Amir Attaran, a professor of public health at Ottawa, has just published an article in a reputable epidemiology journal in support of licit opium production in Afghanistan. Short version:
- There’s a huge unmet need worldwide for pain relief from morphine;
- So let’s shift Afghan opium production into licit channels, rather than vainly trying to destroy it.
Basically, he makes the same case I did here two years ago (here and here), though with proper references and all. References and peer review are powerful medicine – fair enough – and this time the topic’s been picked up in the high-frequency blogosphere by Charli Carpenter and Matt Yglesias.
Attaran doesn’t cite the 2009 paper by Greenfield, Paoli, and Reuter (GPR below), criticising a similar scheme, and published in another proper academic journal edited by our own Mark Kleiman. So I’ll have a go here at a rejoinder.
From the GPR abstract (footnote):
Legal medicinal opium production is an improbable answer for at least five reasons (their numbering, my reformatting):
1. illegal production will continue;
2. diversion from the legal market to the illegal market is inevitable;
3. diversion will involve further corruption;
4. there may not be a market;
5. Afghanistan lacks the institutional capacity to support a legal pharmaceutical industry.
I have two issues with this: the framing, and a straw man fallacy under point 2 over diversion. Let’s start with the straw man (the variant is attacking a weak version of the opposing case rather than the strongest).
GPRs’ diversion argument is based on single-minded extrapolation from India. Indian technology is old-fashioned. Poppies are induced to seep opium gum by manual slitting of unripe poppy heads. The dried gum is collected by hand and taken to factories for simple processing. Combined with decentralised control, this creates opportunities for diversion of legally grown gum into illicit channels. This happens on a large scale: estimated at 100-200 tons of opium a year, about 30% of production. India is a much more law-abiding country than Afghanistan, where diversion would therefore be greater. QED.
I was clearly wrong to cite India as a model in my 2008 post. (However, the diversion does not condemn India’s industry. The harm done by 20 tons of illegal heroin does not negate, and probably does not outweigh, the good done by twice as much licit morphine). But I gave first place to Turkey, correctly.
Mansfield (2001) gives a detailed first-hand account of the contrasting systems. In Turkey production takes place on the arid Anatolian plateau, also by small farmers, on similarly tiny, one-acre plots. The differences from India are two. First, the farmers must not produce gum but poppy straw (basically the entire dried plant, minus the seeds taken out to sell to pastry-cooks). The straw is harvested crudely, trucked to a single large factory, and processed into morphine, incidentally emitting an ‘unbearable stench’ – a feature that seriously discourages diversion by itself. The developed-country producers – Australia, France, and Spain – also use the poppy straw method, on much bigger mechanised farms.
The statement in GPR that production by the opium straw method is capital intensive is IMHO seriously misleading. It’s irrelevantly true for rich countries. Mansfield states (my italics) that “in Turkey and India, opium poppy is still largely cultivated, as it always has been in a labour intensive way”. So in Turkey, only the processing is more capital intensive.
The other difference in Turkey is that the control system is much tougher and more centralised, though there is local buy-in by the presence of local officials and witnesses. Potential sanctions for violations are severe – one to five years in (Turkish!) prison. The result of all these factors: diversion and illicit production in Turkey are negligible.
It’s not wishful thinking, but a pretty mainstream view, to expect that a switch to opium straw technology combined with Turkish-style controls would seriously crimp the possibilities for diversion. The GPR argument from India is thus valid only against a particular weak feature of one proposal for organising licit Afghan production, that of the Senlis Council, not Attaran or me.
So let’s simply stipulate that the Afghan licit production system should be a clone of the successful Turkish one, not the faulty Indian one. The big smelly factory will have to be paid for and tightly controlled by foreign donors, with the sort of intrusive security you get in a diamond mine.
GPR briefly and superficially consider applying the Turkish model. They dismiss it on the mistaken ground of capital intensity, and a more reasonable pessimism about bringing Afghan governance anywhere close to Turkish standards of control at village level.
This brings us to the broader question of framing.
Is the goal pain relief, COIN and development, or drugs control?
What’s the problem to which for GPR “this isn’t the answer”? Theirs seems to be the flow of illicit opium from Afghanistan into the world market, on the way supporting the Taliban’s insurrection. So their frame is “drug control + COIN and development”. Mine and Attaran’s is primarily pain relief: the scandal of millions of poor Africans and Asians dying of cancer and HIV without the morphine you and I can expect to ease our own passing; and secondarily finding some honest work and income for Afghan peasants, independently of Taliban predation. Our frame is “pain relief + COIN and development”. The contrast in framing probably determines your response to these arguments.
GPR’s point 4 illustrates this perfectly. They take the world licit market for medical opioids as given, therefore an obstacle to any Afghan initiative. But the first premise Attaran and I start from is that this administered and cartelised world market is broken, and that major initiatives are required to jump-start demand and provision in poor countries. Quite why their governments put in such ridiculously low needs figures to the INCB is not clear; perhaps their doctors have bought into the drugs war rhetoric too much. Whatever the reason, if you start from meeting the basic human need to to be free from great pain, you must think of a massive expansion of the world opium supply – up to sixfold, by my back-of-an-envelope calculations. This would initially have to be underwritten by rich-country governments and charities. The planned expansion is also politically vital to outflank opposition from the existing licit growing cartel, crucially including Turkey, whose technical assistance would be essential.
Once you posit a big expansion in licit supply, the secondary question arises who should produce it. There’s apparently plenty of room for Afghan farmers as well as existing producers; why not give them a chance? After all, they already know how to grow the stuff.
You might reply: the Afghan situation, with massive failures of security and governance, is so bad that you would only make things worse. This is the essence of GPR’s argument apart from the fallacious Indian analogy. And of course there’s a lot of evidence pointing their way.
Their case is strong but not conclusive. I’ll concede their point 1, that substitution would probably not by itself cut illegal production and Taliban funding – though since there’s no alternative way of achieving this goal, we might as well forget about it as a discriminator between projected initiatives. I don’t accept the story of inevitably massive diversion (2) and accompanying corruption (3) for reasons already given, but there would be some. (What would the Taliban do with a hijacked truck of opium straw? Sell it again to the government with false papers perhaps; corruption but no diversion.) However, to pay off overall the strategy only needs to provide a relative gain in the licit economy and non-corrupt part of government, even if it does nothing to reduce the absolute scale of illicit opium smuggling and the flow of rents to the Taliban, warlords, and corrupt officials. So why not take the gamble?
The tone of the paper reflects a deep pessimism about Afghan governance. But if Afghanistan is a failed state that cannot manage any important task with a modicum of effectiveness and honesty, the inference is that the war is lost and the foreign coalition should get out. That may well be true. But the Obama Administration and its allies have decided to stay in militarily. Therefore their civilian policy must be based on the presumption that an effective non-Taliban government in Afghanistan is possible. The situation is terrible, and none of the the available strategies is remotely assured of success. The test for a policy initiative is not then “more likely to work than not” but “having a better chance than what we’re failing at now”. Trying to copy the Turkish system of licit opium production, known to work fine in a poorish country, therefore looks to me a reasonable flutter with a limited downside and a lot of upside.
So here’s my Mark 2 proposal:
- A group of rich countries to declare their support for a rapid and controlled expansion of world licit morphine delivery, and push through the necessary changes in quotas and interpretation under the UN convention.
- An angel – a rich country like Norway or a philanthropist like Bill Gates – to finance an initial stock of 10 tons of morphine (ca. $100m delivered to the pharmacy shelf, at 1c per mg – but a single bulk buyer would surely pay less). This represents
1003,000 tons of opium straw.
- A group of poor countries to commit to expanding delivery of painkillers to their citizens, with the support of the aforesaid angels.
And finally the Afghan icing on the poppyseed cake:
- The USA and its allies to fund a village scheme in Afghanistan on Turkish lines for the licit purchase of the
1003,000 tons of opium straw in reasonably secure areas, and one pilot-scale smelly high-security processing plant at Bagram.
[Update: my mistake. It would be ca. 100 tons of opium gum, but straw is 30 times less concentrated, according to Wikipedia.]
Point 5. GPR score a debating-point win against a particular, over-optimistic claim by the Senlis people that Afghanistan could produce other bio-pharmaceuticals. A single foreign-run opium-straw factory, which is all I’m proposing, is not a whole pharmaceutical industry and would not lead to one. It hasn’t done so in Anatolia.
A sixth negative argument has been made to me by Mark Kleiman, which is IMHO stronger than any of GPRs’: if your policy is crop substitution, why not spend the money encouraging apricots or currants or other unproblematic crops? To which I would only reply: why not try both? It can’t be that a range of crops are equally suited to the same pieces of land over large areas. The evidence that Afghan peasants can successfully grow opium in current chaotic conditions is that they do so; these conditions are similarly evidently unfavourable for apricots etc, since they aren’t growing them so much. As a substitution crop, licit opium certainly has some significant and specific additional costs and problems, but also some possible unique benefits. Apricots have their own less dramatic pros and cons. Only experience can tell us which substitution works. Which brings us back to the need for pilot projects not theorising.
I don’t have access to the final version of GPR, only a draft, so I paraphrase rather than cite. The passage I do cite from the abstract is the same as a paragraph in the draft, so I’ll assume the changes are nor significant. Apologies if I’m wrong.