Poverty and Opium Poppy Cultivation in Afghanistan: Open Letter to UNODC and Reply

Barnett R. Rubin

On Sunday, January 13, I sent a letter (via email) to Antonio Maria Costa, the administrator of the UN Office of Drugs and Crime in Vienna. In that letter I challenged several assertions in UNODC's Afghanistan Opium Survey 2007, in particular these:

First, opium cultivation in Afghanistan is no longer associated with poverty – quite the opposite. Hilmand, Kandahar and three other opium-producing provinces in the south are the richest and most fertile, in the past the breadbasket of the nation and a main source of earnings. They have now opted for illicit opium on an unprecedented scale (5,744 tons), while the much poorer northern region is abandoning the poppy crops.

Second, opium cultivation in Afghanistan is now closely linked to insurgency. The Taliban today control vast swathes of land in Hilmand, Kandahar and along the Pakistani border. By preventing national authorities and international agencies from working, insurgents have allowed greed and corruption to turn orchards, wheat and vegetable fields into poppy fields.

I have appended the full text of the letter to the end of this post and also posted it here. The main point of my letter was:

I believe that the assertions in the two paragraphs are wrong, not supported by evidence, and are being used in support of a policy that will greatly hinder achievement of the over-riding goals of the Afghanistan Compact, “to improve the lives of Afghan people and to contribute to national, regional, and global peace and security.” The statements also contradict other well known policies of the United Nations: the estimated average per capita income of the residents of Hilmand province, the “richest” province in the supposedly richest part of Afghanistan, is estimated to be $1 per person per day. As you know, the first of the United Nations Millennium Development Goals is to “Reduce by half the proportion of people living on less than a dollar a day.” The United Nations thus has defined the average income in “rich” Hilmand as the threshold of absolute destitution.

UNODC replied by raising a straw man and implicitly admitting that the statement in the Opium Survey 2007 was wrong:

Dear Mr. Rubin,

Thank you for letter (via e-mail) to Mr. Costa dated 13 January. We appreciate the fact that you respect and rely on UNODC's research, and the trouble that you have taken to put forward an extensive letter with a number of thought-provoking points.
We are aware of the policy impact of UNODC's work. We also realize the complexities of promoting both security and development, and the need to eradicate poverty and not just opium.

In due course we will post a discussion paper on the UNODC website (www.unodc.org) that will present some evidence to show that poverty is not the single, exclusive driver of poppy cultivation in Afghanistan. When the paper is uploaded, you may wish to make a link from the UNODC website to your blog.
Mr. Costa will also put forward his views on the subject at the Joint Cooperation and Monitoring Board in Tokyo.

Thank you again for raising these important issues.

Sincerely,

Walter Kemp
Office of the Executive Director
UNODC

I replied:

In your response you say that the discussion paper you plan to post will "present some evidence to show that poverty is not the single, exclusive driver of poppy cultivation in Afghanistan." Such a paper would be welcome, but please note that there is not a single person in the world, including me, who has ever claimed that poverty is the "single, exclusive driver of opium cultivation in Afghanistan." David Mansfield, on whose work for UNODC and others I have relied heavily, argues that the main driver is insecurity, and this is the hypothesis borne out by the data presented in the 2007 Opium Survey.

UNODC's 2007 Opium Survey, however, stated, “opium cultivation in Afghanistan is no longer associated with poverty.” To the average reader this would mean that poverty is not a driver of opium cultivation in Afghanistan. Your statement in your reply to me, implying that poverty is one of several drivers of opium poppy cultivation in Afghanistan, directly contradicts the statement in UNODC's 2007 Opium Survey. I hope your paper will include a correction of the serious error.

The text of my letter follows (PDF version here):

13 January 2008
Antonio Maria Costa
Administrator, United Nations Office on Drugs and Crime
Vienna, Austria
Via electronic mail

Dear Mr. Costa:

I regret that we have not met in over a year, since we testified together at the U.S. House of Representatives Committee on International Relations on September 20, 2006. I am writing now to follow up on an informal query I sent to your office that has remained unanswered, probably because of the informal and indirect way that I sent it.

I am now writing formally to request a response. At the hearing I had the pleasure of meeting the head of your office in New York, Simone Monasebian, with whom I have developed very good cooperative relations. After the publication last summer of UNODC’s Afghanistan Opium Poppy Survey 2007, I wrote to Simone informally to ask her to pass on a query to UNODC headquarters. I asked UNODC to provide the empirical basis on which the Survey made the following statements:

First, opium cultivation in Afghanistan is no longer associated with poverty – quite the opposite. Hilmand, Kandahar and three other opium-producing provinces in the south are the richest and most fertile, in the past the breadbasket of the nation and a main source of earnings. They have now opted for illicit opium on an unprecedented scale (5,744 tons), while the much poorer northern region is abandoning the poppy crops.

Second, opium cultivation in Afghanistan is now closely linked to insurgency. The Taliban today control vast swathes of land in Hilmand, Kandahar and along the Pakistani border. By preventing national authorities and international agencies from working, insurgents have allowed greed and corruption to turn orchards, wheat and vegetable fields into poppy fields.

I have not yet received an answer to this informal query, which, as I noted, could easily have been misplaced. I am therefore writing to explain why I consider this question to be important and to request an answer by January 21, in advance of the February 6, 2008, meeting of the Afghanistan Joint Cooperation and Monitoring Board, which will meet in Tokyo to discuss action on counter-narcotics. I consider this matter important, because these two paragraphs are cited by proponents of expanded forced eradication of the opium poppy crops. I believe that the assertions in the two paragraphs are wrong, not supported by evidence, and are being used in support of a policy that will greatly hinder achievement of the over-riding goals of the Afghanistan Compact, “to improve the lives of Afghan people and to contribute to national, regional, and global peace and security.” The statements also contradict other well known policies of the United Nations: the estimated average per capita income of the residents of Hilmand province, the “richest” province in the supposedly richest part of Afghanistan, is estimated to be $1 per person per day. As you know, the first of the United Nations Millennium Development Goals is to “Reduce by half the proportion of people living on less than a dollar a day.” The United Nations thus has defined the average income in “rich” Hilmand as the threshold of absolute destitution.

In my discussions with policy makers about counter-narcotics in Afghanistan, I have at times encountered a dismissive attitude toward research that does not conform to their policy preferences and “academic” forms of argument in general. But UNODC has an extensive and highly respected research department full of academic experts, for whose work I have enormous respect and on which I have often relied. While policy makers cite this work because it justifies what they want to do rather than because they believe the analysis of United Nations Agencies (you may compare your experience with that of your colleague Mohamed El Baradei of the International Atomic Energy Agency), I believe that I owe your researchers a reasoned response based on data and established principles of analysis. I hope you will bear with me as I proceed through this exercise.

The first paragraph makes two empirical assertions:

1. That “Hilmand, Kandahar and three other opium-producing provinces in the south are the richest and most fertile [provinces in Afghanistan], especially compared to “the much poorer northern region.”
2. That because in the past few years there are trends toward reduction of poppy cultivation in the north and its concentration in southern provinces, “opium cultivation in Afghanistan is no longer associated with poverty.”

Nowhere does the report define what it means by “rich” and “poor” provinces or how this is measured. David Mansfield, a researcher who has worked for UNODC in the past, with his co-author, Adam Pain, believes that the assertion is based on “the finding that households in these provinces reported higher average annual incomes ($3,316 for poppy-growing and $2,480 for others) to UNODC surveyors than those in the north ($2,690 for poppy-growing and $1,851 for others) or centre ($1,897 for poppy-growing and $1,487 for others).” There are many ways in which this data is inadequate as a justification for policies such as requiring forced eradication of poppy crops in insecure areas under Taliban control adjacent to areas of Pakistan where Afghan Taliban, Pakistani Taliban, and al-Qaida are well positioned to exploit discontent in either country. Mansfield and Pain mention some of those technical and academic points such as lack of reporting of sample size and statistical significance. They also note that household income is notoriously difficult to estimate and unreliably reported. I will not repeat their arguments here, though I would appreciate a response to them.

The argument fails primarily because of the ecological fallacy, an error of inference from aggregate statistics that I warned my students against back when I was an assistant professor of political science. The arguments make assertions about the “north” and “south” by aggregating provincial averages for all provinces. Yet is it not true that every province in the south is “richer” even by this flawed measure, than every province in the north. Mansfield and Pain note:

Household data produced by the Central Statistics Office of Afghanistan in 2004 and collected by the 2005 National Risk and Vulnerability Assessment (NRVA) rank the southern provinces relatively low in terms of social and economic well-being. Of the 34 provinces, Helmand ranked 6th, Kandahar 15th, Uruzgan 32nd and Zabul 33rd. The seven northern provinces ranked higher: Jawzjan 1st, Balkh 9th, Baghlan 11th, Samangan 13th, Bamyan 18th, Faryab 25th and Sar-i-Pul 31st. These rankings do not substantiate the argument that farmers in the south are significantly wealthier than those elsewhere in the country. Moreover, in 2005, Helmand reported some of the country’s worst school enrolment rates for children aged between 6 and 13, and one of the highest illiteracy rates. Given the intensity of the conflict in the south, these indicators are likely to have dropped further over the two years since the data were collected.

Let me simplify. Of these eleven provinces, the estimated ranking from most to least well-being is:

1. Jawzjan (N)
2. Hilmand (S)
3. Balkh (N)
4. Baghlan (N)
5. Qandahar (S)
6. Samangan (N)
7. Bamyan (N)
8. Faryab (N)
9. Sar-i Pul (N)
10. Uruzgan (S)
11. Zabul (S)

I am not sure which is the fifth southern province to which UNODC is referring. The inclusion of Farah, Nimruz, or Paktika would not change the overall picture, which is that there is much greater difference in social and economic well-being within both the south and the north than between the two regions considered as a whole.

This fallacy is related to the major conclusion of the paragraph, which is frequently cited by proponents of eradication: that “opium cultivation in Afghanistan is no longer associated with poverty.” UNODC has produced no evidence to support this assertion, and the available evidence contradicts it.

UNODC’s argument is: higher average household incomes across multi-provincial regions are correlated with increased poppy production in those multi-provincial regions. Therefore poppy production is not associated with poverty. Indeed the second paragraph goes even further, stating that poppy cultivation is due to “greed and corruption.” This is a very grave conclusion, with major policy implications, which should not be taken lightly on the basis of flawed data and faulty reasoning, compounded by negative stereotyping. Yet, in my view, this is what UNODC has done.

Decisions about poppy production are not made by regions consisting of several provinces that are closely identified with particular ethnic groups. Nor are they made by provinces. They are made by households. This too is overly simplified, as any piece of farmland may be owned by one family, sharecropped to another, and may employ labor from yet another family. Just as there is greater variation within north and south than between them, so there is greater variation within each province than there is among them. There are many desperately poor households in even the “richest” provinces. Valid inferences about the relationship of poverty to poppy cultivation must be based on household-level data. Research by Mansfield, the World Bank, and others using household level data is quite clear. I am sure that your research department is quite familiar with the research showing that dependence on opium poppy cultivation is highest among the poorest households. To put it statistically, among households, poverty is correlated with dependence on opium poppy cultivation.

Therefore, those dependent on opium poppy cultivation in Hilmand are likely to be the poorer households in that province, those with an income less than one dollar per person per day. Does UNODC consider such households to be rich, greedy, and corrupt because households in Balkh have an average income of only $0.70 per person per day?

The second paragraph is more complex, as it is phrased so that it can be subject to several interpretations. The key sentence is “By preventing national authorities and international agencies from working, insurgents have allowed greed and corruption to turn orchards, wheat and vegetable fields into poppy fields.”

This statement is true in the following sense. As research by scholars such as Francisco Thoumi of Colombia has demonstrated, the cultivation of raw materials for illicit narcotics migrates to those naturally suitable areas that are most insecure. Hence opium poppy cultivation has migrated from other countries to Afghanistan. Furthermore, within Afghanistan it has migrated from the more secure areas to those where the insurgency is more concentrated. The one way that north and south are indeed very different is that the insurgency is much more widespread in the south and security is worse. That, as you know, is due to the geographical position of the southern Afghanistan rather than its alleged wealth.

Insecurity leads to poppy cultivation in part because, as UNODC says, national authorities and international agencies cannot work where it is insecure. As a result, the government and international community cannot provide security and all of the other supportive public goods necessary to agriculture and other forms of employment, such as financing, technical assistance, and marketing. Instead all of these are supplied by the drug industry.

I will make one parenthetical remark here. The U.S. government says that this does not apply to Helmand, which, if it were a country, would be the fifth largest recipient of U.S. aid in the world. This is a very deceptive statistic. What the U.S. government measures is the amount that it has spent (or authorized) for projects located in Helmand. The single largest and most expensive project in Afghanistan today is the Kajaki Dam, located in Helmand Province. The bulk of U.S. expenditures in Helmand are for this project. As you know, despite the hundreds of millions of dollars spent on that project, it is not yet operational. As the people of Helmand have yet to receive any benefits from this project, it is deceptive to characterize them as its “recipients.”

I have no quarrel with characterizing drug traffickers and their protectors as greedy and corrupt. While policy should be based on analysis of what actions are effective, rather than value judgments alone, certainly this characterization of drug traffickers and their protectors provides moral support for effective measures of interdiction. In my discussions with policy makers, however, they have applied these terms to cultivators of opium poppy in Helmand and used the UNODC statement as justification for eradication. Does UNODC consider opium poppy cultivators in Helmand to be primarily driven by greed and corruption?

Of course, even if that were true, it would not be a reason to carry out eradication, as policies should only be carried out if they are effective. That is not the subject of this note. But I would like to point out that there is a relationship between what I consider to be UNODC’s erroneous arguments and policy on eradication.

The National Drug Control Strategy of Afghanistan states that the Afghan government will “conduct targeted and verified eradication where there is access to alternative livelihoods.” Proponents of increased forced eradication have taken the two paragraphs from the UNODC Survey above as evidence – indeed proof – that Helmand province is such a place and that eradication should therefore be carried out there. I do not think that the fact that the U.S. has spent hundreds of millions of dollars on an unfinished project in a province where the average income per person is one dollar a day and where insecurity has prevented the delivery of education and healthcare constitutes evidence that in Helmand province “access to alternative livelihoods” is available. Does UNODC consider that access to alternative livelihoods is available in Helmand Province and that it therefore should be subject to increased forced poppy eradication?

Some policy makers are aware of these problems and try to compensate for them by arguing that eradication will be targeted against the truly rich, corrupt, and powerful. They have not explained to me yet how they will target the rich owners of land cultivated with poppy without targeting their poor sharecroppers and laborers, who will bear the brunt of the cost and have no access to alternatives. As UNODC’s own outstanding research has documented over the years, the opium economy creates powerful ties of dependency between those who control the economy and the poor who are dependent on it.

I believe that the misleading presentation of research by UNODC is providing a justification for a very mistaken and dangerous policy in Afghanistan. I would appreciate any explanation you can provide of why the assertions in the UNODC Afghanistan Opium Poppy Survey are correct. If I receive a reply by January 21, I will post both this letter and your reply to my blog and circulate them to my mailing list. If I do not receive a reply by that time, I will circulate this letter while awaiting your reply.

Sincerely yours,

Barnett R. Rubin