"[Assume that the figure given above] provides a model for world economic growth in the sense that the actual economic performance of the 18 Western OECD countries over the historical time frame 1500-2001 (red line, cross sectional average of log per capita real income) is a base trajectory for economic evolution that other countries in the world follow, albeit over different time frames, thereby allowing for differing speeds of technological learning across countries. Accordingly, we seek to estimate where other countries lie on the long run base OECD trajectory and to calculate the approximate speed of learning within those countries that is necessary to have achieved their given economic performance.

For many countries, data on annual per capita real income is available only from around 1950, so it is convenient to use the period 1950-2001 to represent recent trends in world economic growth. The OECD historical data base provides estimates of per capita income over the long historical period 1500-2001 but not for all 18 constituent OECD countries and not on an annual basis prior to 1870. Therefore, to obtain a complete base trajectory series for OECD trend growth, we infilled observations over the 1500-1870 using a combination of linear interpolation and coordinate trend fitting, as proposed in Phillips (2004). Next, we formed 9 geographical subgroups from 88 countries in the world economy and used the cross sectional average of their log per capita income to approximate their common growth component. Using this data, we matched the initial and last period incomes of the 9 subgroups in 1950 and 2001, respectively, with the base OECD trajectory and estimated the corresponding speed of learning for each subgroup (G).

Table 1 reports the results. Base OECD trajectory income in 1950 was $5,150 and takes 52 years to reach $20,110 in 2001. Columns 4 and 5 of the table show the base OECD trajectory year corresponding to the observed income levels y_G,1950 and y_G,2001 for each country grouping G. For example, the average initial and final period incomes for the African countries group in 1950 and 2001 were $949 and $1,796 (columns 2 and 3). These figures correspond to base OECD trajectory income in years 1727 and 1860, respectively. We deduce from this calculation that the speed of learning for the African group is approximately 2.61 (column 6). Thus, the African countries can be said to be undergoing growth comparable to that of the OECD base group over 133 years during the industrial revolution in the eighteenth and nineteenth centuries, but the growth experience of these countries is actually compressed into a 52 year time frame during the twentieth century. In an analogous way, initial year 1950 income and final period 2001 income for the Asian Dragons are $1,533 and $18,289, respectively, which places this group on the base OECD trajectory over the period from 1849 to 1995. The Asian Dragons have therefore experienced in 52 years economic growth that is comparable to that of 146 years of growth for the OECD base group up to 1995. The learning speed of the Asian Dragons is 2.86 and is therefore faster than that of the African group. Moreover, the Asian Dragon experience co-relates to a much more recent period of the base OECD trajectory. Similarly, Japan’s speed of technology learning is 1.59 and this implies that it has compressed the last 81 years of base trajectory OECD growth up to 2001 into 52 years.

The table also reveals that the fastest learning countries are China, India and the East Asian group. Remarkably, China has experienced over four centuries of base trajectory OECD growth in the last 52 years taking it to year 1917 levels on the OECD trajectory. India and the East Asian group of countries have experienced more than three and a half centuries of base trajectory growth in 52 years, taking them to mid-nineteenth century OECD levels of income."[1]

I created the image given below afer reading the whole paper (took me two bottles of red wine) but I think it's still ok:


related items:
What Causes Prosperity?, Arnold Kling
Economic Growth & Parameter Heterogeneity, Mahalanobis

[1] Phillips, Peter C.B. and Sul, Donggyu, "Economic Transition and Growth" (May 2005). Cowles Foundation Discussion Paper No. 1514.