When a country wants to measure poverty in its population, it first needs to determine who should be considered as poor. This is the case of individuals who do not attain the level of prosperity or living standards considered as minimal in that country. Economists will talk about a minimal welfare level. Developed countries use income to measure this notion of welfare. They define a poverty line, which is the income threshold allowing purchasing a bundle of goods giving the minimal welfare level. Everyone whose income is below the poverty line is then considered as poor. Because income varies a lot in time in developing countries, they use consumption instead of income in order to measure the welfare of their citizens. The poverty line is then nothing else than a consumption threshold. In practice, income and consumption will be measured in currency units, which allow direct comparison of poverty lines across countries. Of course, before doing this comparison, it is necessary to translate them in the same currency on a scale of purchase power parity.
Martin Ravallion has performed that exercise in his paper Poverty Lines across the World. The first observation he makes is that poverty lines vary a lot across countries. He plotted then the poverty lines in function of countries’ average consumption and obtained the graph reproduced below. The following pattern emerges from this graph: the fifth poorest countries share a common poverty line around 1,25$ a day, then the thresholds increase with mean income. According to Ravallion, these empirical differences can be interpreted in two ways. First they could be the result of the variety of the underlying welfare norms across countries: each nation could have a different welfare level it considers as minimal. The second interpretation would be that if countries share the same welfare norm (same welfare threshold), relative income matters for welfare. His paper argues in favor of the second point of view.
The second explanation is supported by several empirical studies that showed people care for their relative income. Another justification given in the paper is based on Sen’s capability approach: what really matters for (the welfare of) people is to obtain several crucial functioning’s. Those functioning’s are doing’s and being’s an individual achieves in her life, like satisfying her basic human needs or being socially included. If the cost of the human survival does not evolve with mean income of society, the cost of participating to social activities typically will increase with it. Organizing or taking part to major celebrations like weddings is certainly more expensive in absolute terms in Europe than in sub-Saharan countries. Hence, reaching the same welfare is more costly in richer countries, therefore the higher poverty line.
As noticed in the paper, many among the poorest countries compute their poverty lines only taking into account the satisfaction of basic human needs, approximated by food intake. This explains the rather constant poverty thresholds for low average incomes. It does not mean however that deprived people in those countries do not care for social inclusion; it is just the way in which those countries have defined their poverty lines. Richer countries have recognized the relative character of poverty as is attested by the increasing slope of the graph. Following this reasoning, Ravallion argues that the variety of poverty lines across countries can be understood in a welfare consistent way. A global measure of poverty is hence possible. He calibrates therefore a relationship on the empirical data; this relationship gives the poverty line for the global measure in function of mean income. Based on this relationship, Ravallion defines global absolute poverty as the situation of individuals earning less than 1,25$ a day. People are then in relative poverty if they earn less than the poverty line corresponding to the average income of their countries.
If we accept the global definition of poverty proposed by Ravallion, what kind of policies does a country needs to implement in order to reduce its poverty? If this country faces absolute poverty problems, policies aiming at promoting economic growth are well-suited tools. If relative poverty is the main issue, those policies alone may not be sufficient and should ideally be combined with others promoting redistribution among the members of that country.
 Director of research department at the Word Bank.
 M. Ravallion, Poverty Lines across the World, Policy Research Working Paper 5284, the World Bank, April 2010.
 at 2005 PPP
 Even if, strictly speaking, mean consumption and mean income are not the same things, we use them as if they were interchangeable, as a measure of average welfare in one country.
 Luttmer E., “Neighbors as negatives: Relative Earnings and Well-Being”, QJE, 2005.