Getting behind the numbers
Jonathan Andersen (Executive Director for Asia-Pacific Investment Research at Goldman Sachs)
This article originally appeared in the February 17, 2003 issue of South China Morning Post in Hong Kong and is reproduced here with permission from the publisher.
Is something wrong with China's official growth statistics? Since the mid-1990s, official growth figures have been all too predictable: real growth of the gross domestic product of between 7 and 8 per cent in each of the last seven years, and equally stable increases in retail sales, industrial production, money and credit.
At a time when China's neighbours were successively disrupted by the Asian financial crisis, the global information technology boom and the subsequent bust - and when China itself faced the aftermath of a domestic bubble - outside observers have found it difficult to believe that the Chinese economy could yield such consistently buoyant growth.
The most pessimistic view holds that the Chinese economy has barely expanded since the Asian crisis. In the space of a few years beginning in 1996, China suffered a "triple hit" on the economy: a sharp disinflation programme at home; the slowdown in external demand during the Asian crisis; and a wave of closures and layoffs at state-owned enterprises.
The contradiction between these trends and the official growth figures has been the focus of a number of academic studies. One of the best-known, carried out by Professor Thomas Rawski of the University of Pittsburgh, concludes that actual growth rates since 1997 have been significantly overstated. In particular, he argues that overall gross domestic product (GDP) growth in 1998 and 1999 amounted to no more than 2 per cent, with only moderate improvement in 2000.
We share the concerns about the quality of Chinese statistics. Despite enormous improvements in the quality of statistical data, China's National Bureau of Statistics is still an agency caught between the planned and market economies. While staffed with professionals who understand international practice, the bureau remains dependent on direct reporting from larger state-owned or partially state-owned enterprises in industrial sectors, as well as data passed up from provincial offices - who have a strong incentive not to report disappointing or volatile figures. As a result, national accounts statistics have not fully represented some of the most dynamic sectors in China and have been subject to non-economic pressures.
However, we believe the real story is one of excessive smoothing, rather than merely hiding low growth. The evidence suggests that China went through a sharp economic slowdown beginning in 1996, and then a rapid recovery from 2000 onwards - neither of which were adequately captured by official growth statistics. Indeed, our best estimates of current momentum point to growth rates well above the official figure of 8 per cent.
Without going into excruciating detail, there are two ways to compile GDP statistics. The first relies on gross production data, subtracting the cost of inputs to arrive at the value-added figure for each sector of the economy; this is the method used by the Chinese authorities to arrive at the official real growth figure.
The second, which is used by most developed countries, estimates final expenditure demand, using data on consumption, fixed investment, inventory accumulation and net exports. This "expenditure approach" has the advantage of relying on selective sampling rather than old-fashioned direct reporting, putting more power into the hands of professional statisticians.
China's statistics bureau does compile expenditure-side national accounts data, but the figures are not produced frequently enough to allow them to compete with official production-based statistics.
The numbers on expenditure GDP imply a very different growth picture over the last seven years. Expenditure data show a much stronger slowdown in real growth in the late 1990s, from a rate of 12 per cent in 1995 to under 5 per cent in 1999, followed by a rapid recovery in the past two years to over 10 per cent last year.
This view is plausible because, for one, the composition of the slowdown and subsequent pickup is very much in line with the "normal" business cycle seen in other economies: the downturn in the late 1990s was led by a notable deceleration in investment spending, partially offset by fiscal stimulus policies. The recovery has also been led by investment, supported in part by stable growth in consumption demand.
In order to provide an independent measure of growth performance, we have created our own proxy index for the Chinese economy - the Goldman Sachs China Activity Index, or GSCA.
The implied growth rates from the GSCA index are much closer to those derived from the statistics bureau data on expenditure GDP. According to the index, the momentum of growth slowed from 10 per cent in the first half of the 1990s to only 3.5 per cent in 1998, and has since increased significantly, to nearly 10 per cent last year.
In other words, we find that China did indeed go through a relative recession - but is now growing very rapidly. We do not foresee another boom-bust cycle; most indicators point to relatively stable growth in the near term. First and foremost, we believe most of the factors commonly pointed to as macroeconomic risks - the weak banking system, ongoing restructuring and bankruptcy of state-owned enterprises, fiscal pressures and low income growth in rural areas - are overstated in the near term.
Of course, each of these factors has a potentially significant impact on China's economy, but we see them as ongoing medium-term priorities with which China can and should deal before they become a greater drag on growth.
Unlike the onset of the last downturn, China's economy is not forming a "bubble". The first half of the 1990s was characterised by a loss of monetary control, rampant inflation and other factors which virtually guaranteed that policies to stabilise the economy would result in a sharp downturn. Barring unforeseen shocks, we expect the Chinese economy to grow at an average rate of 7.5 per cent to 8 per cent over the next three to five years.