Last night, Kate Mackenzie of FTAlphaville single-handedly did what the entire government, financial markets and media of Australia have failed to do. She used reason to consider a wider framework for future Chinese growth paths than the mantra of steel intensity and urbanisation forever that has come to define Australia’s elite.
Under a new series called “China myths” she chopped Australia’s complacency to pieces. In the end, she may or may not be right but one thing is certain, this is the precisely the kind of debate lacking here. Well done to the FT.
The first effort, cross-posted:
The ever-increasing hunger for steel
It used to be an accepted fact that China’s appetite for steel and steel’s main ingredients — primarily coking coal and iron ore — would continue to rise sharply, not just in absolute terms but at an accelerated pace.
Annual steel consumption had been expected to rise from 2011′s 680m-plus metric tonnes to 1bn metric tonnes by 2020. This forecast has been a mainstay of many China-related predictions for some time, particularly in the mining sector. It was still being cited by BHP Billiton in March, even while the miner’s head of iron ore surprised many with a bearish tone. Since then, however, the world’s two biggest miners began to back away from it and this month Rio Tinto is talking about 1bn tonnes of steel by 2030 — a hazily far-off date.
But the 1bn tonnes-by-2020 was so confidently proclaimed for so long that it warrants further examination, even as faith in this target dies away. How did this forecast evolve, and why is it flawed?
The argument is that on a per capita-basis, China is way behind wealthier countries on its steel consumption. Just look here:
The above graph is from HSBC, which recently added to the inevitable-growth oeuvre with a lengthy note this month arguing that high commodity prices are the new normal (supercycle be damned!) because so many emerging markets are now in the throes of the commodities-intense period of their development. Here’s a typical comment:
…China’s capital stock is only 30% of that of the US. On a per capita basis these numbers are even starker, with a per capita capital stock that is only around 8% of the US.
As for the capital stock, China does have more people than the US so it seems intuitive it would need a lot more steel. And on a per capita basis China does look even more starkly steel-deficient compared to the US or Japan or South Korea.
However, those economies are all significantly wealthy and are well and truly developed (even if South Korea doesn’t quite make the grade everywhere). So what exactly does it prove to argue that China uses less steel per person, when the same could be said of any developing country?
The alternative view is that China has an awful lot of steel already for an economy of its size. And it may not keep growing as fast as it has done in recent years.
That second view, which we first described here after reading a note from Nomura’s China steel analysts, has become so accepted that it’s increasingly difficult to find anyone who is sticking by the “per capita”or capital stock approach (which also assumes continued, and very rapid, GDP growth).
Pimco’s head of Asian credit research, Raja Mukherji, adopts a similar theme in a short note on the subject published this month, and he adds some recent figures on China’s steel capacity which make the argument more compelling.
Mukherji points out the 2008/9 stimulus built up massive demand for steel via big fixed assets and particularly, construction. However, China’s total capacity for producing steel also grew sharply, and now appears to have overshot.
Note that the utilisation rate of China’s steel-producing capacity is falling:
And is likely to become even worse, says Mukherji. China’s current steel capacity is estimated to be about 850m metric tonnes per year, but in 2011 it produced only 683m tonnes. In addition, only 631m of last year’s output was consumed, leaving about 52m tonnes unused. Mukherji writes:
Even if consumption remained flat for 2012, there should only be “true” expected demand for new production of roughly 580 million tons for 2012 (683 million tons minus 52 million tons). Demand is not likely to rise, in our view, because stimulus plans have resulted in an estimated consumption of an additional 120 to 140 million tons of infrastructure-related steel from 2008 to 2012 – and we do not expect new stimulus programs to be announced in the coming years.
We’d point out here that some of this may have been run down in exports; China’s steel mill exports have been rising. They reached 5m tonnes in June while imports have been steady at about 1m tonnes a month. So the 52m may be a bit of an overstatement, unless the ‘consumption’ figure quoted by Pimco includes steel mill exports.
As for Mukherji’s other point — the stimulus — it certainly is reasonable to be cautious. Details of new stimulus programmes are rather elusive; but the central government has been quite clear that it isn’t going to repeat the sort of stimulus launched in 2008.
So Mukherji, and Pimco, think China’s future consumption will look more like this:
He writes (emphasis ours):
“Simply having a large population and a willing government did not spur the type of economic growth, urbanization and steel consumption that China has experienced in the past decade. The catalyst, in our view, was foreign capital investment in the manufacturing sector that was motivated by China’s low-cost labor advantage. An export-based investment model drove economic growth, which led to capital formation, which facilitated urbanization and steel consumption. If this model is in the process of changing, we should expect a negative impact on China’s ability to consume more steel.”
That’s the point: rapid growth doesn’t happen just because it happened in the past, or because it happened somewhere else. And it doesn’t happen because there are lots of people and everyone would like it to happen.
There’s another, more fundamental problem with ever-growing steel consumption forecasts, which Nomura’s Matthew Cross and Ivan Lee discussed in detail almost a year ago. We’ll touch on this only briefly as we wrote about it here – with the example of a comparison of the steel intensity of China’s GDP, which is out of the ball park:
Which might just have something to do with China’s unusually capital-intensive economy.
Cross and Lee also point out that annual increases in steel consumption rates are themselves problematic because the 2008-09 stimulus, which is only petering out this year, was overly focused on steel-intensive infrastructure. There’s much more about their ideas here.
Add to that the inevitable, if delayed, rebalancing of China’s economy, and China’s growing steel consumption seems much less assured.
And the second effort, again cross-posted:
The rapid march towards urbanisation
In our first post in this series, we examined the widely-held belief that China’s steel demand will continue to rise at a rapid rate. FT Alphaville, along with others, contend that such forecasts are on shaky ground. This is, in part, because of the dubiousness of one of the underlying assumptions: that China will rapidly urbanise more of its population. (Here’s a very recent example of this argument, from Stephen Roach.)
The proportion Chinese living in urban areas just passed the 50 per cent mark in the past year but, the story goes, there is more to come. This will in turn mean more industrialisation, more modernisation, a bigger and consuming middle class and of course more GDP growth. In other words:
[…] farmers who once led simple, subsistence-level lives now become factory and service workers in the city, reside in apartments furnished with appliances, occasionally eat out, and perhaps even send their kids to college. In the process, self-sufficient rural households are transformed into workers receiving higher wages and participating in the commodity economy of consumption. As such, urbanization is as much an economic and social transformation as it is a spatial and demographic process.
Sounds great doesn’t it?
The above quote however comes from a paper by Kam Wing Chan in Eurasian Geography and Economics early this year. Chan is a professor of geography at the University of Washington, and he doesn’t agree that this is how things will continue to play out for China. In fact, he argues that even China’s urbanisation to date has been misunderstood and much of it is not true urbanisation, at least not of the sort that some China experts would have us believe. He points to a Brookings Institution paperforecast that China’s middle class will rise from 12 per cent of the population in 2010 to 50 per cent in 2021:
That means that the size of the “consuming class” will surge from ca. 150 million to about 670, a leap of about 520 million in a span of slightly more than a decade! Where to find these additional consumers? The prime driver of this meteoric, euphoric advance in consumption and prosperity, according to that popular narrative, is urbanization. The narrative asserts that China’s rural migrants arriving in cities will become more productive workers and will consume much more, following the experience of other countries in the past.
Firstly, Chan points out that this urbanisation/modernisation model, while representing recent history for much of the West, has not played out everywhere. Latin America has some notable exceptions where increasing urbanisation has mostly just meant that the rural poor become urban poor. This might be to do with the increasingly global division of labour, writes Chan.
When the UK went through its process of industrialisation /modernisation /urbanisation and subsequently developed a sizeable middle class, offshoring wasn’t quite such a big deal as it is today, and automation was a lot less advanced too. We wonder if colonialism might have also played a part, but there are certainly other more recent instances that are closer to the industrialisation /modernisation /urbanisation/middle class ideal — Japan and South Korea, for example. It just hasn’t played out that way everywhere.
However, the most glaring oversight in most references to China’s urbanisation, says Chan, relates to the country’s “hukou” system of hereditary residency and citizenship rights, which creates important distinctions within the urban population.
Most people interested in China will have heard of the hukou system, but studies of China’s urbanisation are surprisingly light on analysis of its effect, according to Chan, and even those who do examine it run into problems with the inconsistency of official statistics and their “unsystematic lexicon”.
A short history: in the Maoist era (1949-1978), peasants were forbidden to move to the cities and “confined to tilling the soil to grow food for urban workers”. This began to change in 1979, but the rural peasants who moved to cities were only allowed to hold menial and often dangerous jobs there.
As a simple overview, the hukou system includes both a ‘type’ (agricultural and non-agricultural) and a registration, which attaches a person to a particular place. People with non-agricultural hukou are far more likely to receive pensions, basic health care, unemployment insurance and so on (these things do exist in China; just not for everyone). This is a simplification of the system – for example moving from one city to another might reduce some of the social services available to someone with non-agricultural hukou — but you get the picture. Hukou is a big determinant of one’s economic opportunities and behaviours.
Some districts carried out changes to their local hukou system early in the last decade but Chan argues that these were mostly ineffective. There has been little progress in hukou reform in recent years.
Because your hukou status theoretically determines whether you are eligible for various benefits and rights, it also determines whether or not you can really join the “middle classes”.
The urban-dwellers with agricultural hukou, or “rural migrant workers”, as they’re often described, continue to live a second-class life:
The great majority are in low-paying jobs and have little purchasing power that would allow them to spend on major appliances as members of the new urban middle class would do (Huang, 2010).13 A very low percentage of them has social insurance coverage (Table 2), and even for those who have “coverage,” the scope is very limited compared to the urban workers.
So there are really two urban populations, the ‘non-agricultural’ hukou, and the ‘agricultural’ hukou. Add them together and you have what Chan calls the de facto urban population:
Guess which one most western reports refer to? The confusion is not helped by the fact that China’s own statistics bureau, with the help of the UN, also began tallying the ‘de facto’ population in the 1980s.
Chan’s analysis suggests that those big arguments over how China defines “cities” are missing the a key point. It’s how you define urban people that matters.
Chan describes the urban-dwelling agricultural hukou workers like so:
The fact that they are purposely held down as a massive, permanent second-class is precisely what supplies China with a huge, almost inexhaustible, pool of super-exploitable labor (Pun, 2005; Chan, 2010b). It is no accident that China has been the world’s largest (and most “competitive”) manufacturing powerhouse since the strategy of allowing migrants to work in the city has been implemented.
That last line chimes with the line in a Pimco note we referred to in our first post of this series, which argues that China’s real growth came from its exports and not its numerous — and frequently misallocated — infrastructure investments. Those infrastructure investments have become a bigger source of growth than exports in recent years, leading some analysts to conclude that this means China’s economy has become less vulnerable to external shocks.
However, Chan’s research suggests China’s urbanisation to date hasn’t created a booming middle class, and it won’t unless there are major reforms to the hukou system.
So we’ve seen a surge of Chinese growth driven by export-focused manufacturing, followed by a surge of Chinese growth driven by infrastructure boom with its attendant financially repressive policies. Neither of these have created a surge in the number of “middle class” — in fact, although China hasn’t published its Gini coefficientfor 11 years, inequality is widely believed to have become more extreme in that time.