Friday, February 23, 2018

Problems with Other Countries’ GDP Data

I’m going to start with a digression.


Most students don’t know the difference between a country, a nation, and a state.

A country is the physical space. It’s what you see borders of in an atlas.

A nation is a set of people with a cohesive culture tying them together.

A state is a government that has “standing”: basically, other states admit that it’s worth engaging with. An effective state is sometimes described as “sovereign” (even though that word applies to royalty, and we don’t have much of that anymore), in that it acts as if someone is in charge of the whole thing.

For most of us, a country, a nation, and a state roughly coincide.

Through this lens, we can see the overarching position that Trumpers share: a sense that the immigration is making the U.S. less of a nation, and a sense that our state no longer controls the entire country.

Most of our problems around the world crop up in places where there’s a worse mismatch between these three ideas. 

Iraq has had problems spanning decades because there’s a country on the map, and a state run out of Baghdad that tries to control the whole country, but there isn’t much of a nation of Iraqis: citizens are Arab or Kurd, or Sunni or Shia.

Somalia is a country on the map, and a nation of people who identify as Somalis. But there hasn’t been a viable state running the place from Mogadishu for about 25 years. This country is the source of the term “failed state” that is tossed around in the media.

And what we roughly call Kurdistan is a nation of people who think of themselves as Kurds, with a state that effectively runs the region and defends itself … but there’s no country on the map.

Now I’m going to take you back to the Thirty Years War. Remember that? Probably not. Anyway, it was a nasty continental war fought in Europe in the 17th century. One thing that came out of it was the idea of Westphalian Sovereignty (because the treaty that ended the war was signed in Westphalia).

Briefly, Westphalian Sovereignty means that other states can either accept you as a state or not. If they do, it means that they respect your state’s decisions regarding everything that goes on inside the borders of your country. For example, this is why the Palestinian Authority makes the case so strongly that they need a Palestinian state.


The problem with Westphalian sovereignty for macroeconomics is that it suggests that if a country produces data about itself for public consumption, other countries should accept that data at face value. Except some governments lie.

The classic example of this is CIA estimates of the size of the economy of the Soviet Union, which were largely based on data provided by the Soviets. Officially, in 1990, the CIA said the Soviet economy was about half the size of the U.S. Critics that were friendly to the CIA marked this down to 40%, those unfriendly to it down to 35%, and an influential Soviet economist estimated 28%. Admittedly, the remnants of the Soviet Union have not had an easy time over the last generation, but the current (overstated) PPP estimates for the region are about 25% of the U.S., and the current (understated) exchange rate based estimates are at about 10%.

Problems with GDP measurement around the world are kept track of by a private consulting firm called World Economics. Their Data Quality Index allows ranking of countries by probable inaccuracy (but you have to pay real money to get their actual GDP estimates). They also have some tough reading there (not required) detailing about a dozen ways in which states fail to accurately track their country’s GDP. Here’s the skinny on the problems:

  • Base years need to be updated periodically, and many countries don’t do this very often.
  • The UN has pushed out 4 increasingly refined systems for counting economies over the last 80 years. But countries are not required to switch to the most recent/best system.
  • Informal economies are tough to count, and are bigger in countries with weaker states.
  • Poor countries have trouble affording better GDP measurement.
  • Government’s interfere in the calculations.
  • Much of what government’s purchase has a much bigger divergence between price and value than consumer goods do, so countries with big government sectors are probably measured more poorly.
  • Financial sectors are tough to measure, which is a problem for richer countries with generally thicker financial sectors.

Economists are particularly concerned about China’s GDP numbers. China may be the second biggest economy in the world, but it ranks around the middle in accuracy of GDP. In fact, World Economics devotes an entire section to China in its discussion of Asian economies — they have all the problems list above, plus these:

  • China produces its GDP estimates much more quickly than other countries.
  • China never revises its announced GDP figures.
  • The variability of reported GDP is far lower for China than any other country.
  • China has an uncanny ability to announce targets for GDP growth rates, often years in advance, and then hit them time after time.
  • State-owned enterprises (SOE’s) report their output data directly to the state, which assigns values to it, rather than allowing it to be traded on the open market
  • Chinese provinces appear to report a lot of GDP to Beijing that is systematically pared back by Beijing due to overstatements.
  • Double counting of transactions that cross provincial borders seems to be a big problem.
  • Much of the world is aware of the high prices and rental rates for real estate in coastal China. This is absent from Chinese inflation statistics, suggesting that real GDP growth rates are overestimated.

Do note that you can test forensically for data falsification. Researchers looking at this do not find much evidence for this in China. Instead, the evidence for pervasive “guesstimation” seems more convincing.

A Wikileak showed that China’s current Premier, Li Keqiang, described his country’s GDP as “man made”. The Economist went far enough to construct an index of Keqiang’s personal favorite economic indicators — rail volume, electricity consumption, and loan disbursement — as an alternative to official GDP statistics. A good source for analysis and data on China’s economy is Balding’s World (Balding is an economics professor working in Beijing).

No comments:

Post a Comment