What the Stevens UN Paper Really Reveals About Drug Policy

Recovery sits beyond the binary. It is neither permissive nor punitive. It is simply the stubborn belief that people can change, and that we owe each other the chance to do so.

Every so often a piece of academic work arrives that unsettles the furniture a little. It does not tell you anything wildly new, but it rearranges what you already sensed into a shape you can finally point to. The recent Stevens analysis of the 2024 United Nations drug policy statements is one of those pieces. It suggests, more plainly than most scholars dare, that the global debate on drugs is not truly about evidence at all. It is about values. Always has been.

What the authors did was deceptively tidy. They took each country’s official UN statement and treated it not as diplomatic wallpaper but as a moral expression of how that country understands the human person and the social order. Feed these coded statements through a network model and, like iron filings swirling around twin magnets, two constellations appear. One cluster radiates emancipative values, personal autonomy, human rights language and the familiar vocabulary of decriminalisation harm reduction and increasingly legalisation . The other gathers around themes of tradition, communal duty, the safeguarding of families and neighbourhoods, the preservation of social order and the hopeful conviction that a society worthy of its young should aspire to be drug free

It is elegant work, genuinely illuminating. Anyone who has sat through these international gatherings will recognise it immediately. Countries do not sort themselves by data strength. They sort themselves by the stories they tell about what they fear, what they value and what a good society ought to look like.

And yet the moment the picture sharpens, its absences become equally striking. The model makes two moral languages vividly visible and, in doing so, unintentionally hides a third. Recovery, as lived by families, mutual aid communities and people who have rebuilt their lives from the ground up, does not sit neatly within either camp. Its values dignity, responsibility, solidarity, hope, transformation, perseverance are not captured by the liberal language of individual autonomy, nor are they identical to the traditional emphasis on social order and communal protection. Recovery speaks a different moral dialect altogether, one rooted in the belief that people can change and that communities, when they are functioning well, help carry that change into being. Recovery is neither permissive nor punitive. It is relational. Communal. It belongs to a moral grammar that blends compassion with accountability and believes, quite seriously, that people can change.

And here is where the Stevens paper becomes even more interesting. Because a missing worldview in a model is not always a philosophical problem. Quite often it is a methodological one.

Every map tells you what is there and what is missing. In this case, the missing piece is the moral grammar of recovery dignity, responsibility, hope and change.

There is also the small matter of the tools used. Stevens is not some neutral cartographer wandering into the UN forest and discovering two great moral tribes living there in the wild. He brings his own compass, his own mapmaking kit and, truth be told, his own preferred number of tribes. He codes the statements through a framework he designed, feeds them into a model calibrated according to what he himself finds “interpretable”, and then reports, with a straight face, that the universe has conveniently arranged itself into two tidy constellations. It is perfectly respectable scholarship. But it does mean the model can only ever reveal the moral weather patterns it was built to detect. When you design an instrument to pick up only autonomy at one end and obedience at the other, it will dutifully tell you the world consists of people who like freedom and people who like rules, and nothing in between. Recovery the stubborn, hopeful, communal worldview that actually keeps families going simply sails past the sensors like a ghost through a security scanner at Heathrow. No alarms. No trace. Just a quiet “nothing to declare”. The work here is still useful, but the map it draws is necessarily incomplete. There is a whole moral island missing from it.

Another small detail in the paper deserves a little more daylight. In the sociogram, the United Kingdom is tugged toward the traditionalist cluster. Not because Britain resembles the more authoritarian regimes at its core, but because our UN statement emphasised border control, drug seizures and the aspirational language of a drug free world while avoiding explicit reference to harm reduction. Yet back home the British establishment rarely misses a chance to signal its comfort with harm reduction orthodoxy, all the while presiding over the long, quiet starvation of residential rehabilitation and long term recovery supports. Internationally we flex prohibitionist credentials to reassure one set of allies; domestically we whisper liberal cues to placate another. Even the network analysis quietly exposes the contradiction. The UK drifts, neither fish nor fowl, triangulating its way through a culture war instead of committing to build a system capable of transforming lives.

This is where the Stevens paper earns its keep. Not by offering a final answer, but by showing us the fractures inside the debate and the gaps between rhetoric and reality. It gives us a map, and like all maps, it reveals as much by what it fails to draw as by what it includes. The absence of a recovery oriented worldview is not a flaw of scholarship, but an invitation. It suggests that perhaps we have been trying to describe a very human landscape with tools not designed for its contours. Recovery does not belong to the logic of autonomy or the logic of obedience. It inhabits a third moral space, one that blends rights with responsibilities, compassion with duty, personal change with communal belonging.

This is where nations perform their moral identities. They speak of evidence, of course, but the gravitational pull underneath is always cultural.

If anything, the Stevens analysis deepens the conversation. Evidence matters, of course it does. But evidence is never interpreted in a moral vacuum. It is always filtered through what societies believe about purpose, dignity, obligation and the meaning of the common good. If we want drug policy that heals rather than manages decline, we must make room for the moral traditions that take human change seriously.

The paper has mapped the two dominant tribes on the UN stage. The third tribe the one that rebuilds damaged lives rather than theorises about them is quieter, humbler and not easily captured in a dataset. It does not fly grand banners. It simply does the work of helping people rise again. Perhaps the next generation of scholarship will learn to see it. Until then, the map remains incomplete, but at least we can now see the outline of the missing continent.

You can read the full paper here How and why consensus fractured at the 2024 session of the UN Commission on narcotic drugs: an exploratory study of international drug policy constellations using social network analysis and qualitative comparative analysis

Alex Stevens Felipe Krause & Martin Bouchard

Received 29 Aug 2025, Accepted 12 Nov 2025, Published online: 05 Dec 2025

Alex then responded here with this on Linked in on December 15th 2025

Sociogram of countries’ policy positions stated at the 2024 Un Commission on Narcotic Drugs

Mapping the moral universe of drug policy

Alex Stevens

Professor of Criminology, University of Sheffield

December 15, 2025

I’m grateful to Anne Marie Ward for her eloquent commentary on the article I wrote with Felipe Krause and Martin Bouchard on the fracturing of consensus at the 2024 Commission on Narcotic Drugs. Our article uses concepts and methods I developed for my 2024 book on drug policy constellations. It displays, in the form of a network sociogram, the two constellations of policy actors and ideas that we found in the statements made by countries at that year’s CND meeting in Vienna. These two constellations form around traditionalist ideas like punitive prohibition and the pursuit of a ‘drug-free’ world, and liberal ideas like harm reduction, the right to health, and gender equality.

Ward’s commentary is gracious in praising our article as ‘elegant’ and ‘illuminating’, despite the rather fractious tone of our previous discussions. She makes several interesting points. The two I want to discuss here are about the reflexivity of policy constellations, and whether recovery represents - as Ward claims - a ‘third tribe’, distinct from those who support either liberal or traditionalist positions in drug policy.

The reflexivity of policy constellations

Ward writes that ‘Stevens is not some neutral cartographer wandering into the UN forest and discovering two great moral tribes living there in the wild. He brings his own compass, his own mapmaking kit and, truth be told, his own preferred number of tribes.’ Some may see this as an accusation of scientific impurity; that I have allowed my preferences to lead me to confirming my prior assumptions, rather than following the data. But I see it differently.

When Giulia Zampini and I came up with the concept of the policy constellation, we thought of it as essentially reflexive. What you see in studying policy, as in mapping stars and landscapes, depends on where you’re looking from. In astronomy, the view of astral constellations would change radically from different vantage points in our galaxy. Even on Earth, different cultures use different names for different clusters of stars. As Ward suggests, it is the cartographer (not the terrain) that draws the map.

But we did not design our study, as Ward claims, ‘to pick up only autonomy at one end and obedience at the other’ of the potential influences on countries’ drug policy positions. Our analysis also included data on whether their populations prefer decisions to be justified on the basis of religion or rational deliberation (as measured by the World Values Survey), and on the level of human development of each country. The data (not our preconceptions) led us to the provisional conclusion that, of these potential influences, the population level of commitment to emancipative values is the most important influence on the stances each country took at the 2024 CND.

We tried analysing the data in ways that produced a different number of constellations, but none of them made sense. None produced a map with three constellations, including the one that Ward says is missing from our analysis; the tribe of supporters of recovery.

Recovery and the moral universe of drug policy

Ward praises our article for ‘unsettling’ and ‘rearranging’ what we know about how drug policy works. But she also suggests that the idea of recovery is completely missing from our analysis. This, in turn, got me thinking. What is the place of recovery in the moral universe of international and UK drug policy?

Support for rehabilitation was one of the most popular policy positions for the countries we placed in the traditionalist constellation. Many of them boasted about how many rehabilitation centres they had set up, alongside stating how many arrests they had made, and how many kilos of drugs they had seized. This a moral world that equates recovery with abstinence, and sees no problem with harming people - via coercion and punishment - in order to achieve it. The Russian delegation even framed this as a way to protect human rights; the right to live free of drugs.

Visions of recovery

Ward’s vision of recovery is ‘one that blends rights with responsibilities, compassion with duty, personal change with communal belonging’. Strangely enough, that is also my vision of what harm reduction can be. My own moral commitments are most explicitly explained in an article I wrote on drug policy, harm and human rights. It uses the rationalist moral philosophy of Alan Gewirth to argue that we have rights because of our capacity to act as responsible agents. So we have a duty to do so when we can. Our compassion must be directed towards supporting people to achieve agency and responsibility for their own actions. This can be done by creating bonds of reciprocal solidarity and support.

The creators of the current UK drug strategy might also claim that such moral imperatives underlie their approach. They claim to offer a balanced approach. This has been challenged in various analyses of the strategy, but I have no doubt that several of its many authors genuinely believed that a balanced approach included both harm reduction and recovery. Recovery is surely the inspiration for the title of the UK strategy, From Harm to Hope, almost as explicitly as it was for a previous Scottish drug strategy, The Road to Recover.

Universal support for recovery

Ward’s claim that our analysis misses out recovery is belied by the presence of recovery as one of the biggest dots in our sociogram. It was supported in 20 of the 99 statements that countries made to the 2024 meeting of the CND. Its supporters included countries like China and Bahrain, who we placed in the traditionalist constellation because of their support for conformity with the previous consensus of punitive prohibition. But the supporters of recovery also included more countries from the liberal constellation, including Canada, Belgium, Norway, and the USA. Each of these countries also supported harm reduction in their CND statement.

Ward may argue that these countries (and the UK) are not committed to her vision of recovery. She wants to carve out a position where recovery is in a separate moral world from harm reduction. She regularly presents harm reduction as a counsel of despair, denying hope for change. Many harm reductionists - including me - disagree. We see harm recovery and recovery as morally and practically compatible. Both recovery and harm reduction promote the inherent worth of human life, and the rights of people to control their own fates. Without harm reduction, many people will die before they get the chance to gain control of their drug use. Without support for recovery, many will find it difficult to escape cycles of relapse and harm.

The liberal morality of communal solidarity

The key difference that Ward draws between her vision of recovery and the traditionalist and liberal constellations we found is the belief in change through communal solidarity. She pursues ‘a moral grammar that blends compassion with accountability and believes, quite seriously, that people can change’. This is fascinating stuff, because it points to a key ambivalence in contemporary liberal positions on drug policy. Is this to be a neo-liberal - even libertarian - belief in individual freedom in which there is ‘no such things as society‘? Or is it to be an older version of liberalism, which retains respect for the duties of citizens to be active in supporting each other?

As James Nicholls has recently reminded us, there are different visions of the good society within liberalism. It is not just about the ‘negative’ freedom to be left alone by the state, but can also be about the ‘positive’ freedom to promote communal wellbeing. So perhaps Ward’s vision of recovery is in search of a return to this older style of liberalism, where rights and responsibilities are better balanced, and the bonds of solidarity have not been sacrificed to the capricious gods of the free market.

Beyond the map

The map we drew of the moral universe of drug policy is at far too large a scale to pick out the many complexities of people’s nuanced relationships with drugs. Highlighting only two constellations is inevitably reductive, obscuring the existence of myriad sub-constellations of actors and ideas. But it does help to do what we set out to do, which is to describe and explain the fracturing of consensus on international drug policy at a particular time and place.

Personally, I doubt that splitting the drug policy universe to make a third, distinct group that supports recovery would be either analytically or practically productive. Rather, I would prefer that we find ways to build solidarity and mutual support between people who prioritise recovery and those who focus on harm reduction, while reducing the power and influence of those countries and politicians who want to punish and abuse the human rights of people who use drugs.

Many of the ideas that led to our analysis of international drug policy constellations are discussed in the text and appendices of my book and website on Drug Policy Constellations.

And I then responded a few days later with this here on substact but Alex has remained quiet since

Recovery Is Not a Methodological Error

Why drug policy keeps managing decline instead of expecting change.

ANNEMARIE WARD

DEC 24, 2025

This essay forms part of an ongoing public exchange with Professor Alex Stevens following his recent paper on the fracturing of consensus at the 2024 UN Commission on Narcotic Drugs. My initial response to that paper, which argues that recovery occupies a neglected moral space beyond the harm-reduction versus prohibition binary, can be read here. Professor Stevens’ subsequent reply, defending the policy constellations framework and situating recovery within liberal harm-reduction traditions, can be read here.

What follows is not a recap of those exchanges, but a clarification of where the real disagreement lies, and why it matters beyond this conversation.

---

Alex, thank you for the seriousness of your reply. I recognise the deliberate change in temperature, from our earlier exchanges to something more collegiate, and I am content to meet you there. I am also aware that our previous discussions have sometimes been read as “fractious”, but that is usually what happens when lived consequences collide with systems that prefer to discuss values in the abstract. I am less interested in winning an argument than in refusing to let the deepest disagreements be softened into technical ones.

Because that softening is exactly what is happening here, and it is worth naming it plainly.

What you have done, very skilfully, is attempt a domestication. You have reframed my critique as an internal disagreement within liberal politics and then narrowed the terrain so that the live question becomes whether a “third tribe” appears as a distinct cluster in your sociogram. Once the argument is recast as a matter of analytical parsimony or methodological plausibility, the more unsettling claim I was making about moral blind spots, institutional incentives and ideological capture can be safely set aside. I am not imputing bad faith here. I am describing a familiar pattern in how contested policy fields stabilise themselves: by translating moral challenges into methodological ones until the challenge no longer threatens the settlement.

There are three moves in your reply that do this work, and they matter far beyond our exchange.

The first is your embrace of reflexivity without allowing it to bite. You acknowledge that constellations depend on vantage point, that cartographers draw maps, that different observers would see different skies. But this acknowledgement is immediately neutralised by the claim that the data led you there anyway, that alternative configurations were tried and did not “make sense”. Reflexivity becomes a philosophical courtesy rather than a destabilising insight. Yet your own paper documents how many discretionary choices shape the outcome: what counts as a policy position, which positions are excluded as too common, which countries are dropped, how modularity is tuned, which results are deemed interpretable. None of this is illegitimate. But it does mean the map can only ever show what the instrument is capable of detecting. To acknowledge that and then proceed as if the map still simply reveals the moral universe as it is, rather than as filtered through those choices, is to blunt the force of reflexivity while retaining its language. Even if I accept your constellations exactly as drawn, the question I am raising remains untouched: why recovery, while universally endorsed in language, is consistently avoided as an organising principle in practice.

The second move is the collapse of recovery back into harm reduction by redefining harm reduction in its most morally ambitious form. You suggest that my description of recovery as dignity, responsibility, solidarity, communal bonds and transformation is also your understanding of what harm reduction can be. Philosophically, that may well be true for you, and I do not doubt your sincerity. But this shifts the debate away from what I am actually critiquing. I am not arguing that harm reduction cannot support recovery, and I never have. The repeated attempt to cast my position as “anti harm reduction” serves only to resurrect a false binary in which recovery and harm reduction are treated as opposing moral projects, rather than as practices whose value depends on how they are structured, funded and governed. I am arguing that the dominant institutional form it has taken in many jurisdictions does not require recovery to succeed, and therefore does not build systems that seriously expect it.Systems are judged not by their best intentions but by what they produce at scale. It is perfectly possible to speak the language of agency while constructing services that are organised around survival alone. This is how systems can endlessly fund overdose prevention while leaving residential rehabilitation with long waiting lists, unstable funding and fractured aftercare, and still describe themselves as “balanced”. In recovery terms, this is the difference between the leaflet and the meeting.

Results may vary by instrument

The third move is to reframe the “third moral space” I identified as merely an internal ambivalence within liberalism, a tension between thin libertarianism and thicker, more communitarian traditions. There is truth in that framing, but it also functions to keep recovery safely inside your liberal constellation rather than allowing it to stand as a critique of both dominant camps. The recovery worldview I am describing is not simply liberalism done better. It is a moral grammar with its own centre of gravity, one that insists that people are not finished products, that responsibility is possible, that community is constitutive rather than optional, and that the common good includes moral formation, not merely risk management. You can label that liberal, communitarian, Christian personalist, civic republican, or simply human. The label matters less than the fact that it does not sit comfortably on the autonomy versus obedience axis that structures your explanation of the fracture.

This brings me to the point where I think we are talking past each other.

I am not claiming that recovery is absent as a word. I am claiming it is absent as a governing logic.

You are right to note that recovery appears prominently in your sociogram and is claimed by countries across both constellations. That is precisely the problem. When a concept is universally praised yet structurally unfunded, operationally sidelined and politically instrumentalised, it becomes liturgical. Everyone recites it. Almost no one is accountable to it.

The meaningful question is not whether recovery appears in diplomatic statements. It is what happens when recovery makes demands. Do systems fund long-term, abstinence-oriented pathways with the same seriousness that they fund immediate risk containment? Do they build coherent routes from detox to residential rehabilitation to supported housing, employment and community belonging, or do they construct revolving doors where people are kept alive but left orbiting the same harms indefinitely? Do they measure change in lives, or activity in services? Do they protect mutual aid and recovery communities as civic assets, or treat them as optional extras to be praised rhetorically while real decisions are made elsewhere?

On those measures, recovery is consistently honoured in language and starved in practice. That is why I described it as a missing continent. Not because it never appears on the map, but because the map is drawn at a scale that cannot distinguish aspiration from institutional reality.

I accept that this was not the question your paper set out to answer. My concern is that, in practice, these are precisely the kinds of maps that come to stand in for the territory, and in doing so quietly determine which questions are no longer asked.

You say you would rather build solidarity between those who prioritise recovery and those who focus on harm reduction, while reducing the power of punitive actors. I share your desire to reduce punishment and abuse. But solidarity requires truth, not just goodwill. The obstacle to solidarity is not that recovery advocates are hostile to harm reduction. It is that the prevailing institutional settlement has learned to treat survival as success and transformation as a private, optional matter. That is not a failure of compassion. It is a failure of expectation.

So let me be clear about what I am, and am not, arguing. I am not trying to split the universe into three tribes for analytical novelty. I am trying to stop recovery being used as a rhetorical garnish on systems that no longer seriously expect change. This is not a call to return to a punitive past, but a refusal to accept a managed-decline future dressed up as compassion. Recovery is neither permissive nor punitive. It is the stubborn belief that people can change, and that a decent society builds ladders, not just nets and not just cages.

If your framework wants to test whether this moral space is real, there is a way to do it that goes beyond counting words in speeches. Follow the money. Follow the pathways. Follow what is scaled and what is marginal. Follow what is professionally rewarded and what is treated as reputational risk. That is where the moral universe actually reveals itself.

I appreciate the seriousness with which you have engaged. But I am not willing to allow this disagreement to be tidied away as a technical question about clustering. It is a disagreement about what kind of human being our systems assume exists, and whether they are built to call people upwards or merely to keep them from dying. If we continue to reduce that to maps and methods, we will keep managing decline very politely. I am not interested in managing decline. I am interested in recovery, because sooner or later every serious society must decide whether it still believes people can change, or whether it has quietly decided they cannot.

Perhaps it is no accident that this question lands on Christmas Eve, a night that quietly defies managerial thinking by insisting that real change does not come through systems at all, but through people, patience, and time.

Merry Christmas, and peace to all who are doing the slow, unseen work of helping people change…

10 Likes