at different levels? Are the needs of those in the middle determined by what those above them have while the needs of those at the bottom are determined by what those in the middle have? For the purposes of defining poverty, needs are probably best understood in relation to what those in the middle of the society have, and people are further in poverty the further they are from the median. However, the standards, even the standards of poverty, are probably influenced as well by those at the top of the society.22
The emphasis here on the inextricable connection between income distribution and poverty is not new. On the one hand, in high income countries, the standard for defining poverty has often been a relative one, connected to income distribution – for example, defining poverty as existing when a person’s income is less than one-half the mean. Indeed, countries in the European Union use 60 percent of the median income to define the poverty line.23 The argument here, however, is that the same reasoning that leads to a relative definition of poverty in high income countries, a definition involving the distribution of income, applies in low-income countries as well – allowing for qualifications taking account of the extreme conditions of starvation and hunger.
On the other hand, there has been an increasing recognition of the importance of income distribution’s impact on poverty in low-income countries, even within the context of an absolute poverty definition. If economic growth is accompanied by equalization of the distribution of income, absolute poverty (the two-dollars-per-day definition) is reduced more than were the growth to take place without equalization. Because economic growth does not automatically mean greater income equality, there is concern for structuring growth in ways that would accomplish this more favorable reduction of absolute poverty. For example, Azizur Rahman Khan (2005), examining experience in a broad set of low-income countries and using an absolute measure of poverty, argues that increasing employment is a foundation for poverty reduction because of its impact on income distribution. While employment expansion reduces poverty through economic growth, its impact through shifting income distribution tends to be greater. As Khan (p. 8) notes: “A crude estimation of the elasticities of headcount rate poverty … shows that the partial elasticity with respect to the Gini ratio of expenditure is higher than the absolute value of the partial elasticity with respect to PPP$ income.” Similarly, in a broad analysis of poverty in Bangladesh, Binayak Sen, David Hulme and their associates (2004), using the absolute concept of poverty, give similar attention to the “growth elasticity” and “inequality elasticity” of poverty to evaluate change in recent decades. They point out, in particular, how, with rising inequality in Bangladesh, economic growth has reduced poverty significantly less than otherwise would have been the case.
The argument here, however, is based on a different definition of poverty and implies not only that growth without greater equality will be less effective in reducing poverty
22 It is, however, admittedly complex and exceedingly difficult, if not impossible, to clearly specify what “the society” is for purposes of understanding the standards by which poverty is determined. People are members of several societies, defined over ever larger geographic and conceptual spaces. Are standards set by the neighborhood? by the city or town? by the nation? by the so-called “global village”?
23 See, for example, Eurostat (2000).