Tyranny of numbers

Claims of rising poverty in Iran

Posted in Poverty by Djavad on March 30, 2009

“Give me a poverty scenario and I will produce the numbers to support it,” is a saying among those who work with poverty numbers denoting the flimsiness behind the science of poverty analysis.  In this season of presidential elections in Iran, a scenario much in demand is that poverty has increased under Ahmadinejad.   There are newspaper reports of research that offer evidence for just such a scenario (for examples, see reports in Persian here, here, and here), that seem influential but have not gone through the usual academic scrutiny.   A few months ago I commented on another high profile poverty report that appeared last year in a journal published by the Central Bank of the Islamic Republic of Iran, using faulty methodology to show that poverty has increased.  A study by a serious researcher, Professor Davoud Souri of Sharif University of Technology in Iran, was published last week on the prominent Persian website Rastak which is dedicated to “free market economics”, is a notch above the rest in academic rigor and therefore worth a closer look.  

He estimates that more than one-third of urban Iranians were in poverty in 2007 and, more shockingly, that this rate has increased during 2004-2007, the first three years of Ahmadinjad’s administration. I say shockingly not because Mr. Ahmadinejad had promised to eradicate poverty–that was  harldy in the cards–but because in these four years Iran received about $200 billion from the rest of the world, some of them poor countries, in oil revenues.  To learn that this inflow of money (nature’s gift) not only did not lift anyone out of poverty, it actually made the poor poorer is indeed  shocking.    

Professor Souri is a knowledgeable econometrician and knows his data well, but there are reasons why his study of poverty in Iran, like many others, should not be taken at face value.  Let us look a closer look at how he arrives at his conclusions.

His first conclusion about high incidence of poverty is really not much of a finding because he pretty much assumes it when he defines everyone under $10 per day ($4.40 in rural areas) as poor.  (I am converting his 35159 rials per person per day poverty threshold for urban Iran by the World Bank PPP conversion factor for rial of 3355 to arrive at $10 per day in PPP dollars.)  This is a lofty standard to which no developing country has been held as far as I know.  It is 2/3 the poverty line in the United States and more than three times the threshold international agencies use to compare countries (the so-called $2 per day).  Another widely reported study uses a poverty line of nearly 8 million rials for a family of five, which translates into $16 per person per day, which is higher than the US poverty line! The problem with these studies is not their very high poverty thresholds, it is that they fail to warn their readers about how their poverty lines compares with those used in other countries.  

For more details on the study of poverty in Iran see my working paper, which was published in a shorter form this month in the Journal of Economic Inequality. I have updated some of the findings in that paper here

Publishing poverty results that use poverty thresholds that are not comparable across countries can confuse international readers and convince unsuspecting journalists in the west, as well as some with an ax to grind, that Iran’s economy is a basket case.  A recent case in point of the latter group appears in the Nowruz address by Israeli president Shimon Perez to Iranian people, in which he said: “I see the suffering of the children [in Iran] and I ask myself, why? This is a country that is so rich…You can’t invest the money in enriched uranium while telling the kids to stay a little hungry and a little ignorant.”  Where he sees the suffering of Iran’s children he does not explain; perhaps he is surmising it from studies that show poverty in Iran on a grand scale!

The stronger point in Souri’s study is that poverty has increased during 2004-07.  This finding should disappoint anyone who voted for Ahmadijead as a leader who would do something for the poor.  It should anger people in oil importing nations who paid through the nose for Iran’s oil in recent years, that the country took $200 billion from other (sometimes poorer) countries only to impoverish its own poor. Is the economic system in Iran so broken that its richer citizens are not satisfied with the $200 billion they pocket from oil revenues and have to rob their own poor? A legitimate question, except that its premise of rising poverty during the recent oil boom is not true.

Stay tuned for an explanation of why the rising poverty scenario claimed in these studies is a statistical artifact.

15 Responses

Subscribe to comments with RSS.

  1. Jermaine said, on May 7, 2013 at 1:11 pm

    I must thank you for the efforts you have put in penning this website.
    I really hope to see the same high-grade content from you later on as well.
    In truth, your creative writing abilities has motivated me to
    get my own, personal blog now 😉

  2. Majid said, on May 19, 2009 at 7:02 pm

    Thank you. I’d like to receive updates from this blog through RSS feed. Is that possible?

  3. Babak said, on May 16, 2009 at 5:28 am

    I am puzzled: if you use PPP to convert the poverty line from rials to read $10 a day, why should it be different from the one used in the US? Doesn’t PPP provide an equalizing measure for comparing prices between the two countries? It strikes me under a PPP conversion one should use the same poverty line, no?

    (I have no formal economic training, so if the answer is obvious, bear with me!)

    • dsalehi said, on May 16, 2009 at 10:10 am

      A good question. Yes, with PPP, $10 buys the same in both countries. Now it is your choice as a policy maker or social scientist if you want to keep the line for Iran the same as the US. US per capita income is more than 6 times that of Iran (in PPP $). It is morally acceptable (to me) that US citizens be willing to help people with less than one-fourth of median incomes. By the same token, I would have to say that Iranians are also morally obligated to help out those with 1/4 of their median income. That would be less that $2 per day. Since their productivity is much lower, Iranians should not and indeed cannot support everyone with less that $10. Any government that promised that would fail and any political activist who eggs on his or her candidate to make such a promise would be setting him up for failure to deliver on his promises.

      • Babak said, on May 16, 2009 at 8:03 pm

        Ok, I see your point. That is valid if you were treating poverty as a notion relative to other members of the *same* society. It makes sense for domestic policy making. But since you are also comparing across countries (for instance when you quoted Shimon Perez) I think an notion of poverty is needed that compares members of *different* countries. Then using PPP makes sense but I think one must use the same poverty line for those countries. If not, then one would miss rising poverty, or misery if you will, even if the GDP falls so low that *no one* in that country would be able to live without sustenance.

        Another thing: do these studies use the similar thresholds of poverty for the past? I ask this because the claim of rising (or any change) in some quantity needs to compare the said quantity at two different times.

    • dsalehi said, on May 16, 2009 at 9:23 pm

      Comparisons poverty rates between US and developing countries are not very informative, but between developing countries are. So, comparing Iran to Turkey and Mexico makes sense but not with US.

      Over time, good studies keep the poverty line constant. Changing it over time confuses the issue.

  4. Mohammad said, on April 9, 2009 at 4:14 pm

    It was great. I am a PhD student (engineering) in Canada (1 year). and before I got here, I thought here all people are very rich or there is no poverty according to some bloggers or mass media. But now I really feel Iranian people are not as poor as what is said. I wrote a paper about economic and social situation of Iranian people for my friends in Iran and showed them that the economic situation of them is not very bad and although it is not as good as here but it is comparable. But they told me your numbers are not correct while I used World Bank numbers. However, I told them I see Canadian people income and expenses and I really feel it. Because they are against AhmadiNejad or Akhoond, they cannot see reality. As I saw your posts here, I want to tank you for confirming some of my analysis as a economy faculty living abroad. I like to follow your blog to be academically familiar with the economy.

    • dsalehi said, on April 10, 2009 at 1:36 am

      Thank you for your comment. I know what you mean about people not believing your number; I hear the same criticism all the time. It is difficult to have a good debate when beliefs are strong but facts weak. The economy has become too complex to understand without some knowledge of economics and good data, but this does not stop people from expressing strong opinions in the face of facts. It will take time and effort to raise the level of the debate from, say, what is the true level of the rate of inflation or poverty to what causes them to be high.

  5. dsalehi said, on April 7, 2009 at 8:43 pm

    The 3355 rials is for 2007, of course; in 2009 it could be as high as 5000. I am sure there is room for error, but the World Bank PPP rate is not as nonsense as you think. To appreciate why $1 is worth quite a bit less than 10,000 rials (not to mention the 20,000 you prefer) consider what $1 would buy in the US–practically nothing. If you tip someone $1 in a hotel or a restaurant they would be insulted; you cannot buy a loaf of bread or a glass of coke for $1; visiting a doctor costs on average more than $100 (though usually 2/3 is paid by private insurance); and tuition in a good private college is about $35,000 a year. Loosely speaking, PPP is that exchange rate that would make these prices, on average, more like Iran’s prices. Those familiar with prices in the US would tell you that 20,000 rials buy a lot more in Iran than what $1 buys in the US.

  6. Amiraz said, on April 7, 2009 at 5:32 pm

    The world Band PPP of Rls 3350 is realy nonsese. So is your number of $10. The dollar shasow price is more than Rls 20000. Souri’s numbers make more sense for the people more close familiarity with Iran.

  7. alihme said, on April 2, 2009 at 12:44 am

    Usually, poverty studies in Iran (if you are lucky enough to find out anything about their method of calculation) are based on calorie intake of individuals. They first calculate the price of a basket of food items that provides 2100 calories (the amount of calorie that body needs to survive) as the threshold for food poverty line and then convert it to the poverty line.
    A legitimate question would be whether this is really an appropriate way of calculating poverty for Iran. Is there a reliable relationship between calorie intake and a person’s income? Does spending less on food really mean earning less money? How about assets (house, car, …) that individual owns? Is this method consistent over years considering the change in the food items of the basket?

    I hope discussions like this, at least help us demand accurate and honest answers from Iranian researchers and prevent us from blindly accepting biased (and popular) claims that don’t match the reality.

  8. iverson said, on April 1, 2009 at 6:30 pm

    I am so so pleased that you have started blogging.

    • dsalehi said, on April 2, 2009 at 5:30 pm

      Thanks! I hope I can keep it up.

  9. Ali said, on April 1, 2009 at 3:49 pm

    Great to have you here sir!

Leave a Reply to Jermaine Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: