Buyers beware of Chinese data

There’s some good news emerging out of China. The economy grew 7.9 per cent overall in the December quarter, beating many estimates and up from 7.4 per cent the previous quarter. For those seeking evidence of a structural shift toward a more sustainable model, official consumption numbers were encouraging, accounting for 4 per cent of GDP growth or over half of GDP growth over the quarter. Fixed investment, which has been driving growth for far too long, contributed 3.9 per cent of growth. Income growth in cities grew a healthy 9.6 per cent and retail sale numbers were up slightly from 11.6 per cent to 12 per cent. There was also improvement in key retail sectors such as car sales, which grew 6.9 per cent.

Is this evidence that China is over the worst of conditions, or are these numbers too good to be true? The prudent response is not only that we do not know, but we cannot know. For those China bears who insist that something seems amiss, there is plenty of evidence for that as well.

First, a comment about the quarterly data and growth figures:

Every quarter, the Chinese National Bureau of Statistics goes through the same ritual. Statistics come in from all over the country. The provinces compile them with impossible speed – around two weeks, or three times as fast as many developed economies with much more efficient processes of data collection.

For example, comparable data in countries such as America take around eight weeks to collect. Remember that China has four times as many people and far inferior bureaucratic and other institutional infrastructure required for data collection and analysis.

The NBS sorts through the data, ‘consults’ with senior Chinese Communist Party officials, and then applies a mysterious methodology to trim them into shape. The various NBS sub-agencies then release figures that are uncannily well aligned with the targets set by political masters in Beijing.

After several years, provincial historical data is tediously retrieved and analysed by Chinese and other economists and statisticians, and official figures are revised. By then, business and media outlets have moved on.

But statisticians, economists and academics will plough on nevertheless. Significant discrepancies are discovered and put to the NBS, and these findings will sometimes be passed on from NBS officials to senior political officers. Occasionally, past practices are condemned, ending with Beijing promising to meticulously address the ‘structural’ and ‘institutional’ flaws in the statistical gathering process.

This brings us back to the official numbers, which are derived from reporting by local officials. China has 1464 counties and over 2800 county-level divisions that essentially do the data gathering and reporting. The data collected by these county-level officials come from more than 30,000 townships.

These county-level officials have massive career and personal incentives to tell Beijing what it wants to hear as regards hitting central targets – whether this be higher growth, an ‘engineered’ slowdown, or the drivers of growth such as fixed investment or consumption. It is the basis for their praise and promotion. While the upside for dishonesty is obvious, there is usually little downside, as it’s unlikely they will be caught, let alone punished, for fudging figures.

Bear in mind that Beijing also has strong political incentives for wanting to spread some optimism around at the moment. The new leadership under incoming President Xi Jinping and incoming Premier Li Keqiang will not be formally confirmed until March. The interregnum period until then is not the time to release bad news.

The point is that we cannot know whether China really arrested seven consecutive quarters of declining growth, just as we cannot know whether the figures for the last seven quarters really were. No one has done any substantive and independent investigation of it.

As one of America’s leading observers of the Chinese economy, Derek Scissors, pointed out to me the relationship between M1 (all physical money in the system as well as most instruments that can be easily converted to cash such as demand deposits) and GDP growth has been complementary: growth in M1 has consistently matched official GDP growth rates. But in late 2011, M1 growth fell below official GDP growth rates for the first time; and M1 is currently at its lowest levels of growth over a fourteen-year period of the study. Yet, GDP growth has mysteriously increased, meaning that the relationship between M1 and HDO growth has been transformed.

Business Spectator readers can make up their own minds. The point is that it is impossible to know what to make of the official 7.9 per cent quarterly growth figure over the most recent quarter.

Even so, we can be reasonably sure that China’s economy actually grew. There is plenty of evidence for that. But the growing interest is in the drivers of Chinese economic growth, particularly trends for fixed investment (which is way too high) and domestic consumption (which is way too low.) This is all about the ‘rebalancing’ of the economy that China has to achieve.

But something is clearly amiss. Even up to November 2012, Chinese officials have been telling us that fixed investment has been growing at well over 20 per cent, far faster than key indices for domestic consumption such as retail sales growth. Yet, we are now told that domestic consumption has contributed more to GDP growth than capital formation – it does not make sense.

Likewise, we are told that car sales improved on the September 2012 quarter and grew 6.9 per cent in the December quarter. The problem is that a significant part of the sub-data used by the NBS considers an item to be purchased (consumed) when it enters the showroom, not when it is driven away by a customer. China's dealerships had 2.2 million unsold cars in their showrooms in June 2012, rising from 900,000 in December 2011. The figure of unsold cars in showrooms could now be as high as 2.5 million. Yet, manufacturing capacity is rising, suggesting that car manufacturers are jumping into the Chinese market with their eyes closed. (According to KPMG’s Global Automotive Executive Survey 2012, China had an estimated 6 million units of unused manufacturing capacity, more than twice the size of the entire car market in Germany. This figure is expected to rise to 9 million units of unused manufacturing capacity by 2016.)

Some of these unsold cars are counted as having been consumed. The same problem applies to many other retail items. Although we cannot know the true figure, it is highly unlikely that domestic consumption exceeded fixed-investment as the major driver of GDP growth in the December quarter.

As the World Bank's most recent East Asia and Pacific Economic Update puts it in offering its 2013 forecast of growth at 8.4 per cent for China, the expansion will be "fuelled by fiscal stimulus and the faster implementation of large investment projects." And all indications seem to be that the same occurred during the December quarter.

So to end, make your own mind up about China’s quarterly figures. As executives at Caterpillar will tell you after its recent US$580 million write-down after discovering "deliberate, multi-year, coordinated accounting misconduct” at Siwei (the Chinese firm it bought last year for $654 million,) it pays to do your due diligence when it comes to China.

Elements of China’s past national statistics have also been the result of "deliberate, multi-year accounting misconduct” even if it may not be "coordinated.” The problem is that we will probably never know until it is too late.

Dr John Lee is the Michael Hintze Fellow and adjunct associate professor at the Centre for International Security Studies, Sydney University. He is also a non-resident senior scholar at the Hudson Institute in Washington DC and a director of the Kokoda Foundation in Canberra.