Why the Economics of Southwest Kyrgyzstan Matter, And Afghanistan

by Joshua Foust on 11/9/2011 · 14 comments

At my day job with the American Security Project, I recorded a podcast with Adjunct Fellow Nick Lockwood, expert on stabilization operations, population engagement and strategic communications. He travels routinely to Afghanistan, and more recently to places like Libya. The topic was primarily about my current research on the economics of southwest Kyrgyzstan, and why that matters for more than just Kyrgyzstan:

American Security Project podcast

The end of that podcast focused on the travails of Maj. Gen. Peter Fuller, who was fired last week for his intemperate remarks. I think his firing was the only good thing to have come out of it, and I explained why in much more detail for The Atlantic:

The problem with defending Fuller’s remarks, however, is two-fold: they were not only an inaccurate description and analysis of the politics of Afghanistan, but they actively undermined the U.S. mission and strategy. Far from being a “truth-teller,” Gen. Allen made a serious error of judgment, both in how he understood Afghanistan and how he chose to express that understanding.

And even from a basic understanding of politics, Gen. Fuller’s remarks were counterproductive. Gen. Fuller’s remarks that Afghans are not grateful enough for the money the U.S. has spent on their army is deeply insulting. No matter the political system, having a foreign general berate an entire political class for their ingratitude is counterproductive and possibly even poisonous for future collaboration — the most guaranteed way to undermine the mission. Basic human empathy can elucidate this very simple fact. Politics is as much as not speaking unpleasant truths as it is speaking pleasant untruths. It is universal to all political systems.

The U.S. strategy is based on supporting and building up the Afghan government. Publicly insulting them undermines that strategy is reason enough to fire Gen. Fuller, regardless of the anything else.

Anyway, there’s a lot here to discuss, I think. Fire away!

Subscribe to receive updates from Registan

This post was written by...

– author of 1848 posts on 17_PersonNotFound.

Joshua Foust is a Fellow at the American Security Project and the author of Afghanistan Journal: Selections from Registan.net. His research focuses primarily on Central and South Asia. Joshua is a correspondent for The Atlantic and a columnist for PBS Need to Know. Joshua appears regularly on the BBC World News, Aljazeera, and international public radio. Joshua's writing has appeared in the Columbia Journalism Review, Foreign Policy’s AfPak Channel, the New York Times, Reuters, and the Christian Science Monitor. Follow him on twitter: @joshuafoust

For information on reproducing this article, see our Terms of Use


Linar Zairov November 9, 2011 at 6:52 pm

Mr. Foust,

It is rare to hear a passionate yet sensible narrative about the challenges that Uzbeks are facing today in the Kyrgyz Republic. However, your analysis seems to be based on questionable data and heavily influenced by stereotypes. For example, your claim that pre-conflict distribution of businesses in Osh was 90% owned by Uzbeks is inconsistent with the facts from representative surveys on households. In fact, your premise that Uzbeks were wealthier than Kyrgyz in Osh is wrong. Several studies based on households surveys for the years 1993 to 2005 clearly indicate that urban Kyrgyz in Osh were wealthier than Uzbeks. However, the difference between urban and rural population was substantial. And, of course, echoing claims that Kyrgyz are less capable to cook or less able to cut hair are simply silly and odd (in contrast with your other serious analytical pieces).

Apart from my annoyance with your inaccurate use of publicly available data, I found your honesty refreshing. In particular, your admission that this particular narrative maybe an artifact of a sample selection bias is the best thing I’ve heard from any western journalist who had anything to do with Kyrgyz Republic. As for your argument that so many journalists and organizations cannot be wrong: if selection bias is pervasive, then it is very likely that most if not all journalists covering the issue are wrong (especially if they use similar channels).

Joshua Foust November 9, 2011 at 7:28 pm


Thanks for the comment (and call me Josh!). I think (I hope?) I pointed out that I was trying to confirm the 90% number. That was given me by two Uzbek businessmen, one in a business association and the other at a microcredit institution. It is not based on a survey but rather perception.

And that was the biggest flaw in this trip. I did not have good access to Kyrgyz businessmen, and the Mayor’s office and local government declined to speak with me. So I got a highly distorted version of both recent and not-so-recent history — like when one Uzbek told me there were no Kyrgyz in Osh in 1990. It’s clearly untrue, but that’s what people say and it’s what they tell each other.

The challenge with the type of trip I did — only two weeks there — is that I can’t develop a perfect picture of what *really* happened. I can, however, get a sense of what people seem to accept as common or accepted wisdom. And there is value to that, I think, so long as it is described as such and isn’t used to make wildly unsupportable policy recommendations.

That’s all I could do anyway. I did not perform a survey, I did not attempt to assemble an oral history, I did not expect to get a perfect timeline of events, and I do not expect to emerge with anything more than a clarified idea of what the major issues are in this region.

I hope that’s what I got across. And this project is still ongoing! So I hope you can stick with me while I compare my interview notes and recordings against the public data I have sitting in piles on my desk, and then give another round of feedback.

Oh, and I am not a journalist.


Linar Zairov November 9, 2011 at 7:50 pm


Thanks for the reply. I was not attempting to discredit your work, but rather comment on things that can be improved. Primary reason for my comments were general annoyance with journalists in that part of the world who operate on a limited information set and have no willingness to accept the possibility of being wrong. Against that background, you are Adriana Lima of analysts. Now that I know you are not a journalist, that makes more sense.

As for works I mentioned, I can’t recall all of them but I recommend looking into Professors Anderson and Pomfret in UK (they have several papers on economic well being in KG), a more recent piece by Esenaliev and Steiner over at German Institute for Economic Research in Berlin, and a good background read by Abilabek Asankanov covering the events in 1990.

I love this quote from Baby Einstein videos: “All truths are easy to understand once they are discovered; the point is to discover them” G.Galilei. I think the issues you are covering require deep and careful search or discovery. Once uncovered, I am sure they all will make perfect sense.

Joshua Foust November 10, 2011 at 8:00 am

Thanks. I found “Consequences of Creating a Market Economy” and I’ll see about getting it without going bankrupt (I hate academic books for that, btw). I also found a publication that seems to more directly address what I want this to turn into, a World Bank piece on spatial disparity and development policy by Gudrun Köchendorfer-Lucius and Boris Pleskovič.

Many thanks for the recommendations!

Linar Zairov November 9, 2011 at 7:56 pm

Although Balihar Sanghera’s pieces are too socialist for my taste, his work (academic and non-academic) is intriguing, insightful, and deserving of careful consideration.

Will November 14, 2011 at 4:26 am

Mr.Linar Zairov has been tirelessly working since June 2010 posting on blogs and twitter to prove how uzbeks were the instigators of June violence. If you are really an honest person, you wouldn’t be criticizing every piece of writing about the June events taking only one side (kyrgyz).
You twist your facts as long as it fits your goal. Based on “household surveys”, which you deliberately fail to cite, you refute the claim that uzbeks were wealthier than kyrgyzs. But in your comment on RFERL (http://www.rferl.org/content/Getting_To_The_Roots_Of_Resentment_In_Kyrgyzstan/2090331.html) you indicated ONE major factor for the 1990 and 2010 Osh events –“a large income inequality between the majority ethnic group and minorities”. Take your another argument there, again you also refer to a hypothetical statistics without citing it, in which you claim that uzbeks do not hold public offices because they are bad at fulfilling their duty, and military service is economically inferior to farming and trade. These are foolish arguments at best. Just look across the border and compare the kyrgyz and uzbek militiary servicemen. Failure to report to a duty is the result of a lack of discipline, not because someone is uzbek or kyrgyz. Also, your argument that farming and trade is economically inferior does not hold water. Now you may think by using terms like “sampling error”, “sample selection bias”, “statistical representation”, etc. you can easily fool the readers. First of all, this is a blog, not an academic journal. Even if there are publicly available data (which you fail to cite), I would take it with a grain of salt based on my experience of conducting surveys in the region. Secondly, there are obvious facts for which one doesn’t need to have a statistical survey. Take, for example, your rebuttal of the claim that kyrgyzs are less capable to cook or cut hair, which is simply out of your frustration, as anyone who visits Osh, even the local kyrgyzs if not blind like you, would notice that these spheres are occupied mostly by uzbeks. As odd or silly it may sound to you, you have implicitly stated almost the same thing on your RFERL post I cited above. Based on your many other posts on other blogs, which are mostly contradictory and one-sided, why I should believe you or your hypothetical “surveys”?

Linar Zairov November 15, 2011 at 2:19 pm

Mr. Will,

If you paid attention to my comments on the issues pertaining to Osh and Jalalabad events, you would have realized that I never argued that Uzbeks are to blame. I have, and still do, argue that jumping the gun and blaming the entire ethnicity and the state for violent murders is not only irresponsible but factually inaccurate. The facts from the HRW indicate that the first victims who were murdered were Kyrgyz. The facts also indicate that out of those recorded and examined deaths, 2/3 were Uzbek and 1/3 were Kyrgyz. Although it is true that majority of dead were Uzbek, one cannot ignore the deaths of innocent children,women, and elderly on the Kyrgyz side as a mere act of defense. My point has been and is that the causes of the violence in Osh and Jalalabad are more complex than what people like you have made it to be- Kyrgyz, who are uniformly inferior in ability to produce, murdered Uzbeks. This line of argument may work well for ‘journalists’ over at Fox News, but a sensible and analytically minded person recognizes that such stories do not fit the facts.

As for my conjecture that income inequality, it has to do with rural Kyrgyz incomes (which make up the majority in the south) and urban Uzbeks. When you compare the unconditional income gap between Uzbeks and Kyrgyz, as has been reported by Esenaliev and Steiner using the 2005 nationally representative survey, the difference actually favors Kyrgyz. I do not know which surveys you conducted, but those commissioned by World Bank and carried out under the supervision of North Carolina State University statisticians, is as good as any US survey.

My point has been and will continue to be that a simple finger pointing and stupid claims that there was a planned and systematic execution of Kyrgyz citizens of Uzbek origin is false. There are more murdering Kyrgyz bastards than Uzbeks who continue to walk free, but it doesn’t mean that the entire ethnic group carries the responsibility for their actions. It shows the failings of the judicial system, governance, and rule of law. One has to be careful in making inference from a small set of observations. In that respect, my comments about sampling bias, representativeness, and sampling error are very much important (albeit, not popular) additions to the debate. If you can’t admit that your claims based on a few observations can be a product of biased sample, you shouldn’t be in the business of analyzing complex social issues.

Once again, all I hear from your writing is “You are defending Kyrgyz; Kyrgyz are to blame; Uzbeks are more skilled; Kyrgyz don’t know how to cut hair; You are using scientific terms; Blah blah blah, I don’t like you.” What evidence or reason do you have to back up any of your stupid claims? How certain are you about these claims? My writing has been about these questions. If you don’t like it, then join the club of haters. I am interested in identifying real causes and real solutions, not bitching. Afterall, I get to live in that environment and you get to be an asshole who runs his mouth, because you don’t have to back up anything you say.

Linar Zairov November 15, 2011 at 6:52 pm

Here is the link to Esenaliev and Steiner paper: http://www.econstor.eu/bitstream/10419/48342/1/75_steiner.pdf
Unconditional consumption is slightly higher for kyrgyz compared to uzbeks, and the reverse is true upon conditioning. both results are statistically insignficant, meaning the uncertainty around the estimates are too big to make a claim in either direction. for all practical purposes, there appears no substantive difference in well-being of uzbeks relative to kyrgyz.

Will November 16, 2011 at 2:01 am

Let me put the record straight from the beginning: I am not blaming the entire kyrgyz nation. What I am trying to get across you and the like-minded kyrgyzs is that denial is not a solution to the problem. You seem to refer to sources that agrees with your thinking, but totally ignore other more reliable sources. HRW has never conducted an independent investigation and their “facts” in their initial report you are referring, while might be true, were driven by word of mouth.

As for your inferiority complex, I did not say one ethnic group was uniformly incapable of producing anything. Now tell me why one ethnic group is disproportionately more represented in certain spheres of business than the other group? What kind of facts do you need?

There was nothing about income inequality between urban and rural groups in your initial comment on RFERL. I simply indicated two of your contradictory opinions. You initially claimed that the violence was solely due to the unequal wealth distribution between the two ethnicities. Here you are stating that urban kyrgyzs were actually better off. So chill out.

Thanks for the link to the paper. Now if you care about statistical flaws, you would have been more cautious about the findings of the unpublished manuscript rather than taking everything on faith. By quick review of the paper I got tons of questions. In order not to sound too critical of the paper, I give two reasons for my skepticism. First, there is a heavy measurement error in their wealth estimates. They rely on wealth measures which are likely under-reported by not including other significant unofficial sources of income. Secondly, I didn’t see anywhere in the paper how they tackled the endogeneity problem which will render all of their OLS results biased. These two are enough not to take the findings seriously.

Instead of name calling, do your homework first otherwise f*** off.

Linar Zairov November 16, 2011 at 1:21 pm


I was wrong to call you ‘asshole’ for two reasons. The first reason is it is inappropriate. The second reason is I realized that it is a wrong term because asshole is an inherent quality. A better term is ignorant. And here are the reasons why.

Your criticism of Esenaliev and Steiner paper missed the essence of their paper. They recognized that any measure of well-being will be marred by measurement error. For that reason, they propose three different measures, one of which is wealth index. It is a standard practice in all well-being studies to use consumption per adult equivalent or per adult member or per member. Given the consistency across three different specifications, they report that statistically the gap is indistinguishable from zero. As for your argument of endogeneity, ethnicity cannot be endogeneous in that specification. Endogeneity requires that Cov(Ethnicity,errorterm)!=0. What a priori theory or evidence do you have that ethnicity is endogeneous? In simple terms, how does wealth or consumption influence ethnicity? It is as exogeneous as it can be. And if it is not, what evidence or theory do you have that the magnitude or direction of bias is substantial? If you can identify that link, I am sure you will be famous. Just because you’ve learned some statistical terms in your summer camp for dummies, doesn’t mean you can make sense of them. If you are not happy with their paper, read Anderson and Pomfret papers, which were published in reputable and peer-reviewed journals. Their conclusion is consistent with the Esenaliev and Steiner paper. The most disturbing part is that you keep bitching about these papers, yet you propose not a single study that show evidence to the contrary. Bitching is easy, we all know that.

The other issue is your dissing of HRW or Kimmo the Finn’s report. What evidence do you have? I didn’t make these numbers up. The number of dead was counted not on the word of mouth but by autopsies and coroner’s examinations (exhuming dead bodies). I have stated before and will say it again: There were dead on both sides, the ratio was approximately 2/3 uzbek and 1/3 kyrgyz. Among the initial hospitalization reports the first 52 hours, a majority were kyrgyz young, old, children, and adults. These numbers are inconsistent with the story that the state, the people, the entire kyrgyz ethnicity had planned and systematically murdered their fellow uzbek citizens. What these numbers show are (1) sequence of events (2) brutality of murders and (3) chaotic nature of violence. How does my statement of facts and questioning of claims that there was a systematic and planned execution of Kyrgyz citizens of Uzbek ethnicity? I have argued before and argue again is that the causes of these events are not as simple as Kyrgyz are murdering assholes and Uzbeks are victims. This claim does not follow from the evidence. I do not defend murderers. I am from a mixed family. I do however defend the right of Kyrgyz citizens of any ethnicity to know the truth of what actually had happened and why.

As for your claim that because most of the hair cutting places and restaurants are owned by Uzbeks in Osh, it means that Kyrgyz are less capable of cutting hair or cooking, I will say again- it is a stupid stupid stupid argument. In contrast to you, I will explain why it is stupid. For starters, this argument is non-sequitur. Ownership of businesses by one ethnicity doesn’t imply that the other ethnicity is less capable of doing it. According to your logic, because majority of cotton or tomato pickers in the US are Mexicans, white Americans or blacks are less capable of picking cotton. If you don’t like it, place any of your favorite examples and observe absurdity of your claim. Second, a more sensible reason is that internal migration of skilled workers was affected by existing hardships after 1991. A great deal of skilled labor from the south migrated to the north, specifically Bishkek. The disproportionate number of internal migrants to the north were kyrgyz. Skilled Uzbeks migrated from rural areas of the south to Osh and Jalalabad. Bishkek has many restaurants and cafes, hair cutting establishments that have cooks and stylists from the south who happened to be Kyrgyz. As for who makes a better dish, it has two answers: 1. If you are racist (and I think you are), then your mind has been made up 2. De Gustibus Non Est Disputandum.

As for my conjecture that income inequality matters, I never reversed it. I do think that it was ‘a’ factor, and an important one at that, in triggering violence. The murdering bastards of Kyrgyz origin in Osh were from rural areas. You do not see reports where local Osh Kyrgyz engaging in muder spree (surely, there were some thugs). In fact, there are many reports of how Kyrgyz sheltered their fellow Uzbek neighbors from those angry thugs.

You speak of one-sided arguments and findings that cannot be taken seriuosly, but you provide not a single reference to a replicable or defensible study. If we are to have a civil debate, please provide arguments and evidence to prove me wrong. Ah, my argument is the causes of violence are complex and are not consistent with the claim that there was a planned and systematic execution of Uzbeks by Kyrgyz ethnicity and the state.

I did my homework. I do it well. And I don’t need an ignorant person to tell me that I am one-sided. I will listen to your arguments if they are indeed arguments and not empty claims. Again, provide us with evidence, evidence that, as you so heroically say, “can be taken seriously.” otherwise eat s..t!

Stefan G November 17, 2011 at 6:03 am

At risk being called an asshole or told to eat shit, or something along those lines, there is one little detail I would want to take issue with.

“Among the initial hospitalization reports the first 52 hours, a majority were kyrgyz young, old, children, and adults. These numbers are inconsistent with the story that the state, the people, the entire kyrgyz ethnicity had planned and systematically murdered their fellow uzbek citizens.”

This is true, but it mainly reflects the fact that many Uzbeks were too afraid to seek medical treatment. I don’t think it particularly useful in determining causality, which is in any case a pretty misguided effort when couched in terms of “they started it.”
Also, I don’t think anybody seriously (i.e. anybody credible) argues the state of Kyrgyzstan and the Kyrgyz people waged a premeditated pogrom or ethnic cleansing. Still, it is worth recalling that it was more than 2/3 of the casualties that were Uzbeks; the figure is closer to 3/4. And let us not forget the disproportionate loss of property and subsequent police harassment.
The hairdresser thing, I cannot speak to, but I have heard many local people complaining that you couldn’t decent grub for ages, because all the best restaurants were run by Uzbeks (or Uyghurs for that matter, a group that got lumped in with the Uzbeks in the zakhvaty that followed). I don’t see that this issue should be matter for heated argument though.
It is in the interest of vested criminal elements to have people squabbling fruitlessly over these issues, because that means nobody will realize and pay attention to how some gangsters have actually enriched themselves as a direct consequence of the Osh events. A community Balkanized over what was a collective tragedy is bad for equitable economic recovery and growth.

Will November 17, 2011 at 2:48 pm

Linar the impulsive and ignorant, this is my final post on this topic. If certain businesses are run mostly by uzbeks, that doesn’t mean kyrgyzs are incapable of doing things. I made this clear, but for reasons not clear to me you keep assuming things that I had never said or implied. The division of labor may be the result of cultural and traditional differences. Uzbeks as city dwellers traditionally controlled these spheres, apparently they (along with uyghurs) have the best cuisine in the region. Although this may have been the source of envy by rural kyrgyzs, I agree with the previous post that it is not something that we should be arguing about.

This is not the appropriate place to discuss someone else’s paper (from your reaction I am thinking that it might be yours ). But let me point out this. Did you pick “ethnicity” out of about two dozen other variables just to prove me I am wrong? Just because table 7 doesn’t show other variables, it doesn’t me they are not included in the regression (read the footnote). If any one of these variables is endogenous, all estimates are biased. Keep this in mind, it may save your bu** in the future.

Just because the first victims at hospitals were kyrgyzs, that doesn’t prove uzbeks were instigators. I agree with the above post, uzbeks MAY have been afraid to seek medical treatment.

You should seek taking anger management class.

Linar Zairov November 17, 2011 at 4:33 pm


Point well made. I agree that most people who go to Osh have a sense that good food is made by Uigurs and Uzbeks. That is not something one can question. What I have questioned was the non-sequitur argument that observing good Uzbek or Uigur cooks in Osh implies Kyrgyz are less skilled to do that job. It is a market place, where people allocate their labor across different jobs in different geographic points. My point is within each sector at a specific geographic point there are people who are more skilled and those who are less skilled. But observing people with lower skills at a particular geographic point having traits “A” do not imply that A are less skilled in a specific activity. Best basketball players in the early NBA years were Jews. Observing that does not imply that blacks or whites were less skilled. It meant there were other opportunities either more appealing or less restrictive for whites and blacks at that time.

I do not call people asshole unless it is the most appropriate term (obviously I was wrong to use asshole). In the case of mysterious Will, he accused me of one-sidedness and nationalism, which are false and offensive, without providing a single sensible evidence or argument. I have been arguing to move the discussion away from simple stories such as Kyrgyz engaged in ethnic cleansing of Uzbeks to a more realistic, factually consistent discussions that may result in real solutions. I have emphasized the role of biases that each one of us have and biases arising from limited channels we often use to obtain ‘evidence’. The matter in question is a complex one, not a simple ‘bad people vs. good people’.

I have never argued that Uzbeks were instigators or Uzbeks are to blame. That is an artifact of hating idiot’s imagination. I have stated that evidence does not support a planned and systematic execution of Kyrgyz citizens of Uzbek ethnicity by the entire Kyrgyz ethnicity and/or the State. Who started it or who is to blame is not what interests me. What interests me is what factors drove ‘seemingly normal’ people to murder each other at a specific place and a specific time. More importantly, what can be done to prevent re-occurrence of such tragic events.

As for Will, my problem is not with him (because I don’t know if this virtual fellow is Will or Jane or Janybek or Rafik). I have a problem with his bitching and accusing me of denial of horrific deaths of my fellow Uzbek citizens. The fact is I care more about my fellow citizens than this virtual critic will ever care for citizens of Kyrgyz Republic. I do not appreciate half-wits who bitch about other’s arguments or works without providing a single argument (not a claim!) or evidence to the contrary. To those classifying anyone who disagrees with a simple generalization as a nationalists or supporter of thugs, I will continue to recommend to consume human fecal matter any day.

Linar Zairov November 17, 2011 at 6:58 pm


Correction: The actual numbers are closer to what you have stated earlier. There were total of 426 dead. Of those identified corpses (381), 276 were Uzbek and 105 were Kyrgyz (http://reliefweb.int/sites/reliefweb.int/files/reliefweb_pdf/node-399972.pdf).

Previous post:

Next post: