The middle-income trap, again?

The rapidly rising economic weight of developing countries – now in its third decade, rather than a product of the crisis, is notable for several interrelated developments:

• Developing economies as a whole have been growing faster than advanced economies since the 1970s, on both aggregate and per capita terms. (Read more about growth poles - .pdf)

• The margin between these growth rates has risen of late, although growth paths have become more synchronized. Decoupled in trend terms, more coupled in cyclical terms? (There is an abundant discussion of “decoupling”: see one example here and one here)

• The share of developing countries in global GDP has been rising since the 1960s and now amounts to about a fourth measured at market exchange rates and close to half in PPP (purchasing power parity).

• Relative to advanced economies, GDP per capita has risen steadily only in East Asia and the Pacific. Among individual countries, only about half have converged to advanced economies.

• Of all the low-income countries in 1960, almost two-thirds remain low-income and only one (Oman) has risen to high-income. About a third, therefore, has made a transition to middle-income status.

And here is the interesting part: Despite the rising weight of developing countries and the optimism that developing countries will support robust global growth even with lackluster expansion in the advanced countries, the large majority of developing countries make very little progress in terms of their incomes per capita. Of the countries that were middle-income in 1960, almost three-fourths remained middle-income or regressed to low-income by 2009. The only ones that made it to high-income are countries in Western Europe, Japan, the newly industrialized states, and two island economies in Latin America. Is this the middle-income trap that Indermitt Gill and Homi Kharas defined in the East Asia Renaissance (World Bank, 2007)?  Can countries move gradually from middle- to high-income, or is any gradual move destined to fail? Or is fifty years too short a period for economic development to take hold?

Here is the plot for thought:

(The horizontal axis plots GNI per capita in 1960 and the vertical plots the values for 2009, in logs and current U.S. dollars.  The red lines mark the middle-income and high-income thresholds (2009 as defined by the World Bank; the ones for 1960 I obtained by deflating the 2009 value with the SDR deflator).