Monday, May 21, 2018

Washington Post catches up to Borepatch in 2012

Study: Authoritarian countries overstate their economic growth rates:
"I find that a 10 percent increase in nighttime lights is associated with a 2.4 percent increase in GDP in the most democratic countries and with a 2.9 percent to 3.4 percent increase in GDP in the most authoritarian ones," Martinez said. The most obvious explanation is that those countries are the most likely to fudge their GDP figures to make their political leaders look good.
It's a clever study, using nighttime lighting as observed from space as a proxy for economic activity.  Of course, we've know this for ages:
The most spectacular Intelligence failure of the Cold War was the failure to recognize the impending collapse of the Warsaw Pact in 1989, and the Soviet Union in 1991.  The CIA certainly reported on plenty of problems as the 1980s wound on towards International Socialism's rendezvous with destiny - as highlighted in this self-serving and mostly missing-the-point analysisfrom the CIA on its own reporting.  This analysis is entirely unsatisfactory, because it completely fails to account for how the USSR was ranked third or forth on the list of the largest World economies.

That wasn't just wrong, it was spectacularly wrong, and had enormous policy implications all throughout the 1980s.  All Western policy makers who used the CIA's analysis as the basis for policy found that they were equally (and equally spectacularly) wrong.

So how did the CIA get into this position?  By being very good at collecting data, but not so good at collecting the right data.  As near as we can tell from unclassified reports, the Agency had a decent grasp of the summary economic data as presented to the Central Committee.  The problem was that the data was bogus.

The Soviet economy was regimented, a rigid command and control system where those at the top set the direction and goals, and those in the factories tried to execute to the 5 year plan.  Factory managers who didn't execute to plan had troubles, and so there was a built in incentive to cheat.  Corners were cut, output was trimmed to where the plan was met with sub-standard, shoddy goods.  Or the numbers were simply made up, and the fraudulent data was sent up the chain.

Further massaging was done in mid level bureaus, because nobody wanted to be the guy who brought the bad news.  Essentially, the data deteriorated its way up the chain to the top, and nobody was very interested in double checking to see if the data as reported matched what had originally been collected.  You see, it wasn't in anyone's interest to do so, and it was very much in everyone's interest not to do so.
Sadly, the WaPo hasn't caught up to Borepatch from 2011, which has the really disturbing take on all of this monkey business:
China is building ten (!) brand new cities each year.  However, it's built with borrowed (leveraged) money, and so rents are high - so high that the cities are deserted.  The Great Mall Of China has 1 - that's one - store open.  There are 64 Million empty Chinese apartments.

It's not just city construction: China's much touted (by Thomas Friedman, and even Barack Obama) High Speed Rail network is supposedly an example of a "moon shot" project.  We have to "keep up with the Chinese".  Except the trains are ghost trains:
Here’s the latest from the South China Morning Post on the dismissal of the nation’s Railways Minister and the engineer in charge of the system’s R&D. Seems there were “severe violations of discipline,” which is usually code for corruption. The larger issue with the vast (16,000 kilometers planned by 2020) endeavor is that it isn’t, in fact, so appropriate to China’s needs. Rather, it may be another symptom of a bubble economy in which vast sums are misspent on underutilized assets. 
The high speed trains are wildly expensive, because the (Chinese government owned) rail network had to issue $300 Billion in bonds to build the network.  And so fares are very high, and people take the slower (but much less expensive) old trains.

But that $300 Billion gets added to China's GDP score.
Remember the $Trillion "Stimulus" program that didn't stimulate anything?  Does anyone know what we got for that money?  Can anyone name a single thing that was built with that money?  But a Trillion dollars got added to the GDP figures.

Economic data are mostly untrustworthy.  A little less so in the West, a little more so in other places.  We can speculate on an algorithm to describe this, based on the finding s of the study reported in the WaPo: the more central control over an economy that a state has, the less reliable the economic data will be.

Now think about how much control the Fed.Gov had over the economy in 1960, and how much control it has today.  What does out algorithm suggest for the veracity of the reported economic data?

To ask the question is to answer it.  Or you can check here - there are many ways to game the data, if the state is motivated.


SiGraybeard said...

I wonder if they'd publish their correlation coefficients on night time lighting vs. GDP? I can see that there would be correlation, but there's a movement toward reducing nighttime sky glow from poorly directed streetlights in the more developed countries; primarily us and the EU. There are dark sky preserves and parks (and various other words) springing up all over the 1st world.

It's like a continuation of environmental movements that lead to the most advanced nations being the cleanest while the most primitive economies are the dirtiest. Compare a US oil port to Nigeria, for example, where spills are so common there are websites devoted to tracking them: .

Will said...

the focus seems to be changing the wavelength of night time lights, primarily street lights, not so much turning them off. San Jose and nearby cities did this due to the local Lick Observatory.

No idea if that helped. Must be frustrating for them, as there wasn't much night light when it was built in the 1880's. Now, if you get up any of the hills around the Bay Area, you can see it is virtually a sea of light.