DRUG LEGALIZATION SERIES
PART 17
If you’re new to the series, start here.
Case Studies Without Cherry-Picking
What worked, what failed, and why.
Executive Summary
The easiest way to write about drug policy case studies is to cheat.
If you love reform, you point to Portugal and pretend the argument is over. If you hate reform, you point to Oregon and pretend the argument is over. If you want a miracle story, you point to British Columbia safer supply. If you want a horror story, you point to the exact same place.
That is lazy analysis.
Real drug policy reform is messier than that. Different models solve different problems. Some reduce punishment. Some reduce overdose risk. Some improve access to care. Some tighten public safety and accountability. Some fail because the policy idea was bad. Others fail because the execution was weak.
That is the point of this chapter.
Part 1.5 on legalization vs. decriminalization vs. regulation, Part 9 on risk-based drug regulation, Part 10 on accountability, and Part 14 on the regulated pharmacy model all pointed toward the same basic claim: serious reform has to separate risks, keep hard rules where risk is high, and stop pretending one policy fits every substance.
Part 16 then argued for pilot, measure, scale. This chapter pressure-tests that framework against real-world drug policy case studies.
The bottom line is blunt:
• Portugal drug decriminalization worked better than critics predicted, but it is not legalization and it is not magic.
• Oregon Measure 110 results show that a good instinct can still fail politically when governance, measurement, and treatment on demand are weak.
• British Columbia safer supply shows why overdose prevention can justify high-risk interventions, but also why diversion control and tighter supervision matter.
• Switzerland heroin-assisted treatment is the strongest real-world proof that a narrow, medically managed high-risk lane can improve outcomes.
• Cannabis legalization outcomes show real gains, real tax revenue, and real problems when commercialization gets loose.
So no, these case studies do not prove every form of drug legalization works.
They prove something more useful: reform works better when it is honest about risk, tied to measurable outcomes, and built with public safety, accountability, and real exits into care.
Why Drug Policy Case Studies Usually Get Cherry-Picked
Most drug policy case studies get abused the same way.
People grab the one example that flatters their ideology and ignore the rest.
That is how you end up with fake certainty.
The honest approach is harder. You have to ask:
• What exactly was the policy?
• What problem was it trying to solve?
• What infrastructure sat underneath it?
• What got better?
• What stayed bad?
• What got worse?
• What lesson actually transfers to another place?
That matters because Portugal drug decriminalization is not the same thing as state cannabis retail. British Columbia safer supply is not the same thing as a broad consumer market. Switzerland heroin-assisted treatment is not the same thing as legal street sales. And Oregon Measure 110 results tell you as much about rollout and system weakness as they do about decriminalization itself.[5][6]
So the goal here is not to find a hero country or a villain state.
The goal is to look at real drug policy case studies and ask what they actually teach about risk-based drug regulation, the regulated pharmacy model, treatment on demand, diversion control, overdose prevention, public safety, and accountability.
Portugal Drug Decriminalization: Better Than the Hype and Better Than the Fear
Portugal drug decriminalization is the case study everybody cites and almost nobody describes accurately.
Portugal did not legalize drugs. It decriminalized possession for personal use and routed those cases through Commissions for the Dissuasion of Drug Addiction under the health system. Trafficking remained illegal. Official Portuguese materials describe dissuasion as a networked strategy meant to reduce use, protect health, prevent exclusion, and bring users into contact with services, often for the first time.[1][2]
That distinction matters, and it is exactly the distinction Part 1.5 already made.
What worked in Portugal drug decriminalization?
• It differentiated users from dealers.
• It moved personal-use cases away from the criminal pipeline.
• It tied policy to a health-oriented response instead of pure punishment.
• It did not trigger the collapse that prohibition hardliners predicted.[1][3][4]
What gets exaggerated?
A lot.
Portugal drug decriminalization is often sold like a one-line policy miracle. It was not. It sat inside a broader shift that involved treatment, prevention, harm reduction, and administrative follow-up. A literature review on Portugal noted that concerns related to heroin declined, but the broader evidence on care impacts remained limited and the policy still operated alongside stronger punishment for dealers and softer punishment for users.[3] A separate empirical study found that decriminalization did not appear to lower illicit drug prices in a way that would suggest a simple surge in access or dependence.[4]
That is the real lesson.
Portugal drug decriminalization worked best as a way to change who gets punished and who gets routed toward help. It did not solve toxic supply by itself. It did not make high-risk drugs safe. And it did not eliminate the need for public safety enforcement against trafficking.
So what transfers?
Portugal tells you that possession does not need to be the center of drug law. It tells you that health-oriented processing can be more rational than criminal processing. But it also tells you that decriminalization is a floor, not a ceiling. It is one useful tool inside drug policy reform. It is not the whole structure.
Oregon Measure 110 Results: Big Idea, Weak Execution
If Portugal gets romanticized, Oregon Measure 110 results get flattened in the opposite direction.
The lazy version says Oregon “proved” decriminalization fails.
That is not serious either.
Oregon’s model did something real. The state’s Behavioral Health Resource Network program says Measure 110 funds tribes and organizations in every county and requires six service areas: screenings, behavioral health assessments, peer support, harm reduction, low-barrier treatment, and transitional or supportive housing.[5] That is not nothing. It is real capacity on paper, real money, and real effort.
But the hardest truth about Oregon Measure 110 results is that a public-health vision still needs execution discipline.
The 2025 Oregon Secretary of State audit was blunt. It said the vision of replacing criminalization with a public health approach remained unfulfilled because of structural and operational weaknesses. The audit found weak integration with the broader behavioral health system, flawed data, unclear goals, insufficient information to determine outcomes, and inconsistent county deflection programs. It also noted that possession was recriminalized in September 2024.[6]
That is devastating, and it matters.
What worked?
• Oregon moved money toward services instead of pure punishment.
• It created statewide networks intended to support treatment, housing, and harm reduction.[5]
• It tried to build an alternative to arrest-first policy.
What failed?
• The system was not strategically integrated.
• The state could not clearly show who was served and what outcomes changed.
• Treatment on demand was not built in a way the public could see and trust.
• Accountability and measurement were too weak to protect the policy politically.[6]
So the lesson from Oregon Measure 110 results is not “reform bad.”
The lesson is harsher and more useful: decriminalization without strong governance, visible performance metrics, integrated care pathways, and measurable accountability gets destroyed fast.
That is why Part 16 mattered. The answer was never “announce reform and hope.” The answer was pilot, measure, scale.
And that is why Oregon Measure 110 results should be read as a warning against sloppy rollout, not as a commandment to go back to the same failed punishment model.
British Columbia Safer Supply: Real Overdose Prevention, Real Warning Signs
British Columbia safer supply is probably the most ideologically fought-over case in modern drug policy.
Health Canada defines safer supply as prescribed medications offered as a safer alternative to the toxic illegal drug supply for people at high risk of overdose. The official rationale is straightforward: prevent overdoses, save lives, and connect people to health and social services.[7]
That is the strongest argument for British Columbia safer supply.
It recognizes a brutal fact: when the illegal supply is poisoned and unstable, some people are going to die while waiting for the perfect treatment journey. In that situation, overdose prevention itself becomes a moral and policy priority.
But the other side of the case matters too.
A 2024 JAMA Internal Medicine study found that British Columbia’s safer opioid supply policy was associated with significant increases in opioid prescriptions and claimants, but it also found a significant increase in opioid-poisoning hospitalizations and no statistically significant change in deaths from apparent opioid toxicity during the study period.[9]
That does not prove the entire model failed. It does prove the evidence is mixed and the design details matter.
And the policy system itself has reacted to that reality. British Columbia now requires all prescribed alternatives to be consumed under direct supervision of a health professional, with limited exceptions, specifically to help prevent diversion to illicit markets.[8]
That is the key point.
British Columbia safer supply is not a clean triumph story and not a clean disaster story. It is a warning against magical thinking from both sides.
What worked?
• It addressed the toxic-supply problem directly.
• It targeted people at very high overdose risk.
• It created an alternative to the deadliest part of the illegal market.[7]
What raised real concerns?
• The evidence on hospitalizations and deaths is not cleanly reassuring.[9]
• Flexibility around carrying and dosing created diversion control concerns serious enough that the province later tightened the rules.[8]
• The model can drift into ideological symbolism if it is not tied to hard auditing and clinical structure.
So the lesson from British Columbia safer supply is not “never do it” and not “scale it blindly.”
The lesson is that the highest-risk lane has to be narrow, supervised, measurable, and tied to treatment on demand. That is exactly where the regulated pharmacy model becomes useful. The high-risk lane cannot run on vibes. It needs rules.
Switzerland Heroin-Assisted Treatment: The Narrow High-Risk Lane That Actually Worked
If you want the strongest case for a tightly controlled medical lane, it is Switzerland heroin-assisted treatment.
This is one of the most important drug policy case studies in the entire debate because it is not broad legalization, not casual decriminalization, and not loose community access. It is a deliberately narrow program for people with severe opioid use disorder.
And it produced real outcomes.
A 2022 review of randomized controlled trials found that heroin-assisted treatment reduced criminal activity across all trials, with several trials finding larger reductions than comparison treatments. The review also noted that HAT was often seen as having a stronger benefit-cost profile than oral methadone because of its effect on crime reduction.[10]
A 2025 nationwide Swiss survey adds something else important: structure. Switzerland’s system is limited to specialized treatment centers. The survey reported 22 operating centers, psychiatry staffing everywhere, psychotherapy in most centers, social workers in nearly all centers, a mean referral-to-initiation time of 10 days, and very rare overdoses and seizures.[11]
That is the part American debates keep skipping.
Switzerland heroin-assisted treatment works inside a narrow, specialized, high-accountability lane. It is not open retail. It is not liberalized chaos. It is a medical program with hard boundaries, staff, monitoring, and fast entry for a highly selected population.
That makes Switzerland heroin-assisted treatment the closest real-world proof of concept to the regulated pharmacy model argued for in Part 14.
What worked?
• It targeted the highest-risk population instead of pretending one lane fits everyone.
• It reduced crime and stabilized care for people who had not done well under standard treatment.[10]
• It built medical structure around access rather than denying access and hoping the street market would teach restraint.[11]
What should not be copied lazily?
• It depends on specialized centers, not generic rollout.
• It works because of tight eligibility, close monitoring, and integrated support.
• It is a high-structure model, not a permissive one.
That is the transferable lesson. When risk is highest, the lane has to get tighter, not looser.
Cannabis Legalization Outcomes: Real Gains, Real Costs
The conversation around cannabis legalization outcomes is usually dumb in a different way.
One side pretends legalization solved everything because stores are licensed and tax revenue exists. The other side pretends any increase in use proves legalization should never have happened.
Neither side is serious.
The most honest reading of cannabis legalization outcomes is that lower-risk adult substances can be moved into legal, age-gated, tested channels, but commercialization creates new problems that still need hard rules.
There are real gains.
Colorado reported in 2025 that marijuana sales had generated more than $3 billion in tax and fee revenue since legalization.[15] Legal channels also replace some illegal sales with tested products, age gates, packaging rules, and visible regulation.
But the costs are real too.
A 2025 JAMA Network Open study found that state recreational cannabis legalization was associated with a 3.28 percentage-point increase in 30-day cannabis use over five years, with a larger increase after retail outlets opened.[12] A 2024 meta-analysis found that recreational legalization was associated with modest but significant increases in past-month youth cannabis use, while medical legalization was not significantly associated with change in past-month youth use.[13] And a 2024 systematic review of traffic injuries found mixed findings: four studies reported increases after legalization, while three reported no significant change.[14]
That is already enough to kill both slogans.
But there is another warning here. A 2025 JAMA Network Open study found that legalizing youth-friendly cannabis edibles and extracts in Canada was associated with a 26% increase in past-year cannabis use and a 43% increase in edible use among adolescents in the provinces that allowed those products.[16]
That is not a reason to go back to prohibition. It is a reason to stop pretending product design, potency, packaging, and marketing do not matter.
So what do cannabis legalization outcomes actually teach?
• Legalization can create regulated channels and real revenue.[15]
• Adult use can go up.[12]
• Youth effects are not zero and are sensitive to policy design.[13][16]
• Public safety issues like impaired driving still require enforcement and better measurement.[14]
• Commercialization can push a lower-risk substance in a worse direction if the rules are weak.
That is why cannabis legalization outcomes reinforce, not weaken, the case for risk-based drug regulation. Lower-risk does not mean zero-risk. It means the lane can be looser than heroin, fentanyl, or a high-risk opioid model, while still being much tighter than a free-for-all.
What These Drug Policy Case Studies Actually Prove
If you strip away the ideology, the real pattern across these drug policy case studies is not complicated.
What worked across the better models
• differentiating users from dealers
• moving the highest-risk population into more structured care
• making overdose prevention a real policy goal
• building treatment on demand instead of referral theater
• using tighter rules as risk rises
• keeping public safety enforcement focused on behavior that harms other people
• making accountability visible
What failed across the weaker models
• vague goals
• poor integration
• weak dashboards
• no baseline outcomes
• loose carry rules when risk is high
• pretending diversion control is rude to talk about
• confusing decriminalization with a full market solution
This is exactly why Part 9 argued for risk-based drug regulation rather than one rule for every substance.
It is why Part 10 kept insisting that reform still needs accountability.
It is why Part 11 rejected fear theater and pushed evidence-based prevention instead.
And it is why the high-risk lane in this series keeps landing in some version of the regulated pharmacy model rather than broad retail access.
So what should policymakers take from these drug policy case studies?
First, not every reform belongs in the same bucket. Portugal drug decriminalization is about removing criminal penalties for possession and pushing people toward health systems. British Columbia safer supply is about short-range overdose prevention in a toxic-supply crisis. Switzerland heroin-assisted treatment is about a highly structured lane for people with severe opioid disorder. Cannabis legalization outcomes are about regulated retail for a lower-risk substance.
Second, the tighter the risk, the tighter the lane. That is the whole logic of risk-based drug regulation.
Third, systems survive politically when they can prove results. The public will tolerate risk far more readily than vagueness.
Fourth, high-risk reform that ignores diversion control is self-sabotage.
Fifth, no case study supports the idea that punishment alone solved the problem better.
The Bottom Line
These drug policy case studies do not say one thing.
They say five or six important things at once.
They say Portugal drug decriminalization is real, useful, and often oversold. They say Oregon Measure 110 results were damaged by weak rollout and weak proof, not just by the idea of decriminalization itself. They say British Columbia safer supply was a serious answer to a serious overdose crisis, but it needed harder diversion control and tighter supervision. They say Switzerland heroin-assisted treatment is powerful evidence for a narrow medical lane. And they say cannabis legalization outcomes are better than prohibition in some ways and sloppier than they should be in others.
That is the adult conclusion.
Not everything worked. Not everything failed. But the better outcomes clustered around the same principles: risk-based drug regulation, tight high-risk controls, treatment on demand, visible accountability, measurable overdose prevention, and real public safety rules.
That is not cherry-picking.
That is pattern recognition.
Frequently Asked Questions
What do these drug policy case studies say about drug legalization overall?
They say broad arguments about drug legalization are usually too vague to be useful. The better question is which substance, which risk level, which access lane, which safeguards, and which metrics.
Is Portugal drug decriminalization proof that reform always works?
No. Portugal drug decriminalization shows that shifting possession away from criminal punishment can work better than critics claim, but it does not prove every reform model works and it does not solve toxic supply by itself.[1][2][3][4]
Why are Oregon Measure 110 results so contested?
Because Oregon Measure 110 results include both real service investments and real execution failures. Oregon did fund networks across the state, but the audit found weak integration, weak data, unclear goals, and uneven implementation.[5][6]
Does British Columbia safer supply support or weaken reform?
Both, depending on what lesson you take. British Columbia safer supply supports the case for direct overdose prevention in a toxic-supply crisis, but it also shows that high-risk prescribing needs hard diversion control, close supervision, and honest evaluation.[7][8][9]
Why does Switzerland heroin-assisted treatment matter so much?
Because Switzerland heroin-assisted treatment is one of the clearest examples of a narrow high-risk medical lane improving outcomes without pretending that the answer is broad retail access. It is the closest real-world cousin to the regulated pharmacy model in this series.[10][11]
What do cannabis legalization outcomes say about commercialization?
Cannabis legalization outcomes say regulated access can outperform prohibition in important ways, but commercialization can still raise adult use, create youth-product problems, and complicate public safety if the rules are weak.[12][13][14][15][16]
References
[1] European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) https://www.euda.europa.eu/drugs-library/decriminalisation-europe-recent-developments-legal-approaches-drug-use-eldd-comparative-study_en
[2] Transform Drug Policy Foundation. https://transformdrugs.org/blog/drug-decriminalisation-in-portugal-setting-the-record-straight
[3] Mendes RO, Pacheco PG, Nunes JPCOV, Crespo PS, Cruz MS. Literature review on the implications of decriminalization for the care of drug users in Portugal and Brazil. https://pubmed.ncbi.nlm.nih.gov/31508758/
[4] Félix S, Portugal P. Drug decriminalization and the price of illicit drugs. https://pubmed.ncbi.nlm.nih.gov/27940068/
[5] Oregon Health Authority. Behavioral Health Resource Network (BHRN) Program. https://www.oregon.gov/OHA/HSD/AMH/Pages/Measure110.aspx
[6] Oregon Secretary of State Audits Division. Oregon Health Authority: Measure 110 Lacks Stability, Coordination, and Clear Results. https://sos.oregon.gov/audits/Pages/audit-2025-29-OHA.aspx
[7] Health Canada. Safer supply: Prescribed medications as a safer alternative to toxic illegal drugs. https://www.canada.ca/en/health-canada/services/opioids/responding-canada-opioid-crisis/safer-supply.html
[8] Province of British Columbia. Prescribed alternatives. https://www2.gov.bc.ca/gov/content/health/practitioner-professional-resources/pharmacare/pharmacare-substance-use-disorder-hub/pa
[9] Nguyen HV, Mital S, Bugden S, McGinty EE. British Columbia’s Safer Opioid Supply Policy and Opioid Outcomes. https://pubmed.ncbi.nlm.nih.gov/38227344/
[10] Smart R, Reuter P. Does heroin-assisted treatment reduce crime? A review of randomized-controlled trials. https://pubmed.ncbi.nlm.nih.gov/34105206/
[11] Meyer M, Quinto A, Guessoum A, Strasser J, Dürsteler KM, Lang UE, Vogel M. Operational and clinical procedures of heroin-assisted treatment in Switzerland: a nation-wide survey study. https://pubmed.ncbi.nlm.nih.gov/41126181/
[12] Hyatt AS, Overhage L, Cook BL. Use of Tobacco and Cannabis Following State-Level Cannabis Legalization. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2836385
[13] Pawar AK, Firmin ES, Wilens TE, Hammond CJ. Systematic Review and Meta-Analysis: Medical and Recreational Cannabis Legalization and Cannabis Use Among Youth in the United States. https://pubmed.ncbi.nlm.nih.gov/38552901/
[14] Dion PM, Lampron J, Rahmani M, Gawargy TA, Cannalonga CP, Tariq K, Desjardins C, Cole V, Boet S. Road hazard: a systematic review of traffic injuries following recreational cannabis legalization. https://pubmed.ncbi.nlm.nih.gov/38951474/
[15] Colorado Department of Revenue. Marijuana sales tax revenue tops $3 billion in latest DOR monthly report. https://tax.colorado.gov/press-release/marijuana-sales-tax-revenue-tops-3-billion-in-latest-dor-monthly-report
[16] Mital S, Nguyen HV. Legalizing Youth-Friendly Cannabis Edibles and Extracts and Adolescent Cannabis Use. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2832970
