IT'S unanimous: China can't grow at a 10 percent rate anymore. Those days are gone, it's said, because the eastern coast, where most of the development has occurred in the past 30 years, is too crowded and because economic theory says there's been too much investment and it's not "efficient" anymore.

Economic theory does indeed say something close to that. Neoclassical theory is all about "marginalism" - the bang one gets for that last buck spent. On the demand side, the fourth ice cream cone is not as enjoyable as the first. On the supply side, the fourth investment dollar doesn't produce the same kick to gross domestic product.

That's a problem when investment makes up 50 percent of gross domestic product, as it does in China. Everyone knows it's too high. Investment can't be as efficient as it used to be.

Look at the steel and concrete industries. Too much capacity. Ten percent growth might have been possible yesterday, but 7.5 percent is tops today. Investment can't punch like it used to.

As often happens in economics, we get the theory right, or mostly right, and we get the measures wrong - sometimes way wrong. Productivity and efficiency are good examples. Theory assumes that all resources are fully deployed. This may be wildly untrue in the short-run. You can still measure productivity and efficiency in the short run - and many do - but the results are usually nonsense.

Labor productivity

The reason is apparent. Take labor productivity, the usual gauge is output per worker, or some close derivative. Sally may be just as good an analyst this year as last year or even better, but if the market drops and no one buys her services, then her measured productivity tanks big time.

The same is true for capital stocks and investment. The most common measure of efficiency here is the incremental capital-output ratio. It's a mouthful of a term but a simple idea that measures how much investment it takes to generate one more unit of GDP.

A high or rising ratio implies low or falling efficiency of investment because more and more investment is required to generate one more dollar or yuan or euro of GDP.

As one might guess, the capital-output ratio can jump around just as much as labor productivity in the short run, and for reasons that have nothing to do with the efficiency of investment. If a country's GDP growth falls close to zero, the ratio shoots off into infinity. If slow growth turns to a slight contraction, the ratio shoots to negative infinity. Even 20-year moving averages can't smooth that kind of distortion.

With this in mind, how inefficient has China's investment been?

Compared with the rest of the world, China's investment efficiency was about as good as it gets.

At an average of 3.4 between 2003 and 2007, China's capital-output ratio was lower than that of most other place, save for Singapore. On a five-year timeframe - when global resources were fully employed - China's investment was 85 percent more efficient than that of the US, twice as efficient as that of South Korea and nearly 4 times more efficient than that of Japan.

Sadly, China's ratio has risen markedly since 2007. It now stands at 4.7. Over the past four years, investment efficiency has fallen by nearly 40 percent. Does this explain why China's GDP growth has slowed? Does this mean China's long-term GDP growth rate must now be 7.5 percent when it used to be 10 percent?

Actually, it means nothing of the sort.

Capital-output ratios the world over soared when the global financial crisis hit in 2008.

Taiwan's jumped to 30. South Korea's soared to 98, and Singapore's rose to 13 in 2008 before plunging to minus 30 in 2009 as growth turned negative. The ratio in the US dropped to minus 49 in 2008. Put those numbers into a five-year moving average and you won't just get distortion; you get utter nonsense.

While investment efficiency is a key determinant of long-run GDP growth, in the short run, the causality is flipped around. That is, growth determines (measured) efficiency, not vice-versa.

Among other things, this means one cannot use the capital-output ratio of the past four to five years to make any inferences about what potential growth might be in the years ahead.

Where does this leave us then?

From a long-term perspective, China's capital-output ratio looks pretty good and not much different from other Asian countries that preceded China in economic development.

If China's growth can remain fast for longer than most think, it may already be running faster than most think, too. For the past two quarters, virtually every province in the country reported GDP growth faster than the national year-on-year average of 7.5 percent. Any statistics professors will tell you it's impossible for everything to be above average, but in China things seem to work differently.

Different numbers

On average, in the second and third quarters, Beijing and Shanghai reported growth of 7.5 percent, but the rest of the eastern coastal provinces reported growth of 9.2 percent. Growth was even faster in the central provinces at about 10.5 percent, and growth in the western provinces averaged 12.4 percent.

So which growth rate is correct? The official figure of 7.5 percent or the regional weighted average rate of 10.1 percent?

From a logistical point of view, the latter would seem to be the only possibility. The national figure can be derived only from the provinces, or at least one would hope so.

Unless the central authorities are handing out predetermined numbers to the provinces, it's hard to see how the national number could be more credible than the sum of the regional numbers.

A few things remain clear. China's growth must slow as per-capita incomes rise and the eastern coastal provinces become saturated with the trappings of development.

Since 2003, growth has slowed markedly in the East, remained high in central provinces and accelerated in the West.

If China wishes to keep growth as rapid as it has been over the past 20 years, it has a much better chance of doing so than most people think. But the baton will have to pass to the inland provinces, which remain relatively undeveloped. In that light, complaints about inefficient infrastructure investment - the so-called "roads to nowhere" -should be tempered.