Fast economic growth, like all good things, must end. But when?
An NBER working paper by Barry Eichengreen, Donghyun Park and Kwanho Shin released on Monday uses data from 1957 through 2007 to suggest when and why slowdowns occur in fast-growing economies.
Other research has looked at the causes of growth collapses or slowdowns in poor countries but, say the authors, little or no attention has been paid to the determinants of specific slowdowns in countries that have reached a moderate level of average income through “catch-up growth”.
While it’s dangerous to extrapolate cross-country historical data series to individual cases, the implications are still important for Tilt countries, especially China.
Here’s the authors’ definition of a slowdown by a fast-growing economy:
The first condition requires that the seven-year average growth rate is 3.5 percent or greater prior to the slowdown (earlier growth was fast).
The second one identifies a growth slowdown with a decline in the seven-year average growth rate by at least by 2 percentage points (the slowdown is non-negligible).
The third condition limits slowdowns to cases in which per capita GDP is greater than $10,000 in 2005 constant prices (ruling out growth crises in not yet successfully developing economies).
This leaves a little over 100 slowdown examples to use to isolate predictive trends, beginning with the relative contributions of labour, capital and total factor productivity to the slowdown.
A warning: you’re about to read 111 words on growth accounting.
In Table 2.1 we see that the contribution of the growth of the capital stock fell from 2.40 per cent to 1.79 per cent around the time of slowdowns. The contribution of labor growth fell more modestly, from 0.89 to just 0.86 per cent, while that of the growth of human capital actually increased (from 0.44 to 0.51 per cent).
Much more dramatic is the decline in the contribution of TFP growth, from 3.04 to 0.09 per cent. Growth slowdowns, in a nutshell, are productivity growth slowdowns. 85 per cent of the slowdown in the rate of growth of output is explained by the slowdown in the rate of TFP growth.
No real surprise — there are only so many workers that can move from the countryside to urban industries, and there are diminishing returns to the import of foreign capital.
Fortunately, growth accounting can only explain so much — the authors also look at the associations between slowdowns and various economic and political indicators.
They find the most important variable to be the level of per-capita income:
The evidence suggests that rapidly growing economies slow down significantly, in the sense that the growth rate downshifts by at least 2 percentage points, when their per capita incomes reach around $17,000 US in year-2005 constant international prices, a level that China should achieve by or soon after 2015.
Followed by the manufacturing employment share: the peak probability of slowdown occurs when manufacturing is around 23 per cent of total employment.
Interestingly, “financial openness, terms of trade shocks, and political regime changes do not appear to have a significant impact on the likelihood of growth slowdowns”, write the authors. And in case you’re wondering, they run separate analysis including and excluding oil-rich states.
So what do the data say can be done to reduce the probability of slowdowns, or at least to prolong increases in capita income prior to their occurrence?
In short, openness to trade, a high contribution of consumption to GDP and not having an undervalued real exchange rate. But we don’t really know why:
The nature of this association remains, at this point, a matter of speculation. It could be that countries that rely on undervalued exchange rates are more vulnerable to external shocks. It may be that real undervaluation that works well as a mechanism for boosting growth in the early stages of development works less well later, when growth becomes more innovation intensive. It may be that real undervaluation allows imbalances and excesses in exportoriented manufacturing build up.
A low consumption contribution makes some sense, even when not disguised by an undervalued real exchange rate. As the authors point out, a low share of GDP from consumption could reflect price distortions and policy decisions that favour tradeables over non-tradeables and undeveloped financial markets that encourage precautionary savings. Logically, a country less dependent on consumption will suffer more from a reduction in trade or marginal efficiency of investment.
Anyway, this speculation brings us to you know where:
Our results can be used to address the question of whether an abrupt slowdown is likely and if so when. Both China’s openness and high investment rate point away from the likelihood of a slowdown. Other considerations, however, suggest that a slowdown may be coming sooner rather than later.
The paper tentatively forecasts a slowdown in China, as defined by the above criteria, “sometime in the next ten years”.
The authors rightly point out that “special caution” is needed in the case of China. FT Alphaville readers can no doubt recite the arguments for why in their sleep: no huge country has grown so fast for so long, regional diversity, stirrings of homegrown innovations, etc.
But the paper is a useful reminder that all good things must end, and a solid effort at beginning to explain why they do.
Full paper in the usual place.
Related links:
China’s sizzlingly (disappointing) growth record – FT Alphaville
China: Unstoppable powerhouse, or ‘more like us’? – FT Alphaville