Case Studies Without Cherry-Picking —…

Q: Do drug decriminalization case studies prove decriminalization works everywhere?

No. Drug decriminalization case studies prove something more useful: outcomes depend on design, services, and measurement. Portugal drug decriminalization results did not come from a single law change. Oregon Measure 110 results show how quickly reforms can be blamed when capacity and governance are weak.

Q: Are Oregon Measure 110 results a reason to abandon reform?

They are a reason to stop doing reform badly. Oregon Measure 110 results highlight governance and implementation failures more than a simple "decriminalization good/bad" story.

Q: What does supervised consumption sites evidence actually say?

It supports the idea that supervised consumption sites can reduce acute harms and connect people to services, and recent U.S. research challenges claims of neighborhood economic damage in NYC.

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

If you want a serious conversation about drug policy reform, you need something stronger than slogans.

You need evidence. You need context. And you need the discipline to stop cherry-picking the one metric that flatters your team.

This post is U.S.-centered by default, because that is where most of the arguments in this series live. But the most useful lessons come from international examples that have been measured over time, plus a few U.S. reforms that have been stress-tested in public.

The problem with most “case study” debates is that they behave like this:

• Reform supporters cite Portugal and stop talking.
• Reform opponents cite Oregon and stop talking.

Both moves are lazy.

Real drug decriminalization case studies show a harder truth: outcomes depend on design, sequencing, treatment capacity, and whether the policy is paired with harm reduction and accountability. The legal label is not the whole story.

In this Part 17, the goal is simple:

• summarize drug decriminalization case studies without hiding tradeoffs
• interpret Portugal drug decriminalization results without turning them into a fairy tale
• interpret Oregon Measure 110 results without pretending they prove prohibition works
• show why Switzerland heroin-assisted treatment is a model of “pilot, measure, scale” done right
• summarize supervised consumption sites evidence without pretending politics will disappear

A few grounding facts before we start:

• The U.S. recorded 79,384 drug overdose deaths in 2024. That is a historic decline from 2023, but it is still catastrophic.[1]
• NIDA’s overdose statistics pull together CDC data and reinforce the obvious point many debates ignore: the U.S. overdose picture is dominated by a fast-changing illegal supply, not a regulated one.[16]
• WHO recommends a package of harm reduction interventions for people who inject drugs, including needle and syringe programmes and opioid agonist treatment, to reduce HIV/HCV transmission and deaths due to overdose prevention gaps.[17][18]
• UNAIDS’ 2026 guidance note on decriminalization of drug use in the context of HIV argues criminalization limits access to health and harm reduction services and is linked to worse HIV outcomes among people who inject drugs.[19]
• Many “policy failures” are actually implementation failures: weak capacity, unclear ownership, missing measurement, and no enforcement plan.[3] That is why drug decriminalization case studies have to be read like system audits, not like moral parables.

If you want a one-line takeaway: the reforms that work are the ones built like systems, not like press releases.

This Chapter Builds on Earlier Parts

If you want definitions first, use these:

• Part 1 — drug legalization is regulation and harm reduction
• Part 1.5 — drug decriminalization vs regulation

If you want the “why prohibition fails” pieces:

• Part 2 and Part 3 — felony consequences
• Part 4 — what regulation buys
• Part 5 — market violence
• Part 6 — modern supply chain

If you want the “how reform is built” pieces:

• Part 7 — funding treatment
• Part 8 — fear-based policy
• Part 9 — risk-based drug regulation
• Part 10 — public safety and accountability
• Part 11 — prevention
• Part 12 and Part 13 — reentry and stability
• Part 14 — regulated pharmacy model
• Part 16 — pilot programs and policy evaluation

Drug Decriminalization Case Studies: A Comparison Table

The table below is the whole point of “without cherry-picking.” It forces each case study to show both wins and losses.

Case study	What changed	What worked	What failed / got messy	What to copy	What not to copy
Portugal	Decriminalized personal possession; moved cases to administrative panels; expanded treatment and harm reduction capacity over time.[4][5]	Reduced criminal justice burden; improved health outcomes alongside harm reduction; long-run evidence does not show a “major use explosion” after decriminalization.[5]	Often overstated; causal claims not always clean; some metrics moved before the law change; youth/young adult use trends are mixed depending on year and measure.[6]	Pair legal change with treatment and harm reduction; use administrative pathways; publish monitoring data.	Don’t sell it as “legalization worked” when it was decriminalization plus a health shift.
Switzerland	Four-pillar policy; introduced Switzerland heroin-assisted treatment for treatment-resistant OUD; built evaluation and communication into the rollout.[7]	HAT was implemented as a scientific and political process; results documented, adaptations made, and the policy moved from experiment to routine care.[7][8]	Resource intensive; politically hard; narrow eligibility; not a mass-market policy.	Run policy like a pilot programs model: evaluation, transparency, iteration.	Don’t pretend this works without strong clinical infrastructure.
Oregon	Measure 110 removed criminal penalties for possession and aimed to expand treatment/recovery services funded by cannabis tax revenue; later legislation recriminalized possession.[3][10]	Large investment in service networks; created a model for attempting to shift response away from incarceration.[3]	State audit found stability, coordination, and measurement problems; later legislation changed the program sharply.[3][10]	Funding streams and service networks matter; policy evaluation systems are not optional.	Don’t remove penalties without building capacity and clear operational ownership.
British Columbia	Safer supply and later a decriminalization exemption pilot; decriminalization exemption expired Jan 31, 2026.[11][12]	Early safer supply pilot findings reported improved stability for participants in qualitative assessments.[12]	Population-level study found increased opioid-poisoning hospitalizations during safer supply policy period; decriminalization pilot ended and remains contested.[13][11]	Monitor diversion, poisonings, and access; measure honestly; adjust rules quickly.	Don’t scale controversial access models without real-time monitoring and guardrails.
New York City	First publicly recognized supervised consumption sites opened Nov 30, 2021.	Study found no significant changes in local economic activity after opening; debates should focus on public health implications.[14]	Politically controversial; legal risk; scale limited; evidence debate hijacked by neighborhood fears.	Use supervised consumption sites evidence to address community “harm” claims; integrate with treatment and services.	Don’t ignore community process and transparency.

Quick U.S. overdose context (because every case study is now argued as if it caused fentanyl):

Year	U.S. overdose deaths	Source
2023	105,007	CDC Data Brief (2003–2023)[2]
2024	79,384	CDC Data Brief (2023–2024)[1]

Case Studies Without Cherry-Picking

This section does not exist to score points. It exists to stop the childish pattern where everyone cites one country and calls it “proof.”

Each case study below includes:

• what changed
• what the best evidence supports
• what critics are right about
• what lessons actually translate to the U.S.

Portugal Drug Decriminalization Results

Portugal is the most cited reform example in the English-speaking internet. That is exactly why it gets abused.

Here is what Portugal actually did, in basic terms:

• It decriminalized personal possession and use up to defined thresholds.
• It kept trafficking criminal.
• It routed cases to administrative “dissuasion” commissions that could recommend treatment, impose fines, or issue warnings.[6]
• It expanded treatment and harm reduction as part of a broader health-centered shift.[5][6]

The strongest peer-reviewed summary in the literature is not a viral chart. It is Hughes and Stevens, which compares trends and concludes the reform did not produce major increases in use and that some problematic use and harms declined, while criminal justice pressures eased.[5]

But “without cherry-picking” means we also say what the critics are right about.

The U.S. ONDCP/OJP memo is blunt about research limitations: it points out that some favorable trends began before decriminalization and that some adverse trends (including increases in drug-related deaths in certain years) were sometimes ignored in popular summaries.[6]

So the honest interpretation of Portugal drug decriminalization results is:

• Portugal reduced criminalization of users and supported a health shift.
• The best-case claim is “no major use explosion and improved health outcomes with harm reduction,” not “decriminalization alone caused everything.”
• The caution is “don’t pretend a single law change explains all outcomes, and don’t hide unfavorable years.”

The U.S. lesson: if you want Portugal drug decriminalization results, you need Portugal’s system posture — administrative response, treatment capacity, harm reduction, and long-run monitoring — not a slogan.

Switzerland Heroin-Assisted Treatment

Switzerland heroin-assisted treatment exists because Swiss policymakers faced a visible crisis and chose pragmatism over ideology.

Uchtenhagen describes how Switzerland moved toward a four-pillar policy and introduced heroin prescription for chronic, treatment-resistant heroin dependence as one innovation.[7]

The reason Switzerland heroin-assisted treatment matters for U.S. debates is not that America can copy it tomorrow.

The reason it matters is that it is “pilot, measure, scale” done the grown-up way:

• extensive scientific and political preparation
• detailed documentation and evaluation
• public communication of results
• adaptation based on findings
• moving from experiment to routine care after results were established[7]

RAND’s evidence synthesis summarizes a broad literature on heroin-assisted treatment (HAT), including outcomes like retention, reduced illicit use, reduced criminal activity, and health functioning, with attention to study design and context.[8]

No, Switzerland heroin-assisted treatment is not a mass solution. It’s a high-structure lane for a narrow, high-risk population.

That is exactly why it is useful in a risk-based drug regulation system: when risk is extreme and standard treatment has failed, the response becomes more structured, not less structured. This is the logic behind a regulated pharmacy model lane (Part 14).

The U.S. lesson: the best high-risk policies are the ones with narrow eligibility, high accountability, and a measurable clinical pathway.

Oregon Measure 110 Results

People cite Oregon Measure 110 results as if it proves either:

• “decriminalization is a disaster,” or
• “decriminalization would have worked if everyone had cooperated.”

Neither framing is serious by itself.

The Oregon Secretary of State’s audit is the adult version of the story: it focuses on governance. It emphasizes that the program lacked stability, coordination, and clear results, and it documents how the program changed through later legislation that recriminalized controlled substances.[3][10]

That matters because it exposes a pattern you see in weak reforms:

• the law changes fast
• the capacity changes slowly
• the measurement is unclear
• the public sees visible disorder and assumes causation
• the political system panics and resets

This is the hard sentence: Oregon Measure 110 results are not only about policy intent. They are about implementation capability.

If reform removes penalties but cannot deliver treatment fast and visibly, the public will treat reform as “softness” and demand a rollback. That is not a moral argument. It is a political reality.

The U.S. lesson: drug decriminalization without treatment capacity, operational ownership, and clear metrics is an invitation to backlash.

British Columbia and the “Safe Supply vs. Safe System” Problem

British Columbia is the modern example of why “pilot” matters.

For safer supply, Health Canada’s early findings summarized qualitative reports of improved stability and reduced reliance on the illegal supply among participants.[12]

But population-level evaluations complicate the story. A JAMA Internal Medicine analysis found the safer opioid supply policy period was associated with increased opioid-related poisoning hospitalizations, while deaths did not significantly change in that study window.[13]

Then add the decriminalization pilot: British Columbia’s government states the exemption launched in 2023, expired Jan 31, 2026, and was not renewed.[11]

That combination is exactly why people lose trust. They see moving parts, contested metrics, and political reversal.

Here is the non-ideological lesson:

• “Safer supply” is not a single intervention.
• It is a category with design choices: eligibility, dispensing rules, supervision, monitoring, and integration with treatment.
• Those choices can produce different outcomes.
• If you don’t measure in real time, you don’t know whether you are lowering risk or moving it.

The U.S. lesson: build a safe system, not just a safe product headline. That means monitoring poisonings, diversion, treatment engagement, and public safety continuously — and adjusting rules fast.

Supervised Consumption Sites Evidence

The U.S. argument against overdose prevention centers often uses the same script:

“If you open these sites, neighborhoods will collapse.”

That claim has to be measured, not debated like mythology.

A 2026 JAMA Network Open study on New York City’s first publicly recognized overdose prevention centers found no significant changes in foot traffic or consumer spending after opening, suggesting debates should focus on public health implications rather than assumed economic harm.[14]

That does not settle every question. It settles one important question: the “it will destroy local business” claim should not be used as a substitute for evidence.

And the supervised consumption sites evidence from Europe is broader: EUDA summaries describe drug consumption rooms as interventions intended to reduce acute harms and connect people to services, and Europe has extensive experience with these models.[15]

The U.S. lesson: supervised consumption sites evidence is strong enough that the debate should be about design, siting, and integration with treatment — not about imagined neighborhood collapse stories.

What the Case Studies Really Say

If you strip away politics, these drug decriminalization case studies keep repeating the same rules. That is why drug decriminalization case studies are more useful than slogans: they show what breaks under real pressure.

Rule one: law changes do not automatically create health capacity.

Portugal drug decriminalization results are inseparable from a wider health-centered approach, administrative pathways, and monitoring.[5][6]

Rule two: the highest-risk lane needs the most structure.

Switzerland heroin-assisted treatment worked because it was narrow, monitored, and integrated into a broader system.[7][8]

Rule three: measurement is not optional.

Oregon Measure 110 results show what happens when governance and public reporting do not keep up with ambitious change.[3][10]

Rule four: harm reduction needs legitimacy, not vibes.

Supervised consumption sites evidence improves political viability when studies address what communities fear (crime, disorder, “economic harm”) with data.[14][15]

Rule five: controversial access models must be piloted, not assumed.

British Columbia shows how quickly trust collapses when policy is contested, outcomes are mixed, and reversal follows.[11][13]

If your reform proposal does not include:

• a clear enforcement plan focused on public safety (not possession)
• treatment capacity and real pathways (treatment on demand)
• monitoring, auditing, and published metrics (policy evaluation)
• a high-structure lane for the highest-risk population (regulated pharmacy model)

…then you do not have reform. You have a gamble.

This is why Part 16 exists. The pilot programs posture is how you learn without betting the farm.

Rollout Timeline for a U.S.-Centered Reform Path

Below is a rollout timeline built from the case studies above: start narrow, measure, scale.

It’s written to be realistic under U.S. federal constraints and state-level operational reality.

gantt
    title Pilot-to-Scale Timeline Using Case Study Lessons
    dateFormat YYYY-MM-DD
    section Build
    Define pilot scope + metrics (publish up front) :a1, 2026-04-01, 45d
    Stand up treatment capacity + pathways (treatment on demand) :a2, 2026-04-15, 90d
    Governance + independent evaluator :a3, 2026-04-15, 60d
    section Pilot
    Launch lowest-risk regulated lane (strict retail rules) :b1, 2026-07-15, 60d
    Launch highest-risk clinical lane (pharmacy / high-structure access) :b2, 2026-08-15, 90d
    Launch supervised services where needed (if local legal posture allows) :b3, 2026-08-15, 90d
    section Measure
    Quarterly public dashboard (overdose, treatment, diversion, public safety) :c1, 2026-10-01, 365d
    Independent interim evaluation + rule adjustments :c2, 2027-01-15, 60d
    section Scale
    Scale / tighten / stop decision gate :d1, 2027-10-15, 30d

If you want the short version: Oregon shows what happens when you change law faster than capacity. Switzerland shows what happens when you build, test, and communicate results. Portugal shows what happens when legal change is paired with a health system posture. NYC shows why claims about neighborhood collapse should be proven, not assumed.

Frequently Asked Questions

Do drug decriminalization case studies prove decriminalization works everywhere?

No. Drug decriminalization case studies prove something more useful: outcomes depend on design, services, and measurement. Portugal drug decriminalization results did not come from a single law change. Oregon Measure 110 results show how quickly reforms can be blamed when capacity and governance are weak.[5][6][3]

Do Portugal drug decriminalization results prove legalization is safe?

No. Portugal did not legalize supply. It decriminalized possession, kept trafficking criminal, and used administrative panels plus health investments.[6] Portugal drug decriminalization results are a lesson in policy posture, not a blank check for commercialization.[5][6]

What does Switzerland heroin-assisted treatment mean for the U.S.?

It shows that the highest-risk population can be managed with a high-structure clinical lane when the system is built deliberately and evaluated transparently.[7][8] That supports a regulated pharmacy model lane (Part 14), not broad retail access for the highest-risk substances.

Are Oregon Measure 110 results a reason to abandon reform?

They are a reason to stop doing reform badly. Oregon Measure 110 results highlight governance and implementation failures more than a simple “decriminalization good/bad” story.[3][10]

What does supervised consumption sites evidence actually say?

It supports the idea that supervised consumption sites can reduce acute harms and connect people to services, and recent U.S. research challenges claims of neighborhood economic damage in NYC.[14][15]

References

[1] Centers for Disease Control and Prevention, National Center for Health Statistics. Drug Overdose Deaths in the United States, 2023–2024 (Data Brief 549). https://www.cdc.gov/nchs/products/databriefs/db549.htm

[2] Centers for Disease Control and Prevention, National Center for Health Statistics. Drug Overdose Deaths in the United States, 2003–2023 (Data Brief 522). https://www.cdc.gov/nchs/data/databriefs/db522.pdf

[3] Oregon Secretary of State Audits Division. Oregon Health Authority: Measure 110 Lacks Stability, Coordination, and Clear Results. https://sos.oregon.gov/audits/Documents/2025-29.pdf

[4] European Union Drugs Agency (EUDA). Portuguese national focal point (Reitox network). https://www.euda.europa.eu/about/partners/reitox/portugal_en

[5] Hughes CE, Stevens A. What can we learn from the Portuguese decriminalization of illicit drugs? (2010 PDF). https://kar.kent.ac.uk/29910/1/Hughes%20%20Stevens%202010.pdf

[6] Office of National Drug Control Policy (archived via OJP). Drug Decriminalization in Portugal: Challenges and Limitations (August 2010). https://www.ojp.gov/pdffiles1/ondcp/Portugal.pdf

[7] Uchtenhagen A. Heroin-assisted treatment in Switzerland: a case study in policy change. Addiction. https://pubmed.ncbi.nlm.nih.gov/19922519/

[8] Smart R, et al. Evidence on the Effectiveness of Heroin-Assisted Treatment. RAND (2018). https://www.rand.org/pubs/working_papers/WR1263.html

[9] Oregon Judicial Department. Measure 110 Circuit Court Cases with Class E Violations (PDF, Aug 31, 2024). https://www.courts.oregon.gov/about/Documents/BM110Statistics.pdf

[10] Oregon Legislative Information System. HB 4002 (2024 Regular Session) overview and history. https://olis.oregonlegislature.gov/liz/2024R1/Measures/Overview/HB4002

[11] Government of British Columbia. Decriminalizing people who use drugs in B.C. (pilot expired Jan 31, 2026). https://www2.gov.bc.ca/gov/content/overdose/decriminalization

[12] Health Canada. Early findings from safer supply pilot projects. https://www.canada.ca/en/health-canada/services/opioids/responding-canada-opioid-crisis/safer-supply/early-findings-safer-supply-pilot-projects.html

[13] Nguyen HV, et al. British Columbia’s Safer Opioid Supply Policy and Opioid-Related Poisoning Hospitalizations and Deaths. JAMA Internal Medicine (PMC). https://pmc.ncbi.nlm.nih.gov/articles/PMC10792500/

[14] Allen B, et al. Overdose Prevention Centers and Neighborhood Commercial Activity in New York City. JAMA Network Open (2026) (PubMed). https://pubmed.ncbi.nlm.nih.gov/41758519/

[15] European Union Drugs Agency (EUDA). Drug consumption rooms: an overview of provision and evidence (Perspectives on drugs). https://www.euda.europa.eu/publications/pods/drug-consumption-rooms_en

[16] National Institute on Drug Abuse (NIDA). Drug Overdose Deaths: Facts and Figures. https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates

[17] World Health Organization (WHO). HIV: People who inject drugs. https://www.who.int/teams/global-hiv-hepatitis-and-stis-programmes/populations/people-who-inject-drugs

[18] World Health Organization (WHO). WHO, UNODC, UNAIDS technical guide for countries to set targets for universal access to HIV prevention, treatment and care for injecting drug users. https://www.who.int/publications/i/item/978924150437

[19] UNAIDS. Decriminalization of drug use in the context of HIV (Guidance note, March 10, 2026). https://www.unaids.org/sites/default/files/2026-03/20260310_Decriminalization-drug-use-HIV.pdf

Previous Next