How Cannabis Prohibition Still Fills Prisons Worldwide

Photo by Maia Fotografia via Pexels.

Somewhere close to 256 million people smoked, vaped, or otherwise consumed cannabis in the past year, according to the UNODC's World Drug Report 2026, released this past June. That makes it the most widely used controlled substance on the planet, more common than any opioid, stimulant, or synthetic drug tracked by the agency. And yet, in the overwhelming majority of the roughly 190 UN member states, possessing it remains a criminal act punishable by fines, jail, or in a handful of jurisdictions, execution.

That contradiction is the whole story here. Usage is climbing everywhere, but what happens to you for using -- or growing, or selling -- depends almost entirely on which side of a border you're standing on. A tourist caught with a joint in Amsterdam faces essentially nothing. Someone caught with the same joint in parts of Southeast Asia could be looking at years behind bars, or worse. Thailand alone managed to flip its own policy in both directions within about three years, decriminalizing in 2022 and then clawing much of that back by mid-2025 -- proof that incarceration numbers tied to cannabis aren't some fixed consequence of the plant itself. They're a policy dial that governments turn up and down, sometimes within a single legislative session.

This piece walks through that range: the zero-tolerance security-state model running through Singapore and much of Southeast Asia, the fractured and often contradictory patchwork of American state and federal law, and the outlier cases of Portugal and Bolivia, two countries that sit on opposite ends of nearly every metric that matters when it comes to who ends up in a cell over cannabis.

A Growing User Base, an Unmoved Legal Map

Photo by Andrew Stutesman via Unsplash.

The World Drug Report 2026, published by the UN Office on Drugs and Crime on June 26, 2026, put the number of past-year cannabis users at approximately 256 million globally. That figure has grown by roughly 40% over the last decade, and the drivers are pretty straightforward: general population growth accounts for part of it, but prevalence itself has also climbed, from about 3.8% of adults a decade ago to roughly 4.8% today. That's not a marginal shift -- it means cannabis use is outpacing population growth on its own terms, spreading into demographics and regions where it was previously rare or well-hidden.

What hasn't moved anywhere near as fast is the legal status attached to that use. Criminalization is still the default framework in most of the countries counted in that 256 million figure. Decriminalization and legalization dominate headlines because they're relatively new and politically newsworthy, but numerically they remain the exception rather than the rule. Most cannabis consumers on earth are, technically speaking, committing a crime every time they use it.

That gap -- enormous and growing usage against a legal map that's barely budged -- is the throughline for everything else in this piece. It explains why enforcement outcomes vary so wildly from one jurisdiction to the next despite consumption patterns that are, relatively speaking, converging. Nobody is arguing that cannabis use is rare or fringe anymore; the data doesn't support that framing. What varies is entirely a matter of statute, enforcement priority, and political will, not underlying behavior. Keep that distinction in mind, because it resurfaces in every regional case that follows, from Singapore's execution chambers to Portugal's administrative panels.

Singapore and Southeast Asia's Zero-Tolerance Model

Photo by KATRIN BOLOVTSOVA via Pexels.

Singapore's Misuse of Drugs Act sets a mandatory death penalty for anyone convicted of trafficking more than 500 grams of cannabis. There's no judicial discretion built in once the trafficking threshold and intent are established -- the sentence is fixed by statute, not left to a judge's weighing of circumstances. As of mid-2026, Singapore had carried out eight drug-related executions that calendar year alone, continuing a pace the country has maintained fairly consistently since it resumed executions after a pandemic-era pause.

Singapore's government has been explicit and consistent about the reasoning: officials frame drug trafficking, cannabis included, as a threat to national stability and public order rather than a public health issue to be managed through treatment or harm reduction. That framing matters because it changes the entire policy toolkit available. Public health approaches lean toward diversion programs, medical treatment, and social services. National security framing leans toward deterrence, interdiction, and, in Singapore's case, capital punishment as the ultimate deterrent signal.

Singapore isn't isolated in this posture -- it reflects a wider pattern across parts of Southeast Asia, where drug offenses, including cannabis possession and distribution, have historically been prosecuted with a severity closer to how other countries treat arms trafficking or violent crime than how they treat a controlled substance now legal for adult use in parts of North America and Europe. Vietnam, Malaysia, and Indonesia each maintain harsh statutory penalties for cannabis-related offenses, though enforcement intensity and the frequency of capital sentencing varies by country and by year.

Set against this backdrop, the regional contrast becomes almost jarring. Thailand, sitting geographically and culturally within this same neighborhood, spent several years running one of the most permissive cannabis markets in Asia before reversing course. Singapore's zero-tolerance model and Thailand's whiplash didn't happen in different worlds -- they happened next door to each other, within the same regional bloc, often within the same news cycle.

Thailand's Whiplash: From 285,000 Prisoners to Legal Cannabis and Back

Photo by Markus Winkler via Pexels.

Before 2022, Thailand's prison system held the largest incarcerated population in ASEAN, somewhere around 285,000 people. More than 80% of those inmates were there on drug-related charges, a proportion that dwarfs incarceration patterns tied to drug offenses in most other countries with comparably sized justice systems. Cannabis possession and low-level trafficking made up a meaningful share of that population, feeding a prison system that had become, by most accounts, badly overcrowded.

Then, in June 2022, Thailand decriminalized cannabis almost overnight, becoming the first country in Asia to do so. The effect on the prison population was immediate and measurable: roughly 4,200 inmates who had been jailed specifically for cannabis offenses became eligible for release that same day. Dispensaries and cannabis shops proliferated at a pace regulators struggled to track -- estimates by 2024 put the number of licensed shops somewhere between 10,000 and 18,000 nationwide, an astonishing figure for a country that had criminalized the plant just two years earlier.

But the reversal came almost as fast as the opening did. On June 23, 2025, Public Health Minister Somsak Thepsutin signed an order banning those same 10,000 to 18,000 shops from selling cannabis without a medical prescription, effectively ending the recreational retail market that had defined Thailand's post-2022 image. Violations of the new order carry penalties of up to one year in jail alongside a fine of 20,000 baht, and the government has signaled its intent to go further still, formally reclassifying cannabis as a narcotic and rolling back the legal architecture built over the previous three years.

Thailand is, in effect, a live experiment running in real time: an entire country's incarceration exposure tied to cannabis has swung dramatically twice within roughly three years, driven not by any change in how many people were using the plant, but purely by how the law classified it. That's a data point worth sitting with -- it undercuts any claim that incarceration rates reflect some fixed, objective relationship between cannabis and public harm. They reflect whatever a health minister signs on a given Monday.

The U.S. Patchwork: Legal in One State, Arrestable Next Door

In 2025, U.S. states that still prohibit cannabis recorded roughly 186,600 arrests, more than eight times the approximately 22,400 arrests in states where cannabis is legal—highlighting the strong link between legalization and reduced enforcement.

A Marijuana Policy Project report released in April 2026 laid out the domestic contradiction bluntly: in 2025, states with legal adult-use cannabis logged 22,357 cannabis-related arrests combined, while states still operating under prohibition logged 186,581 -- more than eight times as many, despite prohibition states not having anywhere near eight times the combined population of legal states. The disparity isn't subtle, and it's not shrinking on its own.

Zoom out to the national trend line and the picture looks more encouraging on the surface. Annual cannabis arrests across the U.S. have fallen from a peak above 900,000 to just over 200,000 as more states have legalized or decriminalized over the past decade and a half. That's real progress by any measure. But the state-level breakdown makes clear that progress has been uneven rather than uniform -- a resident of a legal state benefits enormously, while a resident of a neighboring prohibition state, sometimes separated by nothing more than a state highway sign, still faces meaningful arrest risk for identical conduct.

Federal incarceration tells a smaller but still relevant piece of this story. Advocacy groups estimated the number of federal prisoners held specifically on marijuana offenses at roughly 10,560 as of 2015; given the marked decline in federal marijuana prosecutions since then, the current figure is likely under 10,000, though exact contemporary counts are harder to pin down given how the Bureau of Prisons categorizes drug offenses. Even with that decline, the U.S. reportedly maintains the highest imprisonment rate for marijuana offenses of any country in the world, ahead of Poland and Romania, two countries that themselves maintain fairly strict cannabis statutes within the EU.

The reason this matters beyond raw statistics is that arrests are the front door to incarceration -- most people don't go straight to prison without first being arrested, charged, and processed through a local system. So even as federal numbers shrink, prohibition states keep feeding local jails and state prisons through sheer volume of arrests, meaning the American incarceration picture on cannabis is really 50 separate pictures, not one.

Portugal's Decriminalization vs. Bolivia's Decade-Long Sentences

Photo by Mikhail Nilov via Pexels.

Portugal's 2001 drug policy overhaul remains the reference point international observers reach for whenever the conversation turns to decriminalization done well. Under that framework, personal possession of cannabis and other drugs is treated as an administrative violation rather than a crime, handled by regional "dissuasion" panels rather than criminal courts. Nobody goes to prison for possessing a personal-use quantity. The model is widely credited with sharply reducing drug-related incarceration in the country over the two decades since, even as broader drug policy debates continue about its effects on usage and public health outcomes more generally.

Bolivia occupies the opposite end of that spectrum entirely. Under current Bolivian law, possession of even one gram of cannabis can carry a sentence of 10 to 25 years in prison. That's not a typo-worthy exaggeration -- it's the statutory range for what, in Portugal, would trigger nothing more than a conversation with a dissuasion commission and possibly a small fine. One gram, roughly enough for a single joint, sits at the center of two entirely different legal universes depending on which country you're standing in.

What makes this comparison useful isn't just the shock value of the gap. It's what the gap reveals about what's actually driving incarceration outcomes worldwide: not usage rates, not measurable public health harm, not any objective difference in how dangerous cannabis is in Lisbon versus La Paz. It's legislative choice, full stop. Portugal and Bolivia don't have wildly different populations of cannabis users relative to their size -- they have wildly different legislatures and enforcement priorities.

One practical note worth underlining, given everything Thailand has demonstrated in just the past few years: these frameworks are not permanent fixtures. Laws shift, sometimes with very little warning, and enforcement practice can shift even faster than the underlying statute. Anyone traveling internationally, relocating, or simply trying to understand their own legal exposure should verify current statutes directly rather than relying on a country's reputation or a classification that might be a year or two out of date.

What the Numbers Actually Show

Photo by Magda Ehlers via Pexels.

Line up the numbers from this piece and a pattern falls into place pretty cleanly. Global user counts keep rising -- up 40% over a decade, now at roughly 256 million people -- while the legal frameworks governing that use stay largely frozen in place, or in Thailand's case, actively reverse course within a matter of months. Nothing about the drug itself changed between 2022 and 2025 in Thailand. The classification did, twice.

Incarceration, it turns out, tracks classification and enforcement priority far more closely than it tracks anything resembling measurable harm or comparative usage rates between countries. Singapore isn't executing more people because Singaporean cannabis use is more dangerous than cannabis use in Portugal. Bolivia isn't sentencing people to a decade or more because Bolivian cannabis carries different risks than cannabis in a Colorado dispensary. These are policy choices, dressed up in the language of public safety, but functionally disconnected from any consistent international standard of what constitutes harm worth that level of punishment.

The American patchwork drives the point home at a smaller, more granular scale. You don't need to cross an ocean to see the disparity -- crossing a state line can be enough. Legalization states logged 22,357 cannabis arrests in 2025; prohibition states logged 186,581. Same country, same plant, same conduct, wildly different legal consequences depending purely on geography.

Put the full range side by side and it's genuinely striking: execution in Singapore, a decade or more in prison in Bolivia, an administrative fine and a conversation with a panel in Portugal. Three governments, one plant, three completely different answers to the question of what possessing it should cost a person. That range isn't a reflection of how different cannabis is from place to place. It's a reflection of how different the politics around it remain.

None of this is fixed the way, say, a chemical property of the plant is fixed. Incarceration tied to cannabis is a policy output, and policy outputs can be changed by a legislature, a ministerial order, or a court ruling -- sometimes within a single year, sometimes within a single signature. Thailand proved that twice over, first by emptying cells in 2022 and then by threatening to refill them in 2025. There's no reason to assume other countries are somehow immune to that same volatility, whichever direction their current laws happen to point.

The widening gap between how many people use cannabis worldwide -- 256 million and climbing -- and how severely a shrinking share of those countries choose to punish that use is going to keep putting pressure on the public-safety justification that underpins most prohibition regimes. It gets harder every year to argue that punitive frameworks are calibrated to actual harm when usage keeps rising across nearly every region tracked by the UNODC while enforcement outcomes remain this scattered and, in places, this severe.

For anyone actually tracking this space rather than just reading about it occasionally, the practical takeaway is to watch classification battles closely, not broad cultural sentiment. A single ministerial order, as Thailand demonstrated in June 2025, can instantly change the legal exposure of tens of thousands of people who did nothing differently the day before that order was signed than the day after.