Globalisation: The measurement of poverty and the poverty of measurement

Guyana the wider world
by Dr Clive Thomas
Stabroek News
April 8, 2001


To make an intelligent appraisal of the debates on globalisation and poverty one needs to be reasonably informed about the technical features of the definitions and measures of global poverty in use. As it is presented in texts the mathematics on this looks formidable, but the ideas behind it are not complicated. Indeed this reflects the real danger of excessive formalism, which present-day economics faces, as technics and the tool kit approach to social analysis replace theories, ideas, and innovative thought. Thus poverty is measured, but the poverty of that measurement is lost in the elegant theorems in which they are presented.

Last week we saw how poverty is measured to give an indication of the magnitude of the problem globally. To recap, based on either income or consumption, these poverty measures allow us to make a "headcount" of the percentage or the total number of persons living below a given poverty line. Overall, the headcount measure has two strong attributes. One is that it is simple and, to most persons, intuitively obvious. It is therefore easily communicated and readily grasped. Secondly, its simplicity and ease of communication makes it easy to develop advocacy on behalf of the poor. Experience has shown that persons are readily motivated to express concern and to do something to alleviate poverty, when they are confronted with the dramatic fact that there are 2.8 billion persons living below the poverty line of US$2 per day, in a world of 6 billion persons.

Beyond these attributes however, the measure has many shortcomings. We need to be familiar with these. One set of shortcomings may be classed as "technical". These generally reflect measurement weaknesses in the indicators. Another set however, may be classed as "conceptual". These derive from inherent weaknesses in the notion of the poverty as it is measured. The latter is the more substantial class of weaknesses.

Technical weaknesses

One example of a technical weakness is that changes in the headcount measure of and by itself does not indicate the condition of those who live below the poverty line. Let us for example assume that the absolute poverty rate in Guyana, which in 1999 was 36.3 percent, is still the same 36.3 percent today in year 2001. This figure makes it appear as if nothing has changed. However, it is quite possible for those living in poverty today to be either further away from the poverty line or closer to it, even though the figure has not changed. Clearly this difference is important. If those who are poor remain poor but move closer to the poverty line, it would clearly take less resources to take them out of poverty, than if it were the other way around.

The depth of poverty

This technical weakness has encouraged the search for a measure that would tell us how far away from the poverty line on average, are the persons who live in poverty. This measure is called the Poverty Gap Index. It is always possible to obtain this index from the same survey data that give the headcount measure. There is no excuse therefore, for this index not being calculated. Yet very often it is not! The Poverty Gap Index seeks to measure the total "deficit" of all the persons who are poor in relation to the poverty line. In other words it measures the depth of poverty. For those technically inclined this is done by multiplying the head-count index by the gap between the average income of those who are poor and the poverty line. This gap is expressed as a proportion or fraction of the poverty line. This measure tells us roughly the amount of national income required to bring the population in poverty above the poverty line.

In the example I gave last week of Guyana, in 1999 the national poverty gap ratio was 12.4. This tells us that it would take about 12 percent of our national income to carry everyone above the poverty line. To recall, the head count measure in Guyana provided an absolute poverty index of 36.3 percent of the population living in poverty and a critical poverty index of 19.1 percent. The latter as we saw is a measure of the amount of income/consumption, required for an individual or household to acquire the food essential for survival. And, the former is the amount of income/consumption required for an individual or household to acquire both the food and non-food items essential for survival.

The severity of poverty

Another technical issue is that just as the poverty line distinguishes between the poor from the non-poor, there is among the poor themselves a distribution of poverty. Some persons are poorer than others, and vice versa. This distributional issue suggests that there are different degrees of severity of poverty. Therefore, while the head-count numbers tell us how many people are poor, and the poverty gap measure tells us what resources are required to move the poor out of poverty, there is still the question of differences in the severity of poverty to be determined. There is a highly technical measure of this. It is called the FGTP2 index. Anyone who follows the various poverty debates and discussions, or reads World Bank and IMF Reports, or for that matter local documents on this topic like the Guyana Human Development Report, the Guyana Survey of Living Conditions, or the National Development Strategy would have come across this measure.

For those technically inclined, the FGTP2 measure seeks to weight the degree of poverty in each poor household in assessing total poverty. The weight used is the square of the individual household's poverty gap. It is this simple! What happens when you square each gap, is that those farthest away from the poverty line (where the gap is large) have more weight in the assessment of poverty than those close to the line (where the gap is small). In this way the measure takes into account differences in the severity of poverty.

Poverty and inequality

This discussion of the distribution of poverty among the poor leads to an important observation, which should be carefully noted. That is, poverty and income distribution or inequality are opposite sides of the same coin of social progress. High levels of poverty are strongly associated with high levels of income and consumption inequality. As we have noted before, at the global level globalisation has led to an increase in the total number of persons living in poverty. This increase mirrors the rapidly widening gap in incomes between rich and poor countries, which we also observed. Within individual countries the picture is the same. Thus to continue with the example of Guyana. Here the poorest 20 percent of the population consumes 9 percent of total consumption, while the richest 20 percent consumes 39 percent of consumption. This is more than four times as much, and is the mirror image of the poverty among us, which affects more than one in every three Guyanese.

Next week we shall look at the conceptual weaknesses in the measures of global poverty. And, on the basis of to-day' s discussion and next week's I shall indicate why supplementary indicators are always necessary if we are to get a correct handle on the situation of global poverty. One needs to recognise at all times, that behind the numbers of the global poor there are real persons, who live lives, dream dreams, and deserve better.