Poverty estimates are usually constructed from household survey data. The head of a household is typically asked about income and consumption levels, and these are used as the measure of well-being (World Bank 2001:17). Most governments have established national "poverty lines" by compiling and pricing a basket of goods meant to reflect the basic human necessities, such as food, clothing, and housing. Many countries have a "food" or "absolute" poverty line calculated from a food basket representing minimum nutritional requirements, and a "basic needs" line that is slightly higher (Deaton 2004:3-4; Coudouel et al. 2002:34).

In 1990, the World Bank began using the measure of $1 per day as an official "international poverty line," meant to roughly approximate the poverty lines of low-income countries (Ravallion et al. 1991; World Bank 1990:27). This measure remains controversial, but has provided a starting point for international comparison and for important poverty initiatives, including the United Nations' Millennium Development Goals.

The World Bank's most recent estimate is that some 1.1 billion people lived below the $1 per day line in 2001. About 46 percent of the population of Sub-Saharan Africa and 31 percent of South Asians live on less than a dollar a day (Chen and Ravallion 2004:1, 30). These numbers have not been static; the distribution of world poverty has changed significantly over the last quarter-century, due in large part to a dramatic drop in the number of poor people in East Asia . Chen and Ravallion broadly estimate that between 1981 and 2001, the number of people living below $1 per day in China declined by over 400 million, while in the rest of the world, the number rose from 850 to 880 million. The number of poor in Sub-Saharan Africa almost doubled over this period (Chen and Ravallion 2004:17, 20). In addition, many more people around the world live only slightly above the $1 per day line, suffering many of the symptoms of $1 per day poverty. Some 2.7 billion—almost half the world population—live on less than $2 per day (Chen and Ravallion 2004:16).

As useful as these aggregate numbers are, they tend to mask some important elements of the poverty landscape. For example, not all the poor fall into a single category—some are poorer than others. The depth and distribution of material poverty in different countries can be extremely varied. Weighing how far below the poverty line households fall—their "poverty gap," or gap between household income and the national poverty line—offers a useful measure of the depth of a nation's poverty (World Bank 2001:320).

Another variation on the standard poverty line looks at "relative poverty" by assessing the proportion of a country's population that lives at less than one-third the national consumption average. When this measure is applied, the poverty numbers for Sub-Saharan Africa and South Asia stay relatively similar to those calculated using national poverty lines. But the numbers in other regions soar, rising to 51 percent in Latin America and the Caribbean , and 26 percent in Europe and Central Asia (Hulme et al. 2001:18).

Still another way to measure poverty is to assess whether a household's total assets—cash, property, livestock, transport, and other possessions—fall below a critical level (Barrett and Swallow 2003:9). This approach is consistent with the perceptions of the poor themselves. When poor people are asked about their material concerns, they tend to focus not just on income, but on their lack of assets in general and the insecurity this brings (Narayan et al. 2000b:49).

Because poverty has so many dimensions, monetary measures are not the only, nor necessarily the best, way to count the poor. For example, the conventional household survey approach does not reveal disparities within households, and hence has no way of measuring income or consumption poverty among women, who often hold lower status. Education and health statistics, on the other hand, can be used to get a better perspective on many aspects of poverty, including those that are gender-related (World Bank 2001:27). Life expectancy, child mortality, the incidence of child stunting, literacy rates, and school enrollment are some of the more commonly used nonmonetary indicators. In an effort to address some of the gaps left by money-based assessments, analysts have developed a number of indices that measure multiple dimensions of poverty.The best known is the UN Development Programme's Human Development Index (HDI), a weighted index that includes education, life expectancy, and per-capita GDP (UNDP 2004:139).

For more information, see Data Table 4, "Income Distribution and Poverty."