| |
Contributions from
the Column
Tribune
The importance of basic sanitary services
Human rights: “Make our position unmistakably clear”
No justification for dams
Beautified data
01/2007
|
|
[ Poverty ]
Beautified data
Internationally, World Bank statistics on income are considered reliable indicators. However, some experts blame the Bank of under-estimating the number of poor people in the world and painting unreliable pictures of increasing or decreasing poverty.
[ By Bernd Ludermann ]
Arbitrary definitions need to be made in order to calculate global poverty. The World Bank produces statistics on how many people, in various countries at different times, have incomes below a level defined as an “absolute” minimum. In order to do so, the Bank first had to define that minimum. Sixteen years ago, it analysed the official poverty lines of 33 countries for 1985.
Based on the definitions used by the poorest countries, the Bank then decided that anyone with a daily income equivalent of less than one dollar’s purchasing power in 1985 was “absolutely” or “extremely” poor. Anyone earning up to twice that amount was considered poor. Despite some methodological problems, the Bank’s definition has become internationally accepted. A dollar a day is slightly above India’s official poverty line.
Seven years ago, the World Bank adjusted its international benchmark to 1.08 and 2.15 dollars respectively at 1993 values. According to calculations done by the United Nation Development Programme (UNDP), the new figures are below the poverty lines of the world’s poorest countries.
Thomas Pogge, a philosophy professor at Columbia University, finds the way the World Bank converts currencies and factors in inflation unsuitable for measuring poverty. Pogge elaborated his view at a conference hosted by terre des hommes, a charity, and the Global Policy Forum Europe in Bonn in December.
In a first step, the World Bank converts its dollar definitions into national currencies. Official exchange rates are not suitable for doing so, because they only reflect prices of products traded internationally. The prices of locally produced goods are typically lower in poor countries, and so are living expenses on average. Accordingly, the local purchasing power of a national currency is normally greater than its exchange rate would indicate. The World Bank therefore bases its calculations on “purchasing power parities”, which compare what currencies could actually buy locally in 1993. On this base, the Bank has calculated national equivalents of its absolute poverty line in 1993. If it wants to establish how many people fell below those levels in any other year, it must convert incomes into 1993 currency units. The Bank does so by applying the inflation rates officially reported by the respective governments.
According to Pogge, this approach is misleading. He says its results are distorted because purchasing power parities are based on the prices of goods that do not reflect poor people’s consumption patterns. The same is true of national inflation statistics. The poor normally do not spend money on what is particularly cheap in developing countries, Pogge insists. For instance, they do not employ household helpers. They spend most of their money on food – and that, Pogge points out, is traded internationally and, accordingly, not much cheaper in poor countries than in rich ones. In Pogge’s view, the World Bank therefore systematically over-estimates the purchasing power of the poor.
Up to a point, his objection seems valid. Some commodities, rice for instance, are indeed traded internationally. Others, like African yams and cooking bananas, are almost exclusively grown for local consumption. Moreover, some governments subsidise and control the price of staple foods. Accordingly, Pogge’s claim that the World Bank is over-estimating the purchasing power of the poor particularly applies to those countries where the staples consumed are internationally-traded and not subsidised.
The poor do not buy computers
Every official national inflation rate is based on some kind of basket of consumer goods considered “average”. Such definitions vary from country to country. Therefore, the data are not really comparable. Moreover, what governments include in these baskets changes over time. As countries become richer, more services tend to be included. On top of that, prices change at different rates – for instance, computers and phone charges may become cheaper in relation to food. Combined, these factors make accurate comparison impossible over time. Therefore, Pogge argues that claims of poor people’s rising or falling purchasing power are questionable, as long as they are simply based on such standard data.
He insists that realistic poverty lines have to reflect poor people’s locally prevalent consumption expenses. Together with Sanjay Reddy, an economist, he has calculated that the number of people classified as extremely poor is probably some two-fifths higher than the World Bank indicates.
World Bank economists Shaohua Chen and Martin Ravallion, however, are prepared to defend their figures. They point out, among other things, that they have been blamed by other economists of grossly over-estimating poverty. In this case, the contentious issue is not the method of calculation but the underlying data.
The World Bank relies on data from some 450 household surveys conducted or commissioned by government agencies in nearly 100 developing countries (accounting for over 90% of the people in the developing world). However, the data available for Africa in the 1980s is sparse. Moreover, the surveys are done according to diverging methodologies and only comparable to a limited extent. Instead of using household surveys, therefore, some economists turn to GDP statistics as a base to calculate household incomes.
Such statistics – compiled by Surjit Bhalla, for example – show a much sharper downturn in extreme income poverty in recent decades than World Bank figures indicate.
The reasons for this disparity have not been fully explained yet. One is probably that the World Bank looks only at consumer spending, disregarding savings. Another is that GDP statistics – unlike household surveys – only keep track of income from market transactions. They do not include goods that, like firewood in many countries, do not cost money, but only labour. In this respect, GDP-based calculations under-estimate the incomes of the very poor.
On the other hand, they over-estimate the impact of economic growth on poor people’s incomes. Growth normally goes along with the transformation of free commodities into merchandise. Accordingly, poor people must spend an increasing amount of their monetary income on goods they did not have to pay for before.
Poverty is not just a matter of income
There is another reason why figures purporting to show the global number of people living on less than a dollar a day should be viewed with skepticism. After all, poverty is not simply about income. Basic social services matter, too. Therefore, the UNDP uses statistics of child mortality, life expectancy and education to calculate its Human Development Index (HDI).
Accordingly, the HDI reflects the quality of basic social services. However, it does so only with considerable time-lags. Social Watch, an international network of non-governmental organisations, has therefore developed another index, the Basic Capabilities Index (BCI). According to Social Watch’s Karina Batthyany, the BCI is based on three indicators: mortality of children under-5, births attended by skilled personnel, and the number of children staying in school through to fifth grade.
This index, however, also has some shortcomings. Professor Pogge finds it too focused on children, stressing it fails to take unemployment and malnutrition into account. Batthyany concedes that the BCI does little to improve the way global poverty is measured. Its particular value, however, lies in its usefulness for local advocacy. It measures deficits of basic social services, and the data can easily be compiled – for instance, by non-governmental organisations – for individual areas without cumbersome household surveys. In the Philippines, NGOs used the BCI to remind some local authorities of their duties. The way the BCI is designed, Batthyany says, facilitates such political leverage.
Klaus Heidel of the Heidelberg-based Werkstatt Ökonomie, an non-governmental economics think tank, believes such a politicised approach to using statistics is entirely appropriate as it is impossible to measure poverty in a “neutral”, “objective” or “non-partisan” way. Philosophy professor Pogge agrees. In his view, the World Bank is deliberately understating poverty. Nonetheless, he does not regard statistics for income poverty as useless, as they may indicate trends and can help to identify the countries worst affected by poverty. Heidel says that, as long as their limitations are understood, even poor indicators can tell us something.
References:
UNDP, 2004:
Dollar a day – how much does it say? Poverty in focus, September, with articles by T. N. Srinivasan, Sanjay Reddy, Nanak Kakwani and Martin Ravallion.
http://www.undp-povertycentre.org/ipcpublications.htm
Angus Deaton, 2004:
Measuring poverty. Research program in development studies. Princeton University
http://www.princeton.edu/~deaton/poverty.html
The Economist, 11. March 2004:
Global economic inequality: More or less equal?
Shaohua Chen and Martin Ravallion, 2004:
How have the world's poorest fared since the early 1980s? World Bank Research Observer
Bernd Ludermann
is a freelance journalist based in Essen, specialising in development issues.
bernd.ludermann@t-online.de
|
