Frequently Asked Questions - Multidimensional Poverty Index (MPI)
The MPI was introduced as a new experimental series in 2010, alongside the Inequality-adjusted Human Development Index and the Gender Inequality Index. With the most recent updates, it is now a permanent feature of the HDRs.
The Multidimensional Poverty Index (MPI) identifies multiple deprivations at the household and individual level in health, education and standard of living. It uses micro data from household surveys, and—unlike the Inequality-adjusted Human Development Index—all the indicators needed to construct the measure must come from the same survey. Each person in a given household is classified as poor or non-poor depending on the number of deprivations his or her household experiences. This data are then aggregated into the national measure of poverty. The MPI reflects both the prevalence of multidimensional deprivation, and its intensity—how many deprivations people experience at the same time. It can be used to create a comprehensive picture of people living in poverty, and permits comparisons both across countries, regions and the world and within countries by ethnic group, urban or rural location, as well as other key household and community characteristics. The MPI builds on recent advances in theory and data to present the first global measure of its kind, and offers a valuable complement to income-based poverty measures.
The 2014 Human Development Report (HDR) presents estimates for 91 countries with a combined population of 5.0 billion (75% of the world total). About 1.5 billion people in the countries covered—29% of their entire population—lived in multidimensional poverty between 2005 and 2012.
The MPI identifies overlapping deprivations at the household level across the same three dimensions as the Human Development Index (health, education and living standards) and shows the average number of poor people and deprivations with which poor households contend. For more details see Technical note 5.
The MPI replaced the HPI, which was published from 1997 to 2009. Pioneering in its day, the HPI used country averages to reflect aggregate deprivations in health, education, and standard of living. It could not identify specific individuals, households or larger groups of people as jointly deprived. The MPI addresses this shortcoming by capturing how many people experience overlapping deprivations (prevalence) and how many deprivations they face on average (intensity). The MPI can be broken down by indicator to show how the composition of multidimensional poverty changes for different regions, ethnic groups and so on—with useful implications for policy.
One deprivation alone may not represent poverty. The MPI requires a household to be deprived in multiple indicators at the same time. A person is multidimensionally poor if she/he is deprived in one third or more of the weighted indicators. We also count those who are near-poor by as persons who are deprived in one fifth or more but less than one third of the weighted indicators. Those who are deprived in one half or more are considered being in extreme multidimensional poverty.
We could not include income due to data constraints. Income poverty data come from different surveys, and these surveys often do not have information on health and nutrition. For most countries we are not able to identify whether the same people are income poor and also deprived in all the MPI indicators so could not include income.
We could not include empowerment due to data constraints. The Demographic and Health Surveys (DHS surveys) collect data on women’s empowerment for some countries, but not every DHS survey includes empowerment, and the other surveys do not have these data. Data on men’s empowerment or political freedom are missing.
The MPI relies on two main databases that are publicly available and comparable for most developing countries: the Demographic and Health Survey (DHS) and the Multiple Indicators Cluster Survey (MICS). For several countries, the national household surveys with the same or similar content and questionnaires are used - Argentina, 2005 Encuesta Nacional de Nutrición y Salud (ENNys); Brazil, 2012 Pesquisa Nacional por Amostra de Domicílios (PNAD); China, 2009 China’s Health and Nutrition Survey; Mexico, 2012 Encuesta Nacional de Salud y Nutricion (ENSANUT); State of Palestine, 2006/2007 Palestinian Family Health Survey (PAPFAM), and South Africa, 2012 National Income Dynamics Study (NIDS). Tables 6 and 6A indicate for each country if data come from the DHS, MICS or from a national survey.
We could not include other countries due to data constraints. Comparable data on each of the indicators were not available for other developing nations. There was also a deliberate effort not to use data from surveys conducted earlier than 2005.
The MPI relies on the most recent and reliable data available since 2005. The difference in dates limits direct cross-country comparisons, as circumstances may have improved, or deteriorated, in the intervening years. This is the reason why we do not rank countries based on MPI value.
The MPI complements income poverty measures. It measures various deprivations directly. In practice, although there is a clear overall relationship between MPI and $1.25 per day poverty, the estimates do differ for many countries. This is a topic for further research, but some factors can include provision of public services, as well as different abilities to convert income into outcomes such as good nutrition.
The MPI, like the $1.25 per day line, is a globally comparable measure of poverty. It measures acute multidimensional poverty, and only includes indicators that are available for many countries. National poverty measures are typically monetary measures, and thus capture something different. The fact that there are differences does not mean that the national poverty number, or the MPI headcount is wrong—these simply measure different conceptions of poverty. At the same time, just as national poverty measures, in contrast, are designed to reflect the domestic situation more accurately and often differ in very useful ways from the $1.25 measure, some countries may wish to build a national multidimensional poverty index that is tailored to their context, to complement this international MPI.
No. The MPI is intended to complement monetary measures of poverty, including $1.25 per day estimates. The relationship between these measures, as well as their policy implications and methodological improvement, are priorities for further research.
The MPI methodology shows aspects in which the poor are deprived and help to reveal the inter-connections among those deprivations. This enables policymakers to target resources and design policies more effectively. This is especially useful where the MPI reveals areas or groups characterized by severe deprivation.
The MPI reflects the severe deprivations that people face at the same time. We have described the MPI as a measure of “acute” because it reflects overlapping deprivation in basic needs and also to avoid confusion with the World Bank’s measure of “extreme” poverty that captures those living on less than $1.25 a day.
The critical review of the UNDP’s indices including the MPI was done during the two conferences on measuring human progress organized by the UNDP in February 2012 and March 2013. As an outcome of these critical reviews a certain number of adjustments of the MPI were made. They are justified on the grounds of being more in line with the MDGs. At the same time Oxford Poverty and Human Development Initiative (OPHI) has published their own estimates using the using the original specifications. So, we presented the MPI estimates and the headcounts obtained by HDRO and OPHI.
The difference is in specifications of deprivations in several indicators: School attainment – a household is deprived in school attainment if no member of the household has completed 6 years of education (previously it was 5). Six years is the duration of primary education in most countries, so this change reinforces MDG 2 “Universal primary education.” School attendance – we allow a child of school-entry age one year late enrolment to avoid coding as deprivation a mismatch between the birthdate and the school start date. Further, in the health dimension, for nutrition – a household is deprived if there is a stunted child (instead of underweight child). Because, if a child is stunted, the damage is mostly irreversible. “That child will never learn, nor earn, as much as he or she could have if properly nourished in early life.” Similarly, the child mortality ‘experienced’ in the household is considered as deprivation in health if it has occurred within 5 years before the survey. Previously, there was no limit. This change captures recent improvements in child mortality. We also added ownership of arable lands and livestock to better capture rural poverty.
The MPI constitutes a set of poverty measures. These measures are explained below: Prevalence of poverty (also the term incidence of poverty is used): the proportion of people who are poor according to the MPI (those who are deprived in at least one third of the weighted indicators). Average intensity of poverty: the average number of deprivations poor people experience at the same time. MPI value: The MPI value summarizes information on multiple deprivations into a single number. It is calculated by multiplying the incidence of poverty by the average intensity of poverty. These measures can be unpacked to show the composition of poverty both across countries, regions and the world and within countries by ethnic group, urban/rural location, as well as other key household and community characteristics.
The MPI indicators are drawn from the MDGs as far as the available internationally comparable data allow. The 10 indicators of the MPI are identical, or relate, to MDG indicators: nutrition (MDG 1), child mortality (MDG 4), access to drinking water (MDG 7), access to sanitation facility (MDG 7) and use of an improved source of cooking fuel (MDG 9). The overall MPI can be broken down into its constituent parts, revealing the overlapping needs of families and communities across a range of indicators which so often have been presented in isolation. This helps policymakers to see where challenges lie and what needs to be addressed.
The MPI has some drawbacks, due mainly to data constraints. First, the indicators include both outputs (such as years of schooling) and inputs (such as cooking fuel) as well as one stock indicator (child mortality). Second, the health data are relatively weak and overlook some groups’ deprivations especially for nutrition, though the patterns that emerge are plausible and familiar. Third, in some cases careful judgments were needed to address missing data. But to be considered multidimensionally poor, households must be deprived in at least six standard of living indicators or in three standard of living indicators and one health or education indicator. This requirement makes the MPI less sensitive to minor inaccuracies. Fourth, intra-household inequalities may be severe, but these could not be reflected. Fifth, while the MPI goes well beyond a headcount to include the intensity of poverty experienced, it does not measure inequality among the poor, although decompositions by group can be used to reveal group-based inequalities. Finally, the estimates presented here are based on publicly available data and cover various years between 2005 and 2012, which limits direct cross-country comparability.
Yes. The global MPI estimates are constrained by need for comparability. National teams should use the indicators and weights that make sense. At the country level, however, the multidimensional poverty approach to assessing deprivations at the household level can be tailored using country-specific data and indicators to provide a richer picture of poverty at the country level.
The multidimensional poverty approach can be adapted using indicators and weights that make sense at the country level to create tailored national poverty measures. The MPI can be useful as a guide to helping governments tailor a poverty measure that reflects local indicators and data. In 2009 Mexico, became the first country to adopt a multidimensional poverty measure reflecting multiple deprivations on the household level.
Yes. The MPI methodology can and should be modified to generate national Multidimensional Poverty Measures that reflect local cultural, economic, climatic and other factors. The international MPI was devised as an analytical tool to compare acute poverty across nations.
We estimated the MPI for two or more data points for 38 countries for which suitable data were available. It seems that the MPI can be used to study the changes in poverty pattern over time providing that the data were available from the same survey conducted at different years.
The effects of shocks are difficult to capture in any poverty measure. Because the standard survey data used to estimate the global measure are collected infrequently, the ability to detect changes is limited by the available data fed. The MPI will reflect the impacts of shocks as far as these cause children to leave primary education or to become malnourished, for example. If more frequent data are available at the country or local level, this can be used to seek to capture the effects of larger scale economic and other shocks.