On August 26, the World Bank released a set of updated and improved estimates of the extent of global poverty and its movement over time (Shaohua Chen and Martin Ravallion, The Developing World is Poorer than we Thought, But No Less Successful in the Fight Against Poverty, Policy Research Working Paper 4703, August 2008, Development Research Group, World Bank). They suggest that in 2005 about 26 per cent of the world’s population was living in poverty, which is nine percentage points more than the 17 per cent figure yielded by the methodology adopted hitherto. This implies that 468 million more people, who were not accounted for in the earlier poverty estimates, were living in poverty in 2005.
However, according to the World Bank, this increase in the incidence of poverty in 2005 is not associated with any change in the rate of decline in poverty between 1981 and 2005. In fact, the trend rate of decline in the incidence of poverty is slightly higher at 0.98 per cent a year, as opposed to 0.83 per cent a year, over the 25 year period starting 1981. As a result, the percentage of the population of the developing world living below the poverty line almost halved from 52 per cent to 26 per cent between 1981 and 2005, with the number of poor falling from 1.9 billion to 1.4 billion.
The new estimates are based on a new poverty line of $1.25 per day in 2005 purchasing power parity (PPP) dollars as opposed to the $1.08 a day in the 1993 PPPs used earlier. PPP estimates compute the equivalence between different national currencies and the U.S. dollar, by taking into account differences in the prices of comparable goods prevailing in different countries. Thus if a particular basket of goods and services consumed “on average” across the world costs a hundred dollars in the U.S., the sum in rupees required to purchase the same basket in India would be treated as equivalent to $100 in purchasing power (PPP) terms. In what came to be known as the “Penn effect” (because it was first noted in a joint project of the United Nations and the University of Pennsylvania), the equivalences between national currencies in developing countries and the U.S. dollar as estimated on a purchasing power parity basis were very different from the equivalences captured by actually prevailing exchange rates. Given the role of lower wages in making the prices of goods that are not freely traded much cheaper in poorer countries, PPPs in developing countries equate less units of the local currency to a dollar than indicated by exchange rates.
The practice of computing the values of appropriate bundles of comparable commodities in different countries began in 1968, with a small sample of just 10 countries. Since then, PPPs based on price surveys in an increasing sample of countries have been computed at periodic intervals, with the one preceding the most recent being in 1993. It is because of the release of a better set of PPP estimates for 2005 involving a larger number (146 as opposed to 117) of countries and more attention to the cross-country comparability of goods and services for which prices were being collected, that the World Bank has had to revise its poverty estimates. The new PPP estimates suggest that the 1993 estimates significantly overstated the purchasing power of local currencies in many developing countries. Thus PPPs yielded by the 1993 estimates equated 1.80 yuan and 7 rupees to a dollar in China and India respectively in 2005. Those figures rise to 3.46 yuan and 16 rupees if the estimates based on the 2005 surveys are used.
Such substantially higher PPPs have two implications for estimates of poverty. First, the international poverty line would be lower than it would have been if the inflation-adjusted poverty lines for 2005 were converted to dollars using 1993 PPPs. A lower poverty line would imply less poverty, everything else remaining the same. Second, a higher PPP means that the dollar values of the consumption levels in developing countries would be lower. Lower consumption would increase aggregate poverty for any given international poverty line. The changes in poverty reported above indicate that the second of these effects dominates the first.
The World Bank is not too perturbed by the sharp increase in the number of income poor, partly because the shift to a new PPP affects only the levels of poverty. The rate of change in poverty incidence over time is not affected by the revision, since poverty estimates for previous years are derived using national consumer expenditure or income surveys and national consumer price indices. The rate of change should, therefore, remain the same as before, because it is the same PPP which is applied to values for all years. It is possibly because of the revision of the poverty line and/or of adjustments for the urban bias implicit in the PPP estimates that the new estimates point to a slightly faster rate of reduction in poverty incidence between 1981 and 2005 than indicated by the estimates based on 1993 PPPs.
This faster rate of decline is of considerable significance for the World Bank. The period after 1980 coincides more or less with the period of accelerated globalisation in the post-World War II period, and the new estimates confirm the view favoured by the Bank that globalisation has helped poverty alleviation rather than slowed its pace as critiques of the process argue. Further, If the pace of poverty reduction realised over the past few years persists, the headcount poverty ratio would stand at 16.9 per cent in 2015, as compared with 41.7 per cent in 1990, implying that the developing world as a whole is on track to realise the Millennium Development Goal (MDG) of halving the 1990 poverty rate by that target date. However, this result is largely the result of developments in China, so that excluding China, the developing world is not on track to realise this goal.
There is, however, reason to be sceptical about inter-temporal comparisons and projections of this sort. These comparisons are made by constructing, through means that are not always rigorous, a long term series on global poverty. The number of countries for which consumer expenditure or income surveys were available in each of the years before the mid-1990s was small, necessitating interpolation or extrapolation to generate the relevant estimates. Moreover, the quality and comparability of expenditure and income surveys are known to vary widely across countries and time. Finally, as Sanjay Reddy of Columbia University has argued, PPP estimates “reflect the relative costs (across countries) for a worldwide pattern of consumption prevailing at only one moment in time, and this pattern is constantly changing. The notion that the new PPPs constitute merely an ‘update’ which better captures the reality over the entire period being assessed is badly mistaken” (“The World Bank’s New Poverty Estimates: Digging Deeper into a Hole,” available at www.socialanalysis.org).
If robustness of estimates is a concern for the Bank, then it should not be using these figures for comparisons over time. But the comparison permits the Bank to highlight (as in the title of the paper reporting the new estimates) the ostensible success of the fight against poverty during the globalisation years.
The point is that there is much else that the new (and old) estimates show which could have been highlighted, rather than a suspect inter-temporal comparison. First, China, which accounted for 38 per cent of the world’s poor in 1990, saw its share fall to 12 per cent by 2005. India, on the other hand, saw its share rise from 22 to 30 per cent. There are stark differences in poverty reduction rates across countries. Secondly, a small increase in the poverty line from $1.25 a day to $1.45 a day increases the number of people below the poverty line in all developing countries by 22 per cent (from 1.4 billion to 1.72 billion) and in India by 30 per cent (from 456 million to 590 million). This bunching of the population around the poverty line is of significance since, as Reddy notes, “a human being could not live in the US on $1.25 a day in 2005 (or $1.40 in 2008), nor therefore on an equivalent amount elsewhere.” We, therefore, are still considerably underestimating poverty. Emphasising features such as these, rather than the change in poverty over time, may help fight the complacent view held by many developing country policy-makers that growth is winning the war against poverty.
(C.P. Chandrasekhar is Professor of Economics at Jawaharlal Nehru University, New Delhi.)
Article courtesy, The Hindu newspaper