MPI 2022 FAQs

The Global Multidimensional Poverty Index (MPI) identifies multiple deprivations at the household and individual level in health, education and standard of living, the same three dimensions as the Human Development Index. 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 multidimensionally poor or non-poor depending on the weighted number of deprivations his or her household, and thus, he or she experiences. These data are then aggregated into the national measure of multidimensional poverty. The MPI reflects both the incidence of multidimensional deprivation (a headcount of those in multidimensional poverty) and its intensity (the average deprivation score experienced by poor people). It can be used to create a comprehensive picture of people living in multidimensional poverty, and permits comparisons both across countries and regions, and within countries by ethnic group, urban or rural area, age as well as other key household and community characteristics. The MPI offers a valuable complement to income-based poverty measures. For more details see Technical note.

The 2022 release of the MPI presents estimates for 111 developing countries with a combined population of 6.1 billion (77.5% of the world total). About 1.2 billion people in the countries covered—19.1% of their entire population—live in multidimensional poverty. We could not include other developing countries due to data constraints as comparable data on each of the indicators were not available. There was also a decision not to use data from surveys conducted earlier than 2010.

Using the United Nations Development Programme (UNDP) classification for developing regions, the MPI 2022 covers 10 (out of 20) countries in the Arab States region; 15 (out of 26) countries in East Asia and the Pacific; 13 (out of 17) countries in Europe and Central Asia; 23 (out of 33) countries in Latin America and the Caribbean; 8 (out of 9) countries in South Asia; and 42 (out of 46) countries in Sub-Saharan Africa.

The MPI relies on two main databases that are publicly available and comparable for most developing countries: the ICF Macro Demographic and Health Survey (DHS) and the UNICEF’s Multiple Indicators Cluster Survey (MICS). For several countries, national household surveys with the same or similar content and questionnaires are used: Botswana, 2015/2016 Multi-Topic Household Survey; Bolivia, 2016 Encuesta de Demografía y Salud; Brazil, 2015 Pesquisa Nacional por Amostra de Domicílios (PNAD); China, 2014 China Family Panel Studies; Ecuador, 2018 Encuesta Nacional de Salud y Nutrición (ENSANUT); Jamaica, 2018 Jamaica Survey of Living Conditions (JSLC); Libya, 2014 Pan Arab Population and Family Health Survey (PAPFAM); Mexico, 2020 Encuesta Nacional de Salud y Nutricion (ENSANUT); Morocco, 2017/2018 Pan Arab Population and Family Health Survey (PAPFAM); Peru, 2019 Encuesta Demográfica y de Salud Familiar; Seychelles, 2019 Quarterly Labor Force Survey, and Sri Lanka, 2016 Demographic and Health Survey.

Table 1 of the 2022 release of the MPI indicates for each country if data come from the DHS, MICS or from a national survey.

Since the 2014 Human Development Report (HDR), the statistical programmes used to calculate the MPI have been available at the Human Development Report Office’s (HDRO) website. The statistical programs associated with the MPI 2022 are available online for a large selection of countries (http://hdr.undp.org/en/content/mpi-statistical-programmes).
In 2020 HDRO released, for the first time, programs that calculate the MPI in R, a free software available at https://www.r-project.org/. For now, these programs are available for 4 selected countries (Benin, Republic of Congo, India and Iraq). This is still an experimental phase, and we plan to expand the availability of MPI programs in R; in the meantime, we welcome feedback from the users.

Also, DHS and MICS data are publicly available online. Therefore, national governments, civil society groups, and research communities can replicate the MPI results as well as adapt the programmes to their own country-specific poverty needs.

The MPI was first developed in 2010 by OPHI and UNDP for the UNDP’s flagship Human Development Reports. In February 2012 and March 2013, a critical review of the family of human development indices including the MPI was conducted during the two conferences on measuring human progress organized by UNDP. As an outcome of these critical reviews, several adjustments were made to the MPI. They were justified on the grounds of being more in line with the MDGs. At that time, OPHI continued publishing their own estimates using the original 2010 specifications. In 2018, OPHI and UNDP revised and unified the MPI methodology, and the joint work is available at both OPHI’s and UNDP’s websites. The MPI estimates and analysis are expected to be updated at least once per year to include newly released microdata.

The new round of adjustments to the MPI introduced in 2018 were justified on the grounds of being more in line with the 2030 Agenda. Since 2018 the MPI is therefore a contribution to the implementation and monitoring of Sustainable Development Goal 1 which aims to end poverty in all its forms everywhere, and to the achievement of the Agenda’s ambition and fundamental principle of “Leaving No One Behind”.

The difference between the previous approaches stands in the definition of deprivations for some of the indicators.

• Health dimension:
  - Nutrition: a household is deprived if there is a stunted or an underweight child (instead of only stunted). Because, if a child is stunted, the damage is mostly irreversible. As Anthony Lake of UNICEF described it: “That child will never learn, nor earn, as much as he or she could have if properly nourished in early life.” https://reliefweb.int/report/world/speech-anthony-lake-unicef-executive-director-high-level-meeting-nutrition-new-york-20. Adults above 20 years are considered malnourished if their BMI is below 18.5 kg/m2; for individuals aged 15-19 years age-specific BMI z scores, defined by WHO, are calculated.

- Child mortality: a household is deprived if any child has died in the family in the five-year period preceding the survey. This captures recent improvements in child mortality. When the survey lacks information about the date of the child death, mortality that occurred anytime is used.

In 2019 a methodological adjustment was applied to child mortality. Following the Convention on the Rights of the Child, a child is defined as a person below the age of 18. With this new adjustment, the death of an offspring reported by a mother does not count as a deprivation if the death occurred when the son or daughter was above the age of 18. Previously, there was no age limit and child was understood as a son or daughter of any age.

• Education dimension:
  - School attainment: a household is deprived if no member aged 10 years or older has completed at least six years of schooling. Six years is the duration of primary education in most countries, so this change reinforces SDG 4 “Quality education.”

In 2020 we adjusted the years of schooling indicator by introducing a country-specific age cutoff: No household member of ages 'school entrance age + six' or older has completed six years of schooling.

Previously, the age cutoff was 10 years, but this did not recognize the fact that by age 10 children do not normally complete 6 years of schooling. For example, if the official school entrance age is 6, a child would normally complete 6 years of schooling by age 12; therefore, children of ages 10 or 11 should be considered ineligible for years of schooling. If, as an exception, children of ages 10 or 11 completed 6 years of schooling, this is counted as an achievement. Note that this adjustment has been implemented since 2020 but it was not applied to previous estimates. Although this adjustment is conceptually better, the effect on the empirical estimates is minimal.

- School attendance: a household is deprived if any school-aged child is not attending school up to the age at which he/she would complete grade 8.

• Standard of living:
  - Housing: a household is deprived if it has inadequate housing in at least one of the three: the floor is of natural materials or the roof or the walls are of rudimentary materials.

- Asset ownership: a household is deprived if it does not own more than one of these assets: radio, TV, telephone, computer, animal cart, bicycle, motorbike, or refrigerator, and does not own a car or truck.

- Electricity, improved sanitation, improved drinking water and cooking fuel remain the same.

Further details about the original MPI methodology from 2010 can be found in Alkire and Santos (2010) https://ophi.org.uk/acute-multidimensional-poverty-a-new-index-for-developing-countries/

Further details about the UNDP/HDRO revised methodology from 2014 can be found in Kovacevic and Calderon (2014) http://hdr.undp.org/en/content/undp’s-multidimensional-poverty-index-methodology-paper-2014

Further details about the jointly revised methodology can be found in: Alkire and Jahan (2018) http://hdr.undp.org/en/content/new-global-mpi-2018-aligning-sustainable-development-goals

No. The methodology used for the calculation of the 2022 MPI is identical to the one used for the 2020 and 2021 MPI.

The indicators included in the 2010 MPI were drawn from the Millennium Development Goals (MDGs). These original MPI indicators related 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).

With the subsequent adjustments of the methodology, the 2018 revision of the MPI is a contribution to the implementation and monitoring of Sustainable Development Goal 1 which aims to end poverty in all its forms everywhere, and to the achievement of the Agenda’s ambition and fundamental principle of “Leaving No One Behind”. In addition, national measures of Multidimensional Poverty may directly serve the purpose of monitoring SDG Indicator 1.2.2 (proportion of men, women and children of all ages living in poverty in all its dimensions according to national definitions).

The global MPI shows the overlapping deprivations of people sharing the same household across ten indicators that relate to SDGs 1, 2, 3, 4, 6, 7, and 11. In doing so, it provides a policy tool that can inform policies that seek to address interlinked SDGs and their targets.

Yes. We are continuously working on improving the MPI methodology to better measure the multidimensional aspect of human’s deprivations, to reflect the needs of policy makers and to incorporate new indicators or components when surveys update the information collected.

The MPI constitutes a set of poverty measures. These measures are explained as follows. Headcount or incidence of multidimensional poverty: the proportion of people who are multidimensionally poor according to the MPI (those who are deprived in at least one third of the weighted indicators). Intensity of multidimensional poverty: the average deprivation score experienced by people in multidimensional poverty. The MPI value summarizes information on multiple deprivations into a single number. It is calculated as the product of the incidence of multidimensional poverty and the intensity. These measures can be unpacked to show the composition of multidimensional poverty both across countries and regions, and within countries by ethnic group, urban/rural area, age as well as other key household and community characteristics.

One deprivation alone may not represent poverty. The MPI requires a person to be deprived in multiple indicators at the same time. A household, and so all its members, is multidimensionally poor if it is deprived in one third or more of the weighted indicators. We also count persons who are vulnerable to multidimensional poverty as those 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 living in severe multidimensional poverty.

The number of people living in multidimensional poverty is computed using the United Nations Department of Economic and Social Affairs/Population Division’s total population data for the year 2020 for all countries. This approach assumes a constant headcount ratio (obtained by applying the MPI methodology to the survey) which implies an assumption of no change in multidimensional poverty rates between the year of the survey and 2020.

In addition, every year the release of the MPI provides the number of people living in multidimensional poverty during the year of the survey. In countries with surveys conducted in a two-year period, i.e., 2015/2016, the population of the second year is used.

Total population data was obtained from World Population Prospects: The 2022 Revision. New York. https://esa.un.org/unpd/wpp/. Accessed August 7, 2022.

The 2022 MPI relies on the most recent and reliable data available since 2010. 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 the MPI value.

Data for the 2022 MPI are from 2010–2021, though 4 billion of the 6.1 billion people covered and 945 million people of the 1.2 billion identified as multidimensionally poor (81.3 percent) are captured by surveys from 2016 or later (83 of the 111 countries covered and 35 of those countries have data fielded in 2019 or later).

The MPI methodology shows aspects in which multidimensionally poor people are deprived and helps revealing 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 deprivations.

The multidimensional poverty approach can be adapted using indicators and weights that are more relevant to national context at the country level to create tailored national poverty measures. The MPI can be useful as a guide to help governments tailor a poverty measure that reflects local indicators and data. In 2009, Mexico became the first country to adopt a measure of multidimensional poverty as an official national statistic.

The MPI methodology can be, and often is, modified to generate national measures of multidimensional poverty that reflect local, cultural, economic, climatic and other factors. Such national measures of multidimensional poverty may directly serve the purpose of monitoring SDG Indicator 1.2.2 (proportion of men, women and children of all ages living in poverty in all its dimensions according to national definitions). The MPI was devised as an analytical tool to measure and monitor acute multidimensional poverty across nations.

The MPI has some drawbacks, due mainly to data constraints.

• The indicators may not reflect capabilities but instead reflect outputs (such as years of schooling) or inputs (such as cooking fuel).
• The health dimension indicators are not consistently collected across countries and overlook some groups’ deprivations, especially for nutrition, though the patterns that emerge are plausible and familiar.
• Child deaths that occurred at any time (instead of in the last 5 years previous to the survey) are used in 10 countries because the survey did not collect the date of child deaths. We are currently working on finding a better alternative to these cases.
• In some situations, careful judgments were needed to address data that was not collected. For example, data on nutrition was not collected for children older than 5 years in both DHS and MICS, data on adult nutrition was not collected in MICS, data on adult nutrition was collected only for women of ages 15 and 49 in DHS (though in a small number of countries it was collected for men), and child mortality was not collected in households without women of reproductive age. In these situations, households without eligible members to be measured for nutrition and without eligible members to report child mortality are assumed to be not deprived in these indicators. This is a strong assumption that may bias the results and we are conducting research to find a better alternative. However, 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, or in two health or education indicators. This requirement makes the MPI less sensitive to such assumption.
• Intra-household inequalities may be severe, but these could not be captured.
• While the MPI goes well beyond a headcount ratio to include the intensity of poverty, it does not measure inequality among the poor, although a separate measure of inequality has been published since 2019 (variance of individual deprivation scores of poor people). Also, decompositions by groups can be used to reveal group-based inequalities.
• The estimates presented are mostly based on publicly available data and cover various years between 2010 and 2021, which limits direct cross-country comparability.

The MPI can be used to study changes in multidimensional poverty patterns over time providing that the data at different points in time and MPI estimates are strictly harmonized. If estimates are not harmonized, readers are advised to carefully interpret the changes over time for a particular country because different indicators could be missing from the surveys.

This year, we are releasing changes over time estimates for 81 countries with suitable surveys of which 35 countries have data for three time points and one country, Gambia, has data for four. These estimates are presented in Table 2 of the 2022 MPI publication. This table presents harmonized estimates, and it allows for a strict comparison across time within a country. The harmonization process aligns the indicator definitions, meaning that, where necessary, this process re-creates the indicators in the global MPI so that they are using the same information and deprivation cutoffs in both years. For instance, if a survey in one year collects nutrition information from children under 5 years and for women aged 15–49 years and in the other year it collects nutrition information from children under 5 years only, the harmonized computations will only use information from children under 5 in both years, the common denominator, so that any change in the nutrition indicator is due to a change in the nutritional status of children rather than the inclusion of a group of women who may tend to be more (or less) undernourished.

Users are advised to distinguish between the unharmonized estimates in table 1 and the harmonized estimates in table 2. The first ones represent the best possible MPI estimate for a particular year and survey and include all the information that is available, while the second ones allow for a strict comparison across time (within a country) by removing the components that were not collected in the previous point in time. Unharmonized and harmonized estimates may be identical in some cases, but they may differ in many others.

Because the standard survey data used to estimate the global MPI measure are collected in differing intervals, the ability to detect changes depends upon the frequency of available data. The MPI will reflect the impacts of shocks as far as these cause children to leave primary education or to become undernourished, or lose adequate housing, for example. If more frequent data are available at the country level, this can be used to seek to capture the effects of larger scale economic and other shocks.

The MPI will be a valuable tool to measure the effects of COVID-19 and how the crisis has affected many of the indicators such as nutrition, school attendance, years of schooling and housing. However, the surveys used for the calculation of the 2022 MPI are mostly pre-COVID-19 pandemic. Future surveys will help reveling the impact of the pandemic on multidimensional poverty.

Yes. The global MPI estimates are constrained by need for comparability. National teams should use the indicators and weights that are more relevant to the country context. At the country level, the multidimensional poverty approach to assessing household deprivations can be tailored using country-specific data and indicators to provide a richer and more representative picture of poverty.

Incomplete data on the health dimension were collected in the Brazil’s Pesquisa Nacional por Amostra de Domicílios (PNAD). While PNAD collects rich data on social, economic and demographic characteristics of households and their members, some data that are needed for the MPI computation, such as basic nutrition characteristics (body weight and height of children and adults) are not collected. Although the survey collects household information on deaths of children (of any age), it does not collect the date of the death, so it is not possible to distinguish recent deaths from those that happened more than 5 years prior to the survey. Thus, the child mortality indicator has incomplete information on deaths of children. Given the lack on nutrition, this implies that the imperfect child mortality indicator would receive the full dimension weight of 1/3.

In such a case, the MPI methodology was modified to compensate for this limitation in the health dimension and two adjustments were performed: (i) only child mortality reported by mothers of ages between 15 and 49 was considered, and (ii) a household that is deprived only in child mortality cannot be declared multidimensionally poor. The first adjustment increases the likelihood that the reported deaths happened more recently and that they refer to a child under the age of 18 year, while the second removes the possibility that a household is declared multidimensionally poor based only on the death of a child that happened any time before the survey. This second rule means that an additional deprivation in at least one of the remaining indicators is needed for such a household to be declared multidimensionally poor.

Brazil is the only country that lacks information about nutrition and that lacks information about date of death for the child mortality indicator. There are 7 countries that lack data on deprivations in nutrition but with complete data on child mortality. Likewise, there are another 9 countries without date of death, but with data on deprivations in nutrition, therefore, the weight attached to child mortality is only 1/6 making the MPI estimates less dependent on the treatment of child mortality.

The MPI reflects the multiple deprivations that people experience at the same time. We have described the MPI as a measure of “acute” poverty 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 $1.90 (in 2011 $PPP) a day. The World Bank recently published poverty estimates using an updated poverty line of $2.15 a day based on 2017 PPP dollars (World Bank. 2022. “Poverty and Inequality Platform.” https://pip.worldbank.org /home).

The MPI complements income poverty measures. It measures various deprivations directly. In practice, although there is a clear overall relationship between the MPI and the $1.90 (or $2.15) per day measure, the estimates do differ for many countries. This is a topic for further research, but some factors can include the provision of public services, as well as different abilities to convert income into positive outcomes such as good nutrition.

Including income as an indicator might imply double counting people’s deprivations. The standard of living dimension of the MPI acts as a proxy for economic wellbeing. The MPI, which also includes other dimensions of wellbeing, should be seen as a complementary measure of income poverty that goes beyond the monetary aspect of people’s lives.

An interesting analysis would be to explore the overlap between income poverty and multidimensional poverty, i.e., to distinguish those who are income poor and multidimensionally poor versus those who are poor according to one concept only or those who are not poor at all. However, income data come from different surveys, and these surveys often do not have information on the health dimension. For most countries, we are not able to identify whether the same people are income poor and also deprived in the MPI indicators.

We could not include empowerment due to data constraints. The Demographic and Health Surveys (DHS) 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.