Poor empiricism: The ‘Middle Income Trap’ C.P. Chandrasekhar

Increasing evidence that the era of high growth in Asia may be nearing its end has triggered speculation on ways to revive growth in the region. It has also challenged the belief that more developing countries would like the first generation new industrialisers in Asia (South Korea, Singapore, Taiwan and Hong Kong) transit to developed country status in a relatively short period of time. This has spawned a new industry involving the use of multi-country, inter-temporal GDP numbers to identify the countries that have escaped being stuck in the so-called “middle income trap” and the lessons that can be learned from them. Academic economists (Barry Eichengreen, Donghyun Park, and Kwanho Shin, 2013) and international institutions like the IMF (Regional Economic Outlook: Asia and Pacific, April 2013) and the ADB (Jesus Felipe, March 2012) have jumped on to the bandwagon.

A typical analysis would first use the data to say something of the following kind: Growth slowdowns are more likely to occur when countries reach income levels (measured in PPP terms) that identify them as being in the “middle income range”. But some countries, such as the first tier new industrialisers in Asia, managed to escape this middle income trap. Examining their experience (even though they are few in number) points to what needs to be done if others such as China, India, Indonesia, Malaysia, and Vietnam are to ensure sustained growth that takes them to developed-country status.

However, per capita income, whether measured in nominal or PPP terms, says little about an economy. At similar levels of per capita income countries can be characterised, among other things, by very different distributions of that income, vastly different economic structures, substantial differences in their exposure to foreign trade and investment flows and very different past and current policy environments. This is likely to affect their current performance and their ability to improve on that. It would be naïve to presume that there are particular income levels at which all countries would face similar constraints on growth requiring similar policies to exit from the growth trap. Many poor countries too face severe growth constraints, though some have done quite well in certain periods. And that is true of middle income countries as well. All that can be said is that capitalist development tends to be uneven across geographies and unstable across time. Growth convergence is not the norm.

The effort to work the evidence (by, for example, ‘appropriately’ defining what constitutes middle income) to argue that there exists a specific ‘middle income trap’ is empirically not convincing. To use exceptions to this empirical ‘rule’ to identify policies to escape that trap is even less persuasive. Not surprisingly the conclusions from such exercises are not very revealing and often a reflection of the predilections and prejudices of the authors. A common conclusion is that middle income growth slowdowns are less likely in countries where structural diversification and export growth is high and high-technology products account for a relatively large share of exports. That raises more questions rather than serve as an answer. Why, for example, do only a few countries manage to ensure the required structural diversification and export pattern? And would all countries that manage such diversification, irrespective of number, be necessarily successful?

The difficulty in garnering much from this exercise soon becomes an excuse for advancing assertions as conclusions. The IMF’s latest regional outlook on Asia for example suggests that these economies need to pay attention to “institutions” and infrastructure and exploit the “ample room for easing stringent regulations in product and, in some cases, labor markets.” Institutional strength is seen as reflected in higher political stability, better bureaucratic capability, fewer conflicts and less corruption. Whether weakness in this is the result of underdevelopment or a cause of it is open to discussion. On the other hand, there could be little disagreement that if infrastructural bottlenecks exist growth would be adversely affected.

And there are many who argue that growth in Asia stalled not before they liberalised but after they did. This is based in particular on the evidence that dynamism in Asian economies other than China, and to an extent India, faltered after the 1997 crisis. That crisis, we must recall, was related to the financial liberalisation many of these countries were forced to adopt, either as a quid pro quo for continued access to the export markets on which they were excessively dependent, or because waning manufacturing export competitiveness as a result of rising wage costs and appreciating currencies, pushed them into liberalisation of financial policies in the hope of making financial services the new engine of growth. The result was vulnerability to boom-bust cycles of various kinds that led to the synchronised downturn in many countries (with Thailand, Korea, Malaysia and Indonesia, among them) in 1997-98.

This should possibly lead to two conclusions. The first is that, beyond a point export-driven growth has a way of running into internally generated constraints.  Second, that among the factors that can undermine a country’s growth prospects, even at relatively higher income levels, is excessive liberalisation, especially financial liberalisation. Possibly most countries, whether poor, rich or in some ‘middle income’ range, find their growth has stalled for reasons such as these.

(This article was originally published in the Triple Crisis Blog on 28 May 2013)