Tyranny of numbers

Are living standards as poor as they say?

Western journalists who  travel to Iran get an earful of complaints about how bad things economic are.  The World Bank disagrees.  Iranian GDP per capita in 2007 stood at $10432 (in 2005 Purchasing Power Parity dollars), which is only one-forth that of the of the United States, and less than a thousand dollars below Turkey’s.  In terms of average growth rate of GDP per capita, Iran has actually done better than in the ten year period 1997-2007,  3.5% compared to 2.5% for Turkey. These numbers are readily available from the World Bank’s World Development Indicators website.

There is no shortages of complaints one can have about Iran’s economy (high youth unemployment, high inflation, and stagnant productivity, to name a few) , but a low standard of living is not one of them.

Should you trust official economic data from Iran?

Every time a reporter mentions Iran’s inflation or unemployment data, they feel obliged to assert, sometimes quoting unnamed experts, that actual numbers are probably twice the official rates.  But most people who work with numbers know well that officials can manipulate one year’s inflation rate or two, but if the rates are misreported for a number of years, the mere effect of compounding will soon reveal their hand. If you take an inflation rate twice the reported rate for the last 15 years, the price level in 2008 would be about 130 times higher than it was in 1993, nearly ten times higher than the official rates indicate. It is easy to show that under reporting inflation by 50% each year for 15 years produces some very absurd results.  For example, think of what such under reporting would do to the purchasing power of unskilled construction workers.  Their wages increased by a factor of 20 during the fifteen year period 1993-2008 (you can easily verify this number by asking people in the construction business in Iran, if you do not trust the Central Bank index).   The official data that show a 14-fold increase in prices during the 15-year period, suggesting that unskilled construction workers experienced a 42% increase in real wage over 15 years, or 2.4% per year, which is rather low because it puts the increase in their real wage just below that of per capita GDP.  If one believes that actual   inflation was higher than has been reported, one is saying that construction workers did even worse.  How much worse? Let us assume that the so-called experts are correct and inflation is in fact twice the official rate.  In that case the real construction wage in 2008 should be only 15% of its value in 1993 (then about 6000 rials per day).  Clearly, even a superficial knowledge of the change in the living standards of unskilled workers in Iran would confirm that the 42% increase is much closer to the truth than an 85% decline.

Official unemployment rates are equally deemed suspect.  Iran’s unemployment rate of 12% (2006 census) is about the average for the Middle East, so there is no surprise there.  What is interesting is that if officials are making up the numbers, they are doing a poor job of fixing them.  They have put 2.3 million of the 3 million unemployed in the 15-29 age group and young women’s unemployment rate twice that of men.  Men and women over thirty have unemployment rates of around 5%.  Not the best picture officials should be making up.  But if they are fixing numbers, they are doing an amazing job.  There are at least a dozen surveys that are publicly available in unit record which show the same dismal picture of young vs. adult unemployment (the 2% sample of the 2006 census is available on-line from the website of the Statistical Center of Iran).

Is poverty always rising?

This claim has been made even for recent years when oil revenues have dramatically increased.  I have shown elsewhere that, while the poverty rate did increase in the 1980s, in the last two decades it has been declining, and that it is now quite low by developing country standards.  The World Bank estimates the poverty rate based on the international $2 per day standard in 1998 at about 7% and I estimate it at less than 10% using a higher poverty threshold.  Comparing the cumulative distributions of real household expenditures in the last fifteen years in the Journal of Economic Inequality, I show that later year distributions, say 2007, stochastically dominate older ones, say those in Khatami years, leaving no room for doubt that poverty rates have declined, as they should when money is pouring into the country.

Anyone who believes that the poor in Iran have been simply watching while a couple of hundred billion dollars of oil revenues were injected into the economy must have a very low opinion of them.  Why would they not be able to find any way to lay their hands on some of this money by working in the service sector, to which the rich turn when they are loaded.  It is one thing to believe that the poor lose in relative terms with the inflow of oil money (inequality almost always increases), and quite another to believe that they lose in absolute terms.

I will have more to say about all these issues in future posts, but I thought that exposing  these myths is a good way to introduce the theme of this blog– the tyranny of numbers.