China has made great economic progress in just over two decades, yet growth so far has been driven by heavy investment. For it to continue at the current pace, the general consensus among economists is that consumption needs to increase - but it is difficult to say just how many Chinese are in a position to drive this consumption.
A few years ago, Goldman Sachs put the figure at 100 million people, and saw it climbing to 650 million by 2015. McKinsey Global Institute published similar projections. Others claim the number in 2015 will be less than 300 million.
More confusingly, everyone has a different definition of the "middle class" that is supposed to be behind this growth in consumer spending.
"There's no real definition of middle class," said Jonathan Anderson, chief Asia economist at investment bank UBS. "If you want to identify the number of people who can afford a mobile handset, the answer would run to nearly the entire population. If your definition is how many people have 'significant discretionary power', [it's] 100-125 million."
Others go even further.
"China doesn't have a middle class," said Arthur Kroeber, director of Dragonomics Research and editor of China Economic Quarterly. He believes the term conjures up too many cultural connotations that simply don't apply - like taking the family of five out in the SUV for dinner and a movie. He prefers to divide Chinese consumers into two classes: Surviving China and Consuming China.
Kroeber recently released a series of studies together with Access Asia, a market research consultancy, which say the primary problem lies with the many sets of retail consumption figures the government releases.
The first is deceptively called "total retail sales of consumer goods" - misleading because it counts all goods purchased, including wholesale, rather than goods sold to people in shops; furthermore, it leaves out the service sector entirely. Yet this figure most closely resembles analysts' estimations of the size of China's retail market. In 2005 it was US$832 billion or US$637 per capita.
Then there is the "private consumption" component of GDP, which was US$671 per capita in 2005, according to government statistics.
"This suggests that retail sales equal 95% of private consumption," the report says, "an obvious impossibility."
Taking all this into account, how much is China really spending?
DECEPTIVE FIGURESMcKinsey estimates China's total urban disposable income to be roughly US$622 billion but this just leads into a debate on the urban population is calculated. The government puts it at 562 million, something easily disputed by a careful analysis of census data. The official figures include in the "urban" population millions of residents of small towns far from big cities - most of whom are not very well off.
According to Anderson, the "core" urban population of China is 224 million. This helps explain why Kroeber's report estimates that China's true retail market is only half of what is normally reported: roughly US$450 billion or US$346 per capita. That is about half the size of Malaysia or Thailand and one 30th of Japan.
From this perspective, there hardly seems to be a market at all. But anyone who has been to a Carrefour hypermarket or Gome appliance store can tell you there are plenty of people lining up to buy.
The crucial difference between these two arguments is location and this leads back to Kroeber's thesis of two Chinas.
Surviving China consists of pretty much everyone. "As far as significant retailers are concerned, out of China's 1.3 billion people, 1.2 billion simply don't count," said Kroeber.
The 110 million people identified in the report as viable Chinese consumers - the Consuming China category - are in the top three urban areas: Beijing/Tianjin, Shanghai, and Guangzhou/Shenzhen, and these areas are as close together as Madrid, Belgrade, and Moscow. Given China's geography and current infrastructure, the logistical difficulties of serving these three areas are huge.
If potential sales are great enough, however, they outweigh logistical costs. The rosy figures from McKinsey and Goldman Sachs, which argue that by 2015 some 80% or more of China's urban population will be in the middle class, provide sound argument to outlay for the costs of transportation. The trouble comes when one digs into their definitions. Goldman puts the minimum income for "middle class" status at US$9,000 annually; McKinsey includes gradations from US$3,200 ("lower middle class") to US$12,500 ("upper middle class").
INCOME DISPARITIESIn almost any developed country these levels of income would classify people as impoverished. McKinsey defends its numbers by comparing purchasing power parity, by which it argues Chinese middle class incomes would be the equivalent of between US$13,500 and US$53,900.
This would be an important point if it weren't for the fact that many of the companies relying on these figures are selling cars and expensive clothing. The price of a BMW is not adjusted for purchasing power parity.
In fact, the automobile market makes a good example. Some analysts point to China's growing car craze as evidence of a rise in spending power but they fail to consider that many Chinese may purcase virtually nothing else for years before buying a car.
All of this is not to say that China is going to be poor forever. Kroeber and company are not foretelling doom; rather they are calling for cool heads. They predict that Consuming China will grow to some 270 million a decade from now, with average household consumption rising from US$5,000 to US$10,000 in that time.
Anderson agrees, pointing to the strong investment of the auto industry. Many predicted a glut in the market, but demand is catching up.
Similarly, the reality of a wave of Chinese consumers taking over the world may eventually catch up to the newspaper headlines, but it will be some time still.