enough for problem gamblers to exhaust their resources. State problem-gambling helplines have seen call volumes increase by 150% nationally since 2019; Massachusetts, which launched mobile wagering in 2023, saw a 276% increase in its first full year. University of Oregon researchers found that legalization also correlated with increased intimate-partner violence following home-team losses — a pattern consistent with earlier research on illegal gambling, now scaled to millions of phone-based accounts.

The design of the product explains the mechanism. Same-game parlays, the fastest-growing bet type, carry house edges of 30% or more — several times higher than standard spreads — but offer larger potential payouts that make them psychologically irresistible. In-game micro-betting turns each pitch, possession, and free throw into a wagering opportunity, keeping users in what gambling researchers, following anthropologist Natasha Dow Schüll's influential work, call the "machine zone" — the dissociative flow state that problem gamblers describe as the actual product they're buying. Push notifications are algorithmically timed to the user's betting history. VIP hosts assigned to high-value accounts have, according to multiple pending lawsuits, escalated bets rather than flagging problem behavior. In one Baltimore case, a user's monthly deposits rose from $2,000 to $64,715 over three years while employed at $175,000 — and the company assigned him a personal account manager throughout.

Alcohol's Quiet Reclassification

The most consequential public health development of the last two years may be one that received surprisingly little sustained attention: in January 2025, Surgeon General Vivek Murthy issued an advisory formally linking alcohol consumption to seven cancers — breast, colorectal, esophageal, liver, mouth, throat, and larynx — and called for updated warning labels. The advisory attributed roughly 100,000 cancer cases and 20,000 cancer deaths annually to drinking. The World Health Organization had reached a similar conclusion in 2023: no level of alcohol consumption is safe for human health.

Alcohol warning labels in the United States have not been updated since 1988. They still mention birth defects and driving. They do not mention cancer.

The epidemiology behind those advisories reflects a generational split that the industry is watching carefully. Among Americans under 35, the share who drink at all has fallen from roughly 72% in the early 2000s to around 50% today — a decline driven by Gen Z's broadly health-conscious orientation and their substitution of cannabis for alcohol. But among Americans over 55, drinking rates have risen. Alcohol use disorder in Americans over 65 increased 107% between 2001 and 2013, and female alcohol-attributable deaths in that age group have risen nearly 7% per year since 2012. The archetype of the American problem drinker has quietly shifted from a young man to a middle-aged or older woman — a shift the treatment infrastructure has been slow to acknowledge.

Meanwhile, alcohol-attributable deaths rose 29% between 2016–17 and 2020–21, from approximately 138,000 to 178,000 annually. The industry — $280 to $315 billion in U.S. retail sales, dominated by AB InBev, Diageo, and Constellation Brands — has responded to the declining youth market by pivoting aggressively into ready-to-drink cocktails and hard seltzers while spending $29.9 million lobbying Congress in 2024. Ireland's mandatory cancer warning labels, originally scheduled for May 2026, are now facing industry-orchestrated delays. The U.S. label hasn't moved in 37 years.

Cannabis: The Product Changed Faster Than the Conversation

Cannabis legalization has produced a paradox that the public conversation is not adequately processing. Adolescent use is actually down since legalization began — Monitoring the Future 2024 shows 12th-grade past-year use at 25.8%, below pre-legalization baselines. This is the finding legalization advocates cite, and it is real.

But the product itself has undergone a transformation that makes historical comparisons largely meaningless. THC concentrations in cannabis flower rose from 3–5% in the 1990s to an average of 18–23% today, with premium products at 25–35%. Concentrates — shatter, wax, distillate — now commonly test at 60–90% THC. The plant your parents may have smoked was categorically different from what is legally available at a dispensary in 2026.

The mental health consequences of high-potency use are now well-documented. A landmark 11-site European study published in The Lancet Psychiatry found that daily use of high-potency cannabis was associated with roughly 4.8 times the odds of a first psychotic episode compared to non-use, with researchers estimating that high-potency cannabis accounted for 30% of new psychosis cases in London and 50% in Amsterdam. A 2024 replication in Psychological Medicine confirmed the effect is independent of genetic predisposition. About 30% of regular cannabis users develop some degree of cannabis use disorder — a figure that rises with frequency and potency of use. Cannabinoid hyperemesis syndrome, a condition of severe cyclic vomiting caused by chronic heavy use, was essentially unknown before legalization and is now a routine emergency-department presentation.

What makes this harder to address is the loophole economy that legalization created. The 2018 Farm Bill defined legal hemp as cannabis with less than 0.3% delta-9 THC, inadvertently leaving a range of psychoactive cannabinoids — delta-8, delta-10, THC-O, THCA — entirely unregulated. This produced a gas-station intoxicant market that grew from approximately $200 million in 2020 to roughly $22 billion by 2025, available without age verification in states where recreational cannabis remained illegal. Congress closed that loophole in November 2025, effective November 2026. The products will have been on shelves for seven years before anyone checks ID.

Nicotine's Second Act

The headline statistic about youth vaping is true: the rate fell from 27.5% of middle and high schoolers in 2019 to 5.9% in 2024. This is a real public health achievement, driven by flavor restrictions and increased enforcement. Youth tobacco product use is at a 25-year low.

The footnote is that among the 1.6 million teens still vaping in 2024, daily use nearly doubled — from 15.4% to 28.8% of active users — over the same period. The casual experimenter has left the market. The dependent user remains.

The delivery system has also changed. JUUL, which pioneered nicotine-salt formulations at concentrations three times higher than EU limits, paid over $1.7 billion in settlements to states and class plaintiffs before effectively collapsing as a market force. But its chemistry became the industry standard, and its successors are Chinese-manufactured disposable vapes — Elf Bar, Lost Mary, Geek Bar — that are almost universally unauthorized by the FDA and available at gas stations across the country. Between 2020 and 2022, the FDA found that total nicotine volume sold in tracked retail channels rose 249% even as the number of devices grew only 35%, because each disposable unit now delivers vastly more nicotine per device than cartridge-based predecessors.

The second act is oral nicotine. Philip Morris International paid approximately $16 billion to acquire Swedish Match — maker of Zyn — in 2022. U.S. Zyn shipments hit 223 million cans in a single quarter of 2025, up more than 50% year over year. Nicotine pouches have no tobacco, no smoke, no visible stigma, and a very clean cultural story: Tucker Carlson used them on air; NFL quarterbacks endorse them; TikTok's "Zynfluencer" community has accumulated 1.5 million posts. The product is nicotine delivery optimized for 2026 — invisible, pocket-sized, and socially neutral. Youth use remains modest but is growing. Gen Z's overall nicotine exposure, after decades of decline, is back on the rise.

The Attention Economy as Infrastructure

Social media is the hardest of the five to evaluate because the science is still catching up with the intuition. Neither the DSM-5 nor the ICD-11 recognizes social media addiction as a clinical category. Serious researchers argue the measured effect sizes between screen time and mental health are small and possibly confounded.

That caveat, however, should not obscure what is clearly established. Aza Raskin, who invented infinite scroll, has publicly stated he regrets it. Tristan Harris, a former design ethicist at Google, built a nonprofit — the Center for Humane Technology — around the argument that platforms run a "race to the bottom of the brainstem." The specific mechanics are well-documented: pull-to-refresh mirrors the slot-machine gesture. The notification badge exploits the same variable-reward schedule. Autoplay removes the decision to continue. The For You Page learns your most emotionally vulnerable content categories and serves them at scale.

The Facebook Files, leaked by Frances Haugen in 2021, provided the clearest evidence of corporate knowledge. Internal Meta research found that 32% of teen girls said Instagram made body-image issues worse when they already felt bad about themselves; 13.5% said it made suicidal thoughts worse. A 2023 coalition of 41 state attorneys general sued Meta, citing additional internal documents. Simulated-minor audits by the Center for Countering Digital Hate found TikTok served self-harm content to new teen accounts within 2.6 minutes of signup and eating-disorder content within 8 minutes. Accounts with usernames signaling vulnerability received roughly 12 times more such content.

Teen girls' rates of persistent sadness rose from 36% in 2011 to 57% in 2021, tracking almost precisely with the period when smartphone ownership crossed 50% of adolescents. Whether the phones caused the crisis or merely coincided with it remains genuinely contested among researchers. What is not contested is that the platforms held internal evidence of harm, suppressed it, and continued optimizing for engagement.

One System

The five industries look different on the surface. They occupy different regulatory categories, different cultural spaces, different generational identities. Sports betting is entertainment. Alcohol is social. Cannabis is medicine, or recreation, or both. Nicotine is a habit. Social media is infrastructure. Each has its advocates, its harm-reduction frameworks, its responsibility initiatives.

But strip away the branding and four structural features remain constant across all five. Heavy-user economics: the business model depends on the most harmed consumers. Design for compulsion: variable rewards, frictionless access, and personalized temptation engineered to override self-regulation. Addiction clustering: 73% of lifetime pathological gamblers have alcohol use disorder; adolescent vapers are three times more likely to use combustible tobacco and two to three times more likely to use cannabis; the DSM-5 moved gambling disorder into the substance-use-disorder category in 2013 because the neurobiology is the same. And the tobacco playbook: fund your own responsibility organization, manufacture scientific doubt, shift products when old ones get banned, and invoke personal choice while engineering away the conditions that make choice meaningful.

All of this runs on a substrate the Case-Deaton "deaths of despair" research captured and Surgeon General Murthy's 2023 loneliness advisory quantified: roughly half of American adults report chronic loneliness; in-person time with friends has fallen by two-thirds in two decades; the mortality impact of that loneliness, Murthy wrote, is equivalent to smoking 15 cigarettes a day. An economy organized around engineered compulsion is most profitable in a country where fewer people have somewhere better to put their attention.

What Actually Works

The policy record is clearer than the political conversation suggests. Price intervention works: Scotland's minimum alcohol unit price produced a 13.4% reduction in alcohol-specific deaths within five years, with the largest gains among the poorest quintile. Availability and design restrictions work: Massachusetts's flavored tobacco ban cut cigarette sales 17% in year one; California's e-cigarette flavor ban reduced sales 37% in 18 months; the UK's gambling advertising restrictions during live sports reduced pre-watershed betting ads seen by children by 97%. Warning and disclosure work: graphic tobacco warning labels reduced smoking in every country that implemented them.

What has not worked is voluntary industry self-regulation. The gambling industry's responsible gaming programs, alcohol's "drink responsibly" campaigns, tobacco's nicotine-pouch harm-reduction messaging, and Meta's internal Well-Being team have generated better press releases than outcomes. The business model of each sector requires heavy users. Voluntary programs aimed at reducing heavy use are structurally self-defeating.

The most underused interventions in the United States are excise taxes indexed to inflation, mandatory design standards — nicotine concentration caps, THC potency limits, sportsbook loss limits, default-protective algorithmic settings for minors — and serious federal investment in treatment, which currently reaches fewer than 2% of Americans with alcohol use disorder and a fraction of those with gambling or nicotine disorders.

The Deeper Argument

The standard case against addictive industries is made in the language of public health: deaths, dollars, hospitalizations, lost productivity. That case is true and necessary. But it understates what is actually at stake.

Human agency is not a fixed attribute. It is a capacity that requires conditions — a built environment that does not weaponize your psychology against you, products that are not optimized to override your self-regulation, a social context where the most isolated and vulnerable are not the most profitable demographic. Across all five sectors examined here, those conditions have been systematically dismantled.

The sportsbook's algorithm knows when its VIP customer has had a losing week and times its promotional push accordingly. The platform's recommender system knows which teenager's feed is drifting toward disordered eating content and amplifies it because amplification is engagement. The disposable vape is tuned to deliver nicotine at concentrations that bypass the throat irritation that once limited how much a person could smoke. These are not accidents or oversights. They are product decisions, made by people who understood what they were doing.

The question is not whether to act. The evidence on what works is sufficient. The question is whether the political economy of a country where these industries collectively spent $140 million lobbying Congress in 2024 can generate the institutional will to act before the next engineered compulsion — whatever form it takes — reaches the market.

Ryan is 27. He is not failing at willpower. He is losing a fight he didn't know was designed.