A wealth of data

WHAT IS poverty and when is a person poor? Most would agree that poverty involves not having enough of certain things, or doing without others that richer people take for granted. But what is “enough”, which goods and services really matter, and who should decide these questions—researchers, governments or international agencies—are less tractable issues. Perhaps the poor themselves should have the final word. But this presents its own problems. Tabitha, a 44-year-old woman from a slum outside Nairobi, told researchers from Oxford University that going without meals was “normal for us”. Diminished expectations are only one of the effects of dire poverty.

In the world of international development, most have rallied around the “dollar-a-day” poverty line (or more precisely, the $1.25-a-day measure) and its less acute cousin, $2-a-day poverty. These World Bank measures judge a person to be poor if his income falls short of a given level, adjusted for differences in purchasing power. In principle poverty rates based on these measures count the fraction of people in a country who lack the resources to buy a notional, basic basket of goods.

Despite the many merits of the $1-a-day measure—not least its simplicity—some argue that looking only at income risks impoverishing the debate about poverty. Such complaints can be overdone. Income clearly matters: it determines how much people can buy and therefore whether they can afford to do the things, like eat enough, that critics of income-based measures think are more important. But rising incomes do not always translate into better health, say, or better nutrition. So there is clearly scope for measures of poverty that directly capture the many different ways in which, to quote Amartya Sen, “human lives are battered and diminished”.

A new set of internationally comparable data put together by researchers at the Oxford Poverty and Human Development Initiative at the University of Oxford tries to take Mr Sen's ideas about “the need for a multidimensional view of poverty and deprivation” seriously*. Aided by the improved availability of survey data about living conditions for households in over 100 developing countries, the researchers have come up with a new index, called the Multidimensional Poverty Index (MPI), which the United Nations Development Programme (UNDP) will use in its next “Human Development Report” in October.

The index seeks to build up a picture of the prevalence of poverty based on the fraction of households who lack certain basic things. Some of these are material. Does a family home have a dirt or dung floor? Does it lack a decent toilet? Must members of the household travel more than 30 minutes on foot to get clean water to drink? Do they live without electricity? Others relate to education, such as whether any school-age children are not enrolled or whether nobody in the family has finished primary school. Still others concern health, such as whether any member of a household is malnourished. A household is counted as poor if it is deprived on over 30% of the ten indicators used. Researchers can then calculate the percentage of people in each country who are “multidimensionally poor”.

Looking at many aspects of poverty at once has several benefits. One problem with considering just one indicator is that some deprivations may be a matter of choice. As Mr Sen has argued in his work on poverty, what matters is not whether a person eats “enough” but whether he eats whatever he does out of choice. Fasting is fine; involuntary starvation is not. Some, for instance, may prefer the earthiness of a mud floor to the coldness of a concrete one. But the number of people choosing to be malnourished, illiterate, lacking in basic possessions and drinkers of dirty water all at once is probably fleetingly small. A person deprived along many of these dimensions surely counts as poor.

Measure for measure

By and large, as the chart shows, countries' poverty rates as calculated using the MPI differ quite a lot from those based on their $1-a-day rates. In India, for instance, many more people lack basic things, as measured using the MPI, than earn less than $1.25 a day. The opposite, however, is true of Tanzania, which is doing better at getting its people fed, housed and educated than its income-based poverty rate would suggest.

Since the MPI is calculated by adding lots of different things up, it is possible to work backwards and see what contributes the most to poverty in specific places. In sub-Saharan Africa, the material measures contribute much more to poverty than in South Asia, where the biggest contributor is malnutrition. The authors argue that having this information readily accessible makes it easier for development agencies and governments to decide what to focus on. The MPI also does a better job of uncovering long-term trends. Successful reforms in health or education increase earnings only many years into the future but will show up quickly in the MPI poverty rate.

Much remains to be done to refine the idea. For a start, the things the MPI measures are not particularly useful for middle-income countries, which have figured out how to get their people clean water and enough food but where other kinds of poverty still exist. But the principles on which the MPI is based are simple and easily adapted. An index for areas within a single country could draw on more data and could paint an even more nuanced picture: the Mexican government is already using a variant of the index to help monitor the results of its anti-poverty programmes. Measuring poverty is not the same as alleviating it, of course. But the MPI is a step forward.

* “Acute Multidimensional Poverty: A New Index for Developing Countries”, by Sabina Alkire and Emma Maria Santos. OPHI Working Paper 38, July 2010