In July 1973, the Drug Enforcement Administration opened its doors with a budget of $74.9 million and roughly 1,470 agents. By fiscal year 2024, that budget had grown to $2.61 billion. In 2022, 107,941 Americans died of drug overdoses — the highest annual total ever recorded. More than two-thirds involved synthetic opioids other than methadone — the CDC’s category for fentanyl and fentanyl analogs.
Do the arithmetic. The enforcement apparatus grew thirtyfold. The stated objective — reducing drug-related death — reached its worst outcome in American history. Frame that as failure and you’re still asking the wrong question. The better one: what, exactly, did fifty years of escalating enforcement actually build?
This is an economics argument. The mechanism isn’t peculiar to any era or any drug. It operates wherever prohibition encounters inelastic demand — and it has produced the same institutional output twice in American history: once with the Mafia, once with the cartels. The organisations are different. The mechanism is identical.
The price floor
Start with the demand curve, because that’s where the argument lives.
The Australian Institute of Criminology’s Trends and Issues in Crime and Criminal Justice No. 606, a systematic review of 42 studies covering 462 elasticity estimates, found that on average, a ten percent increase in the price of illicit drugs reduces demand by approximately nine percent. Technically inelastic — but barely. And the figure conceals something important. The heaviest users, who account for the largest share of total revenue, are the least price-responsive of anyone in the market. Addiction is, among other things, an economic relationship with an extremely steep demand curve.
Now add prohibition. What it does, structurally, is set a floor below which no legal supplier can operate. No legitimate entity can absorb the criminal liability. The penalty for participation isn’t a licensing fee or a fine — it’s prison. That liability isn’t a bug; it’s the point. But the effect is to hand the market to the only organisations structured to absorb that particular cost: criminal enterprises.
Not a moral observation. A market structure observation. Criminal organisations don’t dominate prohibited markets because criminals are especially good at business. They dominate because the legal apparatus has eliminated every competitor. Prohibition creates the monopoly. And monopolies insulated from competition by the threat of incarceration for any would-be rival are extraordinarily profitable.
Risk premium becomes supernormal profit. Supernormal profit buys institutional depth: lawyers, accountants, political contacts, logistics networks, territorial enforcement, interorganisational diplomacy. Enforcement keeps competitors out, makes incumbents more profitable, and finances the architecture of permanence.
In 1986, journalist and drug policy analyst Richard Cowan named what he called the Iron Law of Prohibition: as enforcement intensity increases, so does the potency of prohibited substances. More potent substances deliver more value per unit of volume, weight, and legal risk. Beer gave way to spirits during alcohol Prohibition. Bulky opium gave way to refined heroin. The logic is straightforward — if you’re going to risk a felony charge per shipment, you want as much value packed into that shipment as possible.
One kilogram of fentanyl contains approximately 500,000 lethal doses. One kilogram of heroin contains roughly 5,000. That hundred-to-one potency ratio isn’t a pharmacological accident. It’s the Iron Law in preview — and we’ll return to it at full scale.
What prohibition built
Before 1920, American organised crime was fragmented, regional, and opportunistic. Protection rackets, gambling, petty extortion. No institutional continuity across geography or time. Nothing resembling a permanent organisation.
On January 17, 1920, the Eighteenth Amendment came into effect, creating the largest illegal market the country had ever seen — for alcohol, with roughly 130 million people who had been drinking legally the week before. Demand didn’t disappear. It just lost its legal supply chain.
What happened next wasn’t spontaneous. It was structural. Federal investigators estimated that by 1927, the Capone Outfit was generating approximately $60 million annually from bootlegging alone — total revenues from bootlegging, gambling, prostitution, and other enterprises approaching $100 million, roughly $1.7 billion in today’s money. Government investigation estimates, not audited accounts. But the orders of magnitude aren’t seriously disputed.
What that revenue built was not just wealth. It built institutions. Brewmasters and truckers. Lawyers who understood how to corrupt grand juries. Accountants who understood how to hide money. Police and judges who understood the value of being on the right payroll. Distribution networks that crossed state lines. Territorial agreements enforced by violence when necessary and by negotiation when possible.
In May 1929, three months after the St. Valentine’s Day Massacre, Capone convened the Atlantic City Conference — gathering gang leaders from across the country to negotiate territorial boundaries, resolve disputes, and reduce the kind of public bloodshed that attracted federal attention. It was not a criminal gathering. It was an industry summit. The agenda included division of territory, importation logistics, and conflict resolution mechanisms. The participants were not thugs meeting in an alley; they were the executives of a distributed national industry, meeting in a hotel to conduct business.
Prohibition ended December 5, 1933. The Mafia did not.
That’s the structural point that matters most. The institutional architecture built by twelve years of prohibited alcohol — the corruption networks, the money laundering expertise, the territorial control mechanisms, the supply chain management, the diplomatic infrastructure for inter-organisational relations — didn’t dissolve when Repeal happened. It repurposed. Gambling. Loan sharking. Labour racketeering. Eventually, narcotics. The government repealed a law. The institution that law had financed was permanent.
On the violence question: the US murder rate peaked at 9.7 per 100,000 in 1933 — the final year of Prohibition — and fell as legal supply resumed. Chicago is the cleaner specific case. A 2009 study published in the journal Addiction, drawing on data from the Chicago Historical Homicide Project tracking 11,018 homicides between 1870 and 1930, found that total homicide rates in Chicago increased by 21% during Prohibition. The causal relationship between prohibition and violence is contested in the broader literature — some economists find the evidence inconclusive — but the Chicago data and the national peak-in-1933 pattern are both consistent with it.
The question isn’t whether Prohibition caused every murder. The question is what kind of organisations it created, and what happened to those organisations after 1933. They survived. They diversified. The enforcement regime had built them. Repeal couldn’t unmake them.
The cartel factory
On June 17, 1971, President Nixon declared drugs “public enemy number one.” Two years later, the DEA was founded. Since then, the United States has spent an estimated $1 trillion on the war on drugs, encompassing federal operations, state and local enforcement, border interdiction, and international military and intelligence support. The global drug trade is larger now than it was when Nixon spoke.
That doesn’t mean enforcement accomplished nothing. It means enforcement accomplished something other than what it claimed to be accomplishing.
The Colombian cocaine trade of the 1980s is the clearest illustration. The Medellín Cartel, under Pablo Escobar, reached industrial scale precisely during the most intensive enforcement period in DEA history. The cartel didn’t grow despite enforcement; enforcement, by suppressing competition and maintaining premium prices, helped finance its growth. When Colombian and US forces killed Escobar on December 2, 1993, and subsequently dismantled the Medellín and Cali cartels, the response of the global cocaine market was not contraction. It was multiplication and geographic shift.
Analysts call it the balloon effect. Press the drug trade out of one location, it expands in another. Destroy one organisation, and the market it served doesn’t disappear — it gets served by the organisations that survive. Mexican cartels — the Sinaloa, the Gulf, the Zetas, the Jalisco New Generation Cartel — stepped into the gap. But they were not simply drug traffickers. They were diversified criminal conglomerates for whom drug trafficking was the original profit engine. Human trafficking. Fuel theft. Illegal mining. Extortion at industrial scale. The revenue streams the American Mafia developed after Prohibition ended, the Mexican cartels developed while the drug trade was still running.
A caveat worth carrying.
The balloon effect: what the evidence actually shows The displacement of drug supply from one geography to another following interdiction is well-documented in UNODC reports — the shift from Caribbean routes to the US-Mexico border corridor following 1990s interdiction operations is a clear historical case. But whether displacement increases total drug volume reaching users, or merely relocates it, remains contested among economists. Peter Reuter at RAND and colleagues have argued the net effect on total drug volume reaching users is often modest. Others find measurable supply reductions. The honest position is that enforcement displaces supply reliably; whether it reduces it meaningfully in aggregate is a harder question the evidence hasn't settled.
Now add the Iron Law. Through the 2000s and into the 2010s, DEA pressure on prescription opioids — aggressive prosecution of pill mills, rescheduling of certain opioids, pharmaceutical company settlements — achieved a genuine supply reduction in diverted prescription drugs. The population addicted to opioids didn’t stop being addicted. It switched to heroin. Then enforcement increased heroin interdiction pressure. The cartels, responding to the economics as rational actors, switched to fentanyl.
Why fentanyl? The economics are explicit. It can be synthesised from chemical precursors — no agricultural inputs, no coca or poppy cultivation zones to interdict, no dependence on weather or geography. It’s more compact per unit of effect, easier to conceal, cheaper to produce, more profitable per shipment. And it is, per unit of effect, approximately one hundred times more lethal than heroin.
In 2022, 107,941 Americans died of drug overdoses. The DEA’s budget hit $2.61 billion. Both figures were, in their respective categories, all-time records.
The DEA’s enforcement pressure on heroin was, by its own metrics, a genuine operational success. It contributed to a real reduction in heroin availability. It replaced a moderately lethal drug with an extremely lethal one. Not an ironic accident. The Iron Law operating exactly as predicted — at a scale Cowan couldn’t have anticipated when he wrote in 1986 about crack cocaine.
The market learns
In 2011, Ross Ulbricht launched Silk Road on the Tor network, running transactions through Bitcoin escrow. What he built wasn’t merely a marketplace for illegal drugs. It was a full retail experience — vendor ratings, customer reviews, dispute resolution, return policies. The user experience was, in several respects, better than buying drugs from a street dealer. When the FBI shut Silk Road down in October 2013, Ulbricht was later ordered to forfeit $183 million — the court-calculated value of Silk Road’s total sales across its 2.5-year operation, priced at bitcoin’s exchange rates at the time of each transaction.
Within twelve months: more darknet markets were operating than had existed before the closure. More product listings. More vendors. Better operational security. More distributed infrastructure. A Swansea University Global Drug Policy Observatory analysis published in January 2015 documented the expansion directly. A 2016 RAND Europe report found that transactions for illicit drugs on cryptomarkets had tripled and revenues had doubled since Silk Road’s closure.
Same pattern as Escobar. Destroy the incumbent, and the market doesn’t contract — it fragments and adapts.
But there’s a structural difference here that matters. Physical drug markets require geographic dominance. Geographic dominance requires territorial control. Territorial control requires violence to establish and maintain — corner by corner, block by block. That’s where drug-market violence comes from: not addiction, but competition for retail geography in a market that can’t use courts to settle disputes.
Online markets break that dependency. A vendor holds no territory. Requires no street presence. The enforcement tools that work against physical markets — seize the corner, arrest the dealer, map the network — are largely inapplicable. You can’t raid a Tor node.
Do darknet markets reduce violence? Some researchers have argued that disaggregating retail from territorial control should reduce the violence historically associated with corner competition in drug markets. The theoretical argument is coherent, and there's preliminary empirical support in the literature. But the evidence base is developing and causal mechanisms remain contested. Whether darknet markets measurably reduce street-level violence is a genuine empirical question that cannot be answered with confidence from existing data.
The technical layer adapted too. Bitcoin transactions are recorded on a public blockchain. Investigators followed the cryptocurrency trails to identify and prosecute Silk Road vendors. The market’s response was migration toward Monero — a privacy-native cryptocurrency whose transactions cannot be traced with current forensic tools. Enforcement closed one investigative vector. The market opened another.
The market doesn’t contract under enforcement pressure. It re-architectures.
The adaptive logic runs on the same fuel it always has: the profit margin that makes reorganisation worthwhile exists only because prohibition put it there. Better surveillance, improved blockchain forensics, more aggressive international prosecution — none of these change the underlying economics. The question is not whether enforcement can eventually outpace the market’s adaptation. The question is whether the premise was ever sound: that applying criminal liability to goods with inelastic demand would shrink those markets rather than transfer them to whoever was willing to absorb that liability. One government, in 2001, decided to act on what the economics actually implied.
The counter-experiment and the structural problem
In July 2001, Portugal decriminalised personal possession of all illicit drugs. Not legalisation — supply chains remained illegal and criminal enterprises continued to operate them. What changed was the treatment of users: possession became an administrative offence handled by Dissuasion Commissions staffed by social workers, legal professionals, and psychologists. The legal reform was paired with a simultaneous, large-scale expansion of treatment capacity, harm reduction services, and social reintegration programmes. These two elements — the legal change and the treatment investment — cannot be analytically separated. One without the other would likely have produced different results.
The outcomes, drawn from EMCDDA and ECDC monitoring data, are not ambiguous. Drug-related death rates fell from approximately 80 deaths per million in 2001 to 6 per million by 2019 — a reduction of more than ninety percent. AIDS diagnoses among people who inject drugs fell from 518 in 2000 to 13 in 2019. The number of people imprisoned for drug offences fell from 3,863 in 1999 to 1,140 in 2017. HIV infection rates from drug-related transmission dropped from 104.2 per million in 2000 to 4.2 per million by 2015. These are not contested figures. The outcomes of treating drug use as a public health problem rather than a criminal one are not a secret.
But here’s what Portugal did not do: it did not legalise supply. Criminal supply chains continued. The profit margin available to criminal organisations from supplying drugs to Portuguese users was not touched by decriminalisation. Portugal demonstrates the floor of what an alternative regime produces — the minimum achievable improvement when you stop punishing users. It is not the ceiling. The criminal supply problem is structurally intact.
The limits of the Portugal comparison Portugal's conditions in 2001 were specific in ways that constrain direct transferability. A heroin epidemic of catastrophic scale had generated genuine cross-party political consensus that the existing approach wasn't working. Portugal is a small, relatively cohesive society where a single national policy change could be uniformly implemented. The Catholic cultural context allowed harm reduction to be framed as compassion rather than permissiveness. Whether these conditions are replicable in the United States, the United Kingdom, or other large anglophone democracies with more fragmented political systems and stronger prohibitionist constituencies is a genuine constraint on any simple "Portugal proves the alternative works" argument. Portugal proves the alternative produces better health outcomes. It doesn't prove political feasibility elsewhere.
So why doesn’t the alternative spread?
Not ignorance. Most health ministries know what EMCDDA publishes. The answer is that the status quo has organised constituencies, and the alternative does not.
In August 2016, Insys Therapeutics — a pharmaceutical company that had made billions selling Subsys, a fentanyl-based painkiller, and was at the time under federal and state investigation for improper marketing practices — donated $500,000 to Arizonans for Responsible Drug Policy, the campaign opposing Proposition 205, which would have legalised recreational cannabis in Arizona. Insys was simultaneously seeking DEA approval for Syndros, a synthetic THC pharmaceutical. The company’s commercial logic was simultaneous and on record: keep the natural product illegal, sell the patented synthetic version. The $500,000 represented, at the time of campaign finance disclosure, more than a third of all money raised by the anti-legalisation campaign.
A peer-reviewed study published in PLOS One, examining 556 pharmaceutical companies between 1996 and 2019, found that stock market returns were 1.5-2% lower in the ten days following a cannabis legalisation event, with investors anticipating a reduction in annual sales of approximately $3 billion per legalisation event. Pharmaceutical companies don’t merely dislike legalisation politically. They pay for its defeat.
The alcohol industry’s alignment is documented in campaign finance records. In Massachusetts in 2016, the Beer Distributors PAC contributed $25,000 to the campaign opposing recreational marijuana legalisation; the combined donation from Beer Distributors of Massachusetts and Wine & Spirits Wholesalers of Massachusetts to the anti-legalisation campaign totalled $75,000. In Arizona, the Arizona Wine and Spirits Wholesale Association donated $10,000 to the opposing campaign. These are not large sums in absolute terms. The logic they reveal is the same logic CoreCivic disclosed to its shareholders in SEC filings — that reform represents a direct financial risk. Campaign finance records don’t require inference.
CoreCivic, one of the two largest private prison operators in the United States, stated explicitly in its SEC filings that “the demand for our facilities and services could be adversely affected by the relaxation of enforcement efforts, leniency in conviction or parole standards and sentencing practices or through the decriminalization of certain activities that are currently proscribed by our criminal laws.” This is the company telling its shareholders what the drug war is worth to it.
Then there is the DEA itself. The agency has consistently and documentably opposed cannabis rescheduling — most recently, evidence submitted in rescheduling proceedings was significant enough that cannabis industry groups argued the DEA should be removed from the proceedings on grounds of demonstrated bias. This isn’t corruption. It’s the rational behaviour of a bureaucratic institution whose $2.61 billion annual budget is a War on Drugs budget. An end to that war would require a reckoning with that budget. Bureaucracies don’t will their own obsolescence.
This is exactly what Mancur Olson described in The Logic of Collective Action in 1965. When the costs of a policy are concentrated among a well-organised minority with strong financial incentives to lobby, and the benefits of reform are diffuse, delayed, and distributed across a broad population with weak incentives to organise, the concentrated minority wins. Drug prohibition isn’t an exception to this dynamic. It’s one of its clearest examples.
“Soft on drugs” remains an effective electoral attack because it concentrates political cost on reformers. The public health case for reform is diffuse, delayed, and held by people who don’t typically write large cheques to campaign committees.
Criminal organisations don’t lobby for the status quo. They don’t have to. The legal economy has taken care of it.
The closing arithmetic
DEA budget: $74.9 million in 1973. $2.61 billion in 2024. Drug overdose deaths: 107,941 in 2022 — the worst year on record. The enforcement apparatus grew thirtyfold. The stated objective deteriorated to its worst-ever outcome.
The harder claim, grounded in everything established above: the enforcement budget didn’t merely fail to stop criminal industrialisation. In a specific, structural sense, it drove it. By maintaining the price floor that makes criminal drug supply more profitable than almost any legal business on earth. By stimulating, through the Iron Law, the potency escalation that replaced heroin with fentanyl. By funding the institutional appetite of agencies and industries whose financial interests require the continuation of the problem they were created to solve.
None of this is hidden. All of it is documented.
The most powerful constituency for keeping drug markets criminalised is the legal economy — pharmaceutical, alcohol, private prison, law enforcement — that has organised itself around prohibition’s continuation. Criminal organisations don’t lobby for the status quo.
They don’t have to.
Gen AI Disclaimer
Some contents of this page were generated and/or edited with the help of a Generative AI.
Media
Key Sources and References
Payne, Jason, Manning, Matthew, Fleming, Christopher, and Pham, Hien-Thuc. “The price elasticity of demand for illicit drugs: A systematic review.” Australian Institute of Criminology, Trends and Issues in Crime and Criminal Justice No. 606, 2020. https://www.aic.gov.au/publications/tandi/tandi606
Drug Enforcement Administration. “Staffing and Budget.” DEA.gov. https://www.dea.gov/data-and-statistics/staffing-and-budget
Drug Enforcement Administration. “Budget Hearing — Fiscal Year 2024 Request for the Drug Enforcement Administration.” April 27, 2023. https://www.dea.gov/documents/2023/2023-04/2023-04-27/budget-hearing-fiscal-year-2024-request-for-drug-enforcement
Centers for Disease Control and Prevention, National Center for Health Statistics. “Drug Overdose Deaths in the United States, 2002–2022.” NCHS Data Brief No. 491, 2023. https://www.cdc.gov/nchs/products/databriefs/db491.htm
Asbridge, Mark and Weerasinghe, Swarna. “Homicide in Chicago from 1890 to 1930: prohibition and its impact on alcohol- and non-alcohol-related homicides.” Addiction, Vol. 104, No. 3, March 2009. https://pubmed.ncbi.nlm.nih.gov/19207343/
Thornton, Mark. “Alcohol Prohibition Was a Failure.” Cato Institute Policy Analysis No. 157, July 17, 1991. https://www.cato.org/sites/cato.org/files/pubs/pdf/pa157.pdf
Cowan, Richard. “How the Narcs Created Crack.” National Review, December 5, 1986.
United Nations Office on Drugs and Crime. “Drug Trafficking.” UNODC. https://www.unodc.org/unodc/en/drug-trafficking/index.html
Drug Enforcement Administration. “Ross Ulbricht, aka ‘Dread Pirate Roberts,’ Sentenced in Manhattan Federal Court to Life in Prison.” DEA Press Release, May 29, 2015. https://www.dea.gov/press-releases/2015/05/29/ross-ulbricht-aka-dread-pirate-roberts-sentenced-manhattan-federal-court
Aldridge, Judith and Decary-Hetu, David. “A Year Since the Closure of Silk Road: Snapshot of Evolved Dark Net Drug Markets.” GDPO Situation Analysis, Swansea University Global Drug Policy Observatory, January 2015. https://www.swansea.ac.uk/media/A-year-since-the-closure-of-Silk-Road-Snapshot-of-evolved-Dark-Net-drug-markets.pdf
RAND Europe. “Untangling the ‘Dark Web’: New Study Shows Steady Growth of the Online Illicit Drugs Trade.” RAND Europe press release, August 8, 2016. https://www.rand.org/news/press/2016/08/08.html
European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). “Drug Policy Profiles — Portugal.” European Union Drugs Agency, 2011. https://www.euda.europa.eu/system/files/publications/642/PolicyProfile_Portugal_WEB_Final_289201.pdf
Transform Drug Policy Foundation. “Drug decriminalisation in Portugal: setting the record straight.” TransformDrugs.org. https://transformdrugs.org/blog/drug-decriminalisation-in-portugal-setting-the-record-straight
Greenwald, Glenn. “Drug Decriminalization in Portugal: Lessons for Creating Fair and Successful Drug Policies.” Cato Institute White Paper, 2009. https://www.cato.org/sites/cato.org/files/pubs/pdf/greenwald_whitepaper.pdf
Arizona campaign finance records. Insys Therapeutics donation to Arizonans for Responsible Drug Policy, August 31, 2016. Reported in: “A maker of deadly painkillers is bankrolling the opposition to legal marijuana in Arizona.” Washington Post, September 9, 2016. https://www.washingtonpost.com/news/wonk/wp/2016/09/09/a-maker-of-deadly-painkillers-is-bankrolling-the-opposition-to-legal-marijuana-in-arizona/
Bednarek, Ziemowit, Doremus, Jacqueline M., and Stith, Sarah S. “U.S. cannabis laws projected to cost generic and brand pharmaceutical firms billions.” PLOS One, August 31, 2022. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0272492
CoreCivic, Inc. Annual Report (Form 10-K), Fiscal Year 2016. U.S. Securities and Exchange Commission EDGAR. https://www.sec.gov/Archives/edgar/data/1070985/000119312517053982/d310578d10k.htm
Marijuana Moment. “New Evidence Shows DEA Opposes Marijuana Rescheduling And Should Be Removed From Hearings, Cannabis Groups Say.” Marijuana Moment. https://www.marijuanamoment.net/new-evidence-shows-dea-opposes-marijuana-rescheduling-and-should-be-removed-from-hearings-cannabis-groups-say/
Olson, Mancur. The Logic of Collective Action: Public Goods and the Theory of Groups. Harvard University Press, 1965.
The Intercept. “Alcohol Industry Bankrolls Fight Against Legal Pot in Battle of the Buzz.” September 14, 2016. https://theintercept.com/2016/09/14/beer-pot-ballot/
Drug Enforcement Administration. “Facts about Fentanyl.” DEA.gov. https://www.dea.gov/resources/facts-about-fentanyl
CNBC. “America has spent over a trillion dollars fighting the war on drugs. 50 years later, drug use in the U.S. is climbing again.” June 17, 2021. https://www.cnbc.com/2021/06/17/the-us-has-spent-over-a-trillion-dollars-fighting-war-on-drugs.html
Prohibition: An Interactive History / The Mob Museum. “Prohibition Profits Transformed the Mob.” https://prohibition.themobmuseum.org/the-history/the-rise-of-organized-crime/the-mob-during-prohibition/
English 
Español 
Français 
Deutsch 
日本語 
Polski