Poverty and Inequality Platform Methodology Handbook
This site describes how global and regional poverty rates are calculated. This particular page summarizes the methodology in non-technical terms using five steps. The summary is followed by five chapters that dig into these steps in detail.
The five steps are illustrated in the figure below and summarized after the figure. First, household survey data are obtained from relevant sources. Second, the survey data are used to create an estimate of households’ income or consumption, known as welfare aggregates. Third, the welfare aggregates are adjusted for differences in price levels across countries and over time to foster international comparability. Fourth, poverty and inequality are estimated for a particular country for a particular year. Fifth, the estimates of poverty are extrapolated or interpolated to a common year and the population-weighted poverty rate is calculated.
All decisions, assumptions, and protocols involved in these steps and in the chapters that follow are governed by the World Bank’s Global Poverty Working Group, which is composed of staff from the Poverty and Equity Global Practice, the Development Data Group, and the Development Research Group.
Step 1: Acquiring household survey data
Poverty rates are estimated from selected household surveys. In general, the surveys used ask a representative subset of households in a country about their consumption or income. These surveys are often the official surveys used by countries to monitor and report on poverty. Most household surveys from developing countries are obtained through collaboration with countries’ National Statistical Offices, while most data for high-income countries are obtained from the EU Statistics on Income and Living Conditions (EU-SILC) or from the Luxembourg Income Study (LIS) Database. At times, household-level data cannot be obtained in which case, as a second best, aggregated data are used, such as income or consumption by population deciles.
Step 2: Constructing welfare aggregates
The Poverty and Inequality Platform (PIP) primarily utilizes a monetary measure of poverty. Monetary poverty is estimated from an aggregation of household’s income or from the monetary value of their consumption. We refer to such aggregates jointly as welfare aggregates. Welfare aggregates are harmonized across countries and over time to maximize comparability, but country-specific decisions on issues such as whether income or consumption is used, the design of the questionnaire of the household survey, what components that are included in the welfare aggregate, and whether price differences within a country are accounted for, imply that full comparability is not feasible.
Step 3: Converting welfare aggregates
Welfare aggregates are often expressed in local currencies in the prices prevailing at the time of the collection of the data. To compare the consumption of an Indian household in 2011 to the consumption of a Nigerian household in 2018, the welfare aggregates need to be expressed in the same prices. To this end, consumer price indices are used to express all welfare aggregates in local 2017 prices, and purchasing power parities are used to account for price differences between countries. Once all welfare aggregates are expressed in the same units, a common poverty line is needed to estimate poverty. The international poverty line used in PIP is constructed from the poverty lines used by the poorest countries of the world.
Armed with welfare distributions expressed in 2017 PPPs and an international poverty line, poverty and inequality can be calculated and compared across countries and over time. PIP contains a range of different monetary poverty measures, a multidimensional poverty measure, inequality measures, and other distributional statistics.
Most countries do not conduct household surveys every year. Yet, to estimate regional and global poverty for a particular year, one needs an estimate of poverty for every country for the year in question. When a poverty estimate is not available for a given year, the estimates are extrapolated or interpolated from other years. The extrapolations assume that everyone’s income or consumption grows in accordance with per capita growth rates from national accounts between the time of the survey and the year in question. For countries without any household data at all, it is assumed that their poverty rate is equal to the population-weighted average poverty rate in their region. To ensure the quality of the regional and global numbers, coverage rules are used to determine whether a particular reference year has sufficient nearby survey data for global and regional numbers to be presented.