How cool is this?

Is Kazakhstan’s economy worse off than the government reports?

Kazakhstan -- which less than a decade ago was one of the world’s hottest economies -- is cooling faster than a bowl of borsch in a viga.

Or: Kazakhstan is rolling along with its usual cheery growth.

Take your pick.  The national agency on statistics tells both tales.

In a recent news briefing, KazStat announced that the economy was growing 6% per year, according to the business weekly Panorama.  Meanwhile, the agency’s online database tells a darker story.  Compared to 2012, for the period from January through October, Kazakhstani income grew only 2.4% -- the slowest rate since the financial crash of 2009, when the economy shrank by more than 3%.  (This is in terms of the amount of goods and services that the income can buy – a measure called “real income”.)  Per person, income this year grew by less than 1%.  Even the Europeans would grumble about a pace that torpid.

These are not revelations.  KazStat’s data have been showing a similar slowdown for all of 2013, in both the Russian and English versions of the database.  And its reports of nominal income – which is not adjusted for price changes – are consistent with those for real income. 

What’s going on?

Conceivably, both stories are right.  Statisticians measure the economy in terms of the amount of goods and services produced on Kazakhstani soil, called “real gross domestic product”.  This is not exactly the same as income earned by Kazakhstanis.  Some work in other countries, so their income does not show up in Kazakhstan’s GDP.  On the other hand, the contribution of immigrants to our GDP does not show up in Kazakhstani income.  In the oil and gas industry, foreigners build most of the extraction facilities, and much of the resulting income goes overseas.

Cooks and books

Still, it’s odd to see a drop of two-thirds in the growth rate of income and, at the same time, a substantial rise in the growth rate of GDP.  According to KazStat’s database, GDP fell one-half of a percent in 2012 and rose only 1.6% in the first half of 2013, compared to the corresponding period of the previous year.  One might suspect that KazStat is now gilding the lily and that, in truth, an economic slowdown continues. 

Circumstantial evidence supports this hunch:  KazStat’s estimates don’t add up.  The head of the agency, Alikhan Smailov, said GDP grew 6% over the period of January through November, compared to the same period for 2012, reported Panorama.  But the headlines on KazStat’s home Web page say the “short-term economic indicator” was an annual growth rate of 4.8% over the period of January through October this year.  To reconcile these estimates, GDP in November, normally too cold for outdoor work, would have had to have grown two and a half times faster than in the average (and warmer) month of 2013.  This is barely believable. 

So are KazStat’s headlines.  They report growth rates exceeding 10% for agriculture, trade and communications as well as of 7.5% for transport.  But growth rates were only 2.9% for construction, an economic bellwether -- and 2.3% in industry, which includes mining, the traditional engine of economic growth in Kazakhstan.  These are strange lapses for a supposedly robust economy, especially since these two sectors comprise from 35% to 40% of the economy. 

It’s hard to tell what’s happening in construction, because KazStat’s estimates are all over the map.  Compared to the corresponding period of a year earlier, output in this industry reportedly grew 3.1% in 2012, shrank 4.9% in the first quarter of 2013, and grew only .7% in the first half of 2013.  Either construction is a yo-yo or KazStat’s figures are unreliable.

Finally, KazStat’s latest estimate is out of synch with earlier ones.  The agency had calculated that real GDP per capita in 2012 grew only 3.5%, the slowest pace since 2009.     

Okun’s Law

Of course, statistical agencies correct their estimates as time goes by.  That’s normal.  But for an agency to tell simultaneously two stories as contrasting as these – hearty growth in GDP and stomach-churning decline in income -- is disconcerting.   

Finally, KazStat’s estimates flout economic principles.  Although the growth rate of GDP supposedly has risen about 6.5% since 2012, the reported unemployment rates for the two years -- for the nation and for individual oblasts – are virtually unchanged.  For the first three quarters of 2013, the national unemployment rate was 5.2%.  That was just a tenth of one percentage point lower than for 2012.  For no oblast did the unemployment rates for the two periods differ by more than a fifth of a percentage point.

This too is weird.  Usually, the unemployment rate moves in the opposite direction of GDP, and proportionally so.  After all, a growing economy creates jobs, cutting unemployment.  By a hoary rule of thumb, a fall in the rate of economic growth of two percentage points may raise the rate of unemployment by one percentage point.  For Kazakhstan, the year’s reported rise in economic growth of more than six percentage points would lower the unemployment rate by three percentage points.  Of course, this is just a rule of thumb, but the principle matters more than the point estimates:  A large change in the rate of economic growth should induce a large change in the rate of unemployment.  The latter should not stand still.

 

Anomalies also appear in the data across oblasts at a given time.  In 2013, the unemployment rate varied only from 5.6% (in the city of Almaty) to 4.9% (in Akmolinskaya and Aktubinskaya oblasts).  One measure of dispersion in these unemployment data -- the ratio of the standard deviation to the mean -- is only 4.4%.  But the corresponding figure for GDP growth rates is 68% (for the first half of 2013).  Are the unemployment figures even more bogus than the GDP ones?

Let’s lurk in the economic murk

Here's the point:  Foreign investors don’t want perennially rosy numbers.  They want the truth.  They already know that Kazakhstan has educated workers, some political stability and, definitely not least, oil and gas.  KazStat does not have to promise them the Brooklyn Bridge to get them to come.  To the contrary, the agency’s refusal to provide sound statistics discourages investors from committing their dollars for the long run.

KazStat could do a lot to help.  As other major providers of economic data, such as the World Bank and the International Monetary Fund, already do, it could document its figures with care.  Every spreadsheet available to the public should define in detail each data series and its unit of measurement as well as name the original sources. 

For foreign users, KazStat should clean up its English.  Users should not have to tussle with definitions such as this:  “Households’ monetary incomes represent the sum of the money resources received by household’s members in the form of a wages, the income of enterprise activity, social payments (pensions, scholarships, grants and other payments), percent, dividends and other incomes of the property, other monetary receipts.”

KazStat should provide balanced news briefings, noting ambiguities in the data.  And, of course, it should reconcile estimates.

Maybe KazStat will make a New Year’s resolution.  But don’t bet the farm on it.  – Leon Taylor, tayloralmaty@gmail.com

 

Notes

1. Viga is Russian for a blizzard.
2. According to KazStat, the annual growth rate in income fell from 6.9% in 2012 to 2.4% in 2013, for the period of January through October, compared to the same period in the previous year.
3. Here are calculations for the GDP growth rate in November.  In annual terms, the average growth rate was 4.8% for January through October but 6% for January through November.  Denote the average monthly contribution to annual GDP growth as Xi, where i indexes the month (1 for January, 2 for February, etc.).  Then X1 + X2 + … + X10 = 4.8%.  And X1 + X2 + …  + X10 + X11 = 6%.  Solving these two equations gives us X11 = 1.2%.  The average of X1 through X10 is 4.8% / 10 = .48%.  So the November contribution to annual GDP growth, compared to the average of the earlier 10 months, is 1.2% / .48% = 2.5. 
4. A well-known regularity in economics, Okun’s Law, holds that changes in the rate of unemployment is a linear function of changes in the rate of economic activity.  In particular, Y = a – bX, where Y is the change in the unemployment rate and X is the change in GDP (both changes measured in percentage points); and a and b are positive constants.  Clearly, dY/dX = -b:  An increase of one percentage point in GDP reduces the unemployment rate by b of a percentage point for any levels of the two variables.  Changes in X and Y are proportional to one another.   The American macroeconomist N. Gregory Mankiw gives a typical estimate:  Y = 1.5 - .5X.  In this model, a decrease in the rate of economic growth of two percentage points would raise the rate of unemployment by one percentage point.

Good reading

Mankiw, N. Gregory.  Macroeconomics.  Worth Publishers.  Seventh edition.  2010.

References

Agency on Statistics of Kazakhstan.  Various data series.  www.stat.kz

Oksana Kononenko.  BBP Kazakhstana viros na 6%.  Panorama.  December 20, 2013.