Sunday 19 July 2015

Greek lessons for Chinese GDP growth

As well as being an object lesson in why national debt is something that should concern a country's leadership, and why parties attempting to sell their electorates on a dubious mixture of debt-renunciation and increased borrowing-funded spending should be avoided like the plague, Greece also has lessons for us on just how far national statistics can be skewed in the absence of independent auditing and analysis. As Michael Lewis discovered in a 2010 article on the causes of the Greek crisis:

When Papaconstantinou arrived here, last October, the Greek government had estimated its 2009 budget deficit at 3.7 percent. Two weeks later that number was revised upward to 12.5 percent and actually turned out to be nearly 14 percent. He was the man whose job it had been to figure out and explain to the world why. “The second day on the job I had to call a meeting to look at the budget,” he says. “I gathered everyone from the general accounting office, and we started this, like, discovery process.” Each day they discovered some incredible omission. A pension debt of a billion dollars every year somehow remained off the government’s books, where everyone pretended it did not exist, even though the government paid it; the hole in the pension plan for the self-employed was not the 300 million they had assumed but 1.1 billion euros; and so on. “At the end of each day I would say, ‘O.K., guys, is this all?’ And they would say ‘Yeah.’ The next morning there would be this little hand rising in the back of the room: ‘Actually, Minister, there’s this other 100-to-200-million-euro gap.’ ”

Indeed, guaranteeing the independence of the Greek statistical agency to prevent this kind of thing happening again was a key demand of Greece's creditors at the negotiations for Greece's bail-out, a request that has now been accepted by the Greek government and approved by the Greek parliament.

It was with this in mind that I read Christopher Balding's latest post over at the Nanfang on how China's GDP growth has some somewhat suspicious irregularities in it.

Anyone who has followed Chinese affair for a while will be aware that Chinese GDP statistics have some problems with them, such as, for example, the tendency of nearly all provinces and municipalities to publish growth figures higher than the average. It has been assumed, however, that national statistics are more reliable than those collected by local governments, and of course these are the only statistics available so there is little choice in which statistics to use. Balding's piece, however, give very persuasive (but not absolutely conclusive) evidence that the national statistics may also be highly problematic. This includes:
  • What appears to be the wild under-estimation of inflation. The example Balding gives is the private consumer price index (CPI), which has increased by only 8% (not annually: in total) since 2000. The idea that the cost of housing in China could have only increased by this much is utterly preposterous to anyone familiar with the vast increase in housing cost in that country that has left a generation of young families struggling to buy a place to live. It also goes against the findings of independent studies that house prices have increased by roughly 10% per year since 2004. Obviously, if you under-estimate inflation, then you over-estimate actual GDP growth, since GDP growth is adjusted for inflation by apply a deflator roughly equivalent tor inflation - the larger the deflator, the lower GDP growth will be.
  • Estimating the percentage of households renting at 12%,  meaning the 53% inflation in rent costs in the 2000-2014 period applies only to a relatively small portion of the population. To anyone with experience of China during this period, this does seem very low, however obviously you should not decide this figure is inaccurate based simply on anecdotal evidence.
  • Applying, since the year 2000, an 80:20 urban:rural weighting to statistics in a country that has only recently become majority-urban and is still only around 55-60% urban-dwelling. Given that it is urban areas which see the highest growth and the lowest price-inflation, this would clearly slant the figures towards higher growth and lower inflation.
  • 1.1% growth in electricity consumption for the Jan-May 2015 period compared to the previous year. The growth in electrical consumption has long been treated as an unofficial measure of GDP (although there are obvious problems with using it in this fashion) but the discrepancy between the increase in electricity consumption and the 7% GDP growth figure given for 2014 is quite a big one.
  • Corporate profits grew by only 0.6% (I assume this is for 2014 - Balding doesn't give a source in his post). Again, the question is: how this can possibly be in a country where the economy is supposedly growing by 7% a year? Really there are only three options: unobserved massive growth in the state sector (highly unlikely), massive fraud on the part of corporations to avoid tax (unlikely), or the GDP growth figure is wrong.
  • A 20% fall in imports in 2014 that has driven some of Chinese trade partners into recession, yet does not seem to be at all reflected in trends in the GDP statistics. Normally you would expect such a massive decline to be the result of some internal changes visible in the GDP growth figures, but not here.
Again, none of these are absolutely conclusive, but they at least should give anyone treating the Chinese GDP growth figures as even being broadly indicative of the state of the Chinese economy food for thought. China's statistical agency, the National Bureau of Statistics China (NBSC) is every bit as vulnerable to political pressure as Greece's one was, probably more so because in a closed society lacking free media like China manipulation of the statistics would be much easier to get away with.

Whilst I understand the hesitancy of publications like The Economist to concede that the NBSC is doing much more than "smoothing" the data, Balding's evidence (if accurate) hints at much deeper manipulation. Whilst no-one wants to play the role of doom-monger about the state of the Chinese economy (particularly since Gordon G. Chang has that role sewed up), this does suggest that China's GDP growth has been much slower in recent years than official statistics portray them to be, and may well be around the 5-6% mark (or less) that was predicted earlier in the decade - a rate that would be far from disastrous. 

Finally, there is the question of whether it will ever be possible to know the true rate of economic growth. Balding is undertaking research into this very subject,and I wish him luck with this, but it may well be that, as with Greece, the value of an independent statistics agency won't be recognised until a financial crisis has washed over the country.


Ji Xiang said...

I don't doubt China's official statistics are manipulated. At the same time, I don't doubt that 5-6% GDP growth is still absolutely fine. Even 0% would actually be alright. China already has enough money and resources to give its people something approaching a decent life. What is needed is to address the country's terrible social and environmental problems, but especially for the first one meaningful political change is necessary.

Gilman Grundy said...

As I said, 5-6% GDP growth would be far from disastrous. O%, on the other hand, would be a different matter altogether. Li Keqiang (who said that GDP figures were "man made" and that he relies on other indicators) may not use the GDP statistics in decision-making, but as you can see from the Economist article, there are many serious people out there who take the GDP statistics largely at face value. These people presumably do use the GDP statistics in their decision making and may therefore be led into making bad decisions.

Additionally, the GDP statistics do not appear to be made up out of whole cloth - the government does not merely write down a number and publish it. IF rigging is taking place, it is done by exaggerating positives and minimising negatives. Therefore, there would still be a motive for local governments to attempt to skew the statistics through things like off-the-books borrowing inflating local GDP, leading to exactly the kind of problems seen in Greece.

Of course, the above is all theoretical: there is no absolutely conclusive evidence of rigging. The GDP growth figure could be 7%, I just have a hard time believing that this could be the case at a time when imports are rapidly shrinking, and electricity consumption and corporate profits are essentially flat-lining.