World Bank - Most Asians were dead in 1950

Knowledge about the true level of a country's income serves many purposes. It informs us about present and future economic power, about the magnitude of poverty, etc. It can also inform us about the possibility of when a country, e.g. India might host the Olympics - as a spate of articles have speculated recently.

But how do we know the true comparable income of a country? Should we use incomes measured in a common currency, for example US dollars? Most likely not, because one would often get ridiculous results. By this method, we would have concluded that India was richer by 10 per cent in March 2007 compared to a month earlier.

If you recall, that was the month that in a flight of fancy the Ministry of Finance allowed the rupee to appreciate by 10 per cent. [Gee whiz, some people thought - if we let the currency appreciate just 20 per cent a year, India would be ahead of China in the next few years. And we would have zero inflation to boot!]. Economists have long recognised the proclivity of politicians (and markets) to do unusual things. Hence, the development of data on a purchasing power parity basis.

This exercise was started in 1970 with 10 countries (including India) and is repeated every decade or so. The last such exercise was conducted in 2005 and covered more than a 100 countries. According to the "new" PPP estimate, Asian incomes are about 40 per cent lower. Stated differently, the new income estimates imply that most Asians were dead in 1950.

Table 1 shows the two estimates of PPP income in 2005 - the first based on the 1993 PPP survey, and the second based on the new 2005 survey. Something very peculiar is going on here, or as Shakespeare would say, something rotten has happened in Asia. Alone among the vast number of countries, the new income estimates for Asia are wildly off.

The new estimates for developed economies and Latin America are near identical to the old estimate, while estimates for Eastern Europe (including the former Soviet Union) and sub-Saharan Africa are lower by approximately 13 per cent. For 148 countries, developed and developing, the new is much like the old; on an aggregate basis, only 7 per cent lower. But for Asia, we have a different metric - incomes are reported to be 41 per cent lower, and it does not matter whether India and China are included or excluded.

No one I have talked to, or read, has come up with any explanation for this radically off the wall, over the top (bottom, actually) estimate of incomes for Asia. But there are rumours. Could it be that this was an attempt by the international organisations, and China, to portray China as not so rich in order that its exchange rate did not appear so deeply undervalued? Exchange rates tend to appreciate with an increase in per capita income; a lower estimate of PPP income would suggest a lower magnitude of revaluation for the yuan.

As we all know, and as several governments in the developing and developed world contend, the Chinese currency has been deeply undervalued, upwards of 50 per cent, for several years now. This undervaluation may be responsible for all the extra growth that China has witnessed for the last decade or so. The US government has constantly been engaging the Chinese government with facts pertaining to China's trade surplus, its per capita income levels, and the massively undervalued yuan.

As I said, these are just rumours; if there is an alternative, more benign reason for the strange data for Asia, perhaps the World Bank, IMF, (and the agency responsible for the Asian estimates, the Asian Development Bank) could provide us with the an explanation.

There are other more serious implications if we accept the new data at face value - problems for the world much greater than the undervaluation of the Chinese exchange rate. The most important implication is that the new estimates of income, and consumption, projected backwards with domestic GDP growth rates, result in incredibly low incomes for most of Asia in 1950: Incomes low enough to suggest that most Asians were dead in 1950.

The calculation is straightforward and contained in Table 2. There are three steps involved. First, estimates of per capita PPP consumption in 1950, at 1993 PPP prices, are obtained. Second, this estimate is adjusted upward or downward, according to the nearest household survey conducted in each country.

This adjustment is necessary because the official estimator of poverty, the World Bank, insists that this is the only accurate method of estimating consumption of the population, and especially the poor population. For India, this ratio is 87.5 per cent, and for China about the same. Thus, for both countries, the "actual" per capita consumption estimate in 1950 is about 90 per cent of the PPP national accounts estimate.

The next step involves a reduction of consumption by the factor suggested to be "true" in 2005 (reported in Table 1 on an aggregate basis). Finally, the nearest distribution of consumption data are used to derive average consumption for each quintile.

The World Bank defines a person as poor if she consumes, on average, less than $1.08 per day in 1993 prices. This level of consumption is a bare minimum level, a level consistent with bare survival. But according to the new data, all of Asia had an average level of consumption of only 0.71 1993 PPP$ per day. In contrast, the estimate for sub-Saharan Africa is $2.07, and that for Latin America, $2.32. But maybe only some people were very poor in Asia. Not so.

The average consumption of the bottom 80 per cent of the Asian population was 0.49, or less than half the absolute minimum poverty line. Are these numbers plausible? If yes, then you must not be reading it on planet Earth, because your grandparents did not survive the World Bank induced crash of 2005.

The author is Chairman, Oxus Investments, a New Delhi based asset management company. The views expressed are personal.