Mzansi Afrika

From Johannesburg South Africa, a window on the world

Tuesday, June 21, 2005

Measuring global poverty

The rhetoric of global poverty statistics

Ingrid Robeyns
2 May 2004

Last week, the World Bank sent out a press release announcing that global poverty has halved since 1981. According to the press release, the number of absolute poor in the developing countries, defined as those living on less than $1 a day, dropped from 1.5 billion in 1981 to 1.1. billion in 2001. In percentages, this amounts to a reduction of extreme poverty from 40 to 21 percent of the total population in the developing countries. Surely, one would think, there is good reason for optimism.

The World Bank’s poverty statistics are widely cited and presented as clear and objective facts. However, it wouldn’t hurt to dig a little deeper and look at possible rhetorical and political dimensions of these statistics. This is important because the global poverty statistics have become the focus of a severe political and ideological dispute between radical proponents of global capitalism, and the alterglobalists. The latter argue that global capitalism is making life worse for many poor people in the world. Defenders of global capitalism, on the contrary, hold that nothing is more efficient in lifting people out of poverty and enhancing development in the global south than unrestrained free market capitalism. If the statistics show that global poverty goes down, defenders of global capitalism interpret this as the ultimate proof that they are right. The poverty statistics also have another important political function, as they will determine whether the countries of the United Nations will succeed in reducing global poverty with 50% by 2015, as stipulated in the Millennium Development Goals.

However, there are two serious problems with the World Bank’s press statement and its global poverty statistics. The problem with its press release is that it tells us that the extremely poor people are those people who have to live on less than a dollar a day. I guess that to many readers of newspapers in Europe and North-America, this sounds plausible. Surely, you might think, the World Bank knows what it is doing. And we believe that one dollar in Malawi or Bangladesh must buy you a lot because we believe that everything is so cheap there. But this is not what these poverty statistics really measure. The familiar "$1 a day" poverty definition is the shorthand for "$1.08 purchasing power parity for the USA in 1993". In other words, according to this definition, you are poor if you have to get by with fewer commodities than what $1.08 could buy you in the USA in 1993. But what could an American have bought with $1.08 in 1993, which is roughly equivalent to $ 1.40 or Euro 1.17 today ?

Moreover, the problems with the World Bank global poverty statistics are much more serious than only this "slight misrepresentation" in the official press release. The more fundamental problem is that the World Bank’s poverty statistics lack any foundations. The "1 dollar a day" poverty line has never been justified in terms of what poor people could buy with this amount, and many scholars and policy makers argue that it is unlikely that anyone can survive in the long run with the local equivalent of one dollar.

The economist Sanjay Reddy and the philosopher Thomas Pogge (both at Columbia University in New York) have argued that the $1 a day poverty measure is arbitrary and lacks any relevant content. This is probably not the right place to go into details of Reddy’s and Pogge’s critique, which includes a number of technical criticisms, but let me briefly indicate the gist of their analysis. As indicated above, the global poverty statistics lack any conceptual

The global poverty count is not based on how we think of poverty, like being undernourished, and being without shelter and other basic necessities. Second, to compare poverty in countries with different currencies and price levels, the World Bank uses so-called purchasing power parity factors. But these factors take into account the price levels of all goods and services in proportion to their share in international consumption expenditure. In this calculation, the prices of basic necessities play a minor role, whereas they play a huge role in the consumption of the poor. Western tourists often think of developing countries as cheap, but this only holds for services (like restaurant meals, taxi’s etc), but relatively speaking not so much for the basic necessities which make up the consumption of poor people.

Thirdly, the World Bank method is internally inconsistent. In 2000 the World Bank ‘updated’ its definition from "1 dollar purchasing power parity for 1985" to "1.08 dollar purchasing power parity for 1993". Pogge and Reddy show that changing the base year can have enormous effects on the poverty estimates; for example, the poverty estimate for Latin-America for 1993 dropped from 23.5 to 15.5 %, simply by "updating" the base year of the poverty definition. These can hardly be considered minor estimation errors. Finally, the data on which the global
poverty statistics are based, are often of poor quality and therefore unreliable.

These mistakes taken together most likely lead to an underestimation of the extent of extreme poverty in the world. We simply don’t know how many people live in absolute
poverty, and whether their number is increasing or decreasing. In any case it is possible
that extreme poverty has increased in the last decade, in contrast to what the World Bank
would like us to believe.

Sanjay Reddy and Thomas Pogge do not limit themselves to criticizing the World Bank
statistics. Their alternative is that our global poverty measures should start from a clear
conceptualization of what it means to be extremely poor, like not having sufficient food, not being sheltered, not having the absolute basic goods such as clothing. They suggest that we should decide on such a poverty definition through an international dialogue. Once we roughly agree on what the absolute minimum requirements are to escape extreme poverty, we can translate this into a financial amount at the local level. Only with such a method will we be able to count the poor in a meaningful way.

As long as the World Bank keeps measuring poverty with its current methods, we have to be critical of what it presents as neutral facts. And researchers, whether working at the World Bank or at universities, should stop pretending that there is no political and rhetorical dimension to economic theories and policies, and the statistics that support them. (Source)


Post a Comment

<< Home