Image by Ollivier Girard/CIFOR
  • Guide
  • 10 October 2022

Multidimensional poverty: Measures and frameworks to leave no one behind

Authors

Deborah Hardoon , Elena Suckling

Share section

About this guide

This guide seeks to make sense of the growing body of data and evidence around multidimensional poverty, in the context of the leave no one behind commitment. It aims to help those who are interested in using existing multidimensional poverty data and measurement approaches to identify who is left behind in their context.

Poverty is now well established as a multidimensional phenomenon where different deprivations – from health status to political power – interact with each other to make it difficult for people to live a full and decent life. There is now a range of indicators, indices and tools that measure poverty in a multidimensional way. Our briefing describes the concept of multidimensional poverty and its origins.

The leave no one behind commitment puts additional requirements on any data used to analyse poverty. Specifically, it is necessary to be able to disaggregate this data to identify individual and group-based inequalities in poverty outcomes, as well as considering whether people and their voices are included in a meaningful way throughout the data lifecycle.

Disaggregation means analysing the outcome data by particular parts, or variables, such as age or gender. It takes us beyond an understanding of how the average person is doing, or the country/place as a whole, and allows us to understand how different groups are faring, both in their own right and in comparison with other groups.

This guide does not cover all multidimensional poverty measures but provides a sample of ways that multidimensional poverty can be measured. It summarises the strengths and weaknesses of different approaches and presents the main factors to consider with respect to the quality of the underlying data. Critically for data to be useful for a leave no one behind approach, it also presents the scope for disaggregation.

  • Part 1 presents a sample of internationally comparable and publicly accessible measures and indices of multidimensional poverty. These are commonly used by international development actors to compare different contexts and establish common and cross-cutting approaches for measuring multidimensionality.
  • Part 2 presents a sample of data and approaches that facilitate more context-specific measures which recognise that the experience of poverty can be different for different people in different places. They can enable more inclusive and participatory approaches to measuring poverty that leave no one behind.
  • The Useful resources section collates data, approaches and processes related to the multidimensional poverty measures included in this guide, providing evidence and aiding further research.
  • The Appendix breaks down our evaluation of the indicators/indices, providing a top-line overview of the strengths, weaknesses and capacity for disaggregation of each.
Share section

Part 1: Internationally comparable measures

Examining the strengths and weaknesses of selected internationally comparable and publicly accessible measures and indices of multidimensional poverty

This section reviews measures that are developed at the international level, using standardised data collection tools for universally important dimensions of people’s lives. This generates measures that aim to be relevant to, and comparable for, all countries. They can provide a comprehensive global perspective on who is furthest behind in terms of multidimensional poverty but can limit the extent to which the evidence can be localised, with meaningful inclusion of people’s voices throughout the data life cycle.

1.1 The dashboard approach

Dashboards are used to present indicators for different dimensions of poverty alongside each other. This enables the user to identify the full range of indicators that are included in a multidimensional analysis, understand the outcomes of each dimension individually and analyse the relationship between different dimensions.

  • Overview

    Perhaps the broadest and most universally relevant multidimensional poverty framework is articulated by Agenda 2030, which applies to all UN member states, no matter their income level.[1] The first Sustainable Development Goal (SDG) makes it explicit that poverty should be eradicated in “all its forms” – not just in an economic sense – and that this applies “everywhere” – not solely in low-income countries. Although there is a tendency for progress on this goal to be measured only by the international extreme poverty line, there are a range of indicators associated with this goal, which include access to services, rights to land and social protection coverage.

    Moreover, the 16 other goals in the SDG framework highlight the range of needs that must be addressed to achieve meaningful progress for humanity. These include access to decent work, a clean environment, good healthcare, education and gender equality. The dashboard is vast. Across the goals, there are 169 targets, measured by 232 separate quantitative and qualitative indicators, all of which require timely and accurate data for 193 UN member states.

    Dimensions of poverty measured

    1. No poverty
    2. Zero hunger
    3. Good health and well-being
    4. Quality education
    5. Gender equality
    6. Clean water and sanitation
    7. Affordable and clean energy
    8. Decent work and economic growth
    9. Industry, innovation and infrastructure
    10. Reduced inequalities
    11. Sustainable cities and communities
    12. Responsible consumption and production
    13. Climate action
    14. Life below water
    15. Life on land
    16. Peace, justice and strong institutions
    17. Partnerships for the goals

    Data provenance, governance and accessibility

    Where does the underlying data come from?

    National statistical offices are officially required to report data for these indicators. However, in 2019 only 20.5% of data series were filled with data sourced exclusively from countries. ‘Missing data’ gaps can be filled by estimates from custodian UN agencies. When there is national data available, it may be out of date due to the methods used to collect it, or other issues such as difficulties with reporting via statistical systems.

    Who compiles the data into indicators?

    Custodian agencies that lead on one or more indicators compile the data from all countries and are responsible for generating comparable regional and international statistics. Custodians use their own methodology for harmonising, adjusting and estimating data. The United Nations Statistics Division (UNSD) then uploads it into the global indicators database.

    Accessibility

    The data can be downloaded from national data portals, as well as from numerous dashboards and portals that collect regional/global data for a range of indicators.

    Data use: strengths and weaknesses

    Strengths

    • Extremely comprehensive range of indicators which measure many aspects of life and environment and what it takes to live well.
    • International comparability of data that follows international standards and definitions.
    • The breadth of multidimensional goals and indicators, along with the international profile of the SDGs, has given rise to parallel reporting mechanisms by civil society groups. These focus on the dimensions of poverty important to them, using alternative data collection approaches to triangulate and contextualise the findings.

    Weaknesses

    • Large number of equally weighted SDG indicators make it difficult to prioritise measures and interpret overall progress in reducing poverty.
    • International measures may not be appropriate to the local context, particularly when they are formulated and estimated by custodian agencies rather than being sourced from national data sources.
    • The nature of data points varies (e.g., sourced nationally or estimated), and a number of datasets have duplicate values, confusing users as to the correct value to be used and what underlying data this is based on.

    Data disaggregation

    It is a requirement under the SDG framework that the indicators can be disaggregated by different variables, as appropriate for each indicator. Practical guides are available on how to disaggregate the SDG data. In practice, there is limited disaggregation of the SDGs. Performance against the SDGs is presented on the global SDG database and communicated as a national average, rather than identifying the different outcomes for particular individuals or groups.

  • Overview

    The P20 initiative has been developed by Development Initiatives to better understand the data on people who are at risk of being left behind. The P20 is a simple concept that seeks to focus political attention on the poorest 20% of people at all levels – globally, nationally, in every sector and every community – so that action can be taken, and policies and resources can be better targeted to ensure that the poorest 20% of people are included in progress.

    Dimensions of poverty measured

    The P20 initiative uses three key proxy indicators to measure progress: income, nutrition and civil registration and vital statistics (CRVS). If data for these three indicators are not going in the right direction for the P20, it is unlikely that other SDGs will be achieved. The P20 approach can also be used to measure changes in other sectors and outcomes based on the priorities of governments and agencies, so that the P20 can be prioritised across all development efforts, and multiple dimensions of poverty.

    Data provenance, governance and quality

    Where does the underlying data come from?

    This approach can draw on any existing data that covers the whole population (to identify quintiles) and includes data for the bellwether indicators. For example, census data or nationally representative household survey sources such as the World Bank’s PovcalNet, USAID’s Demographic and Health Survey (DHS), and UNICEF’s Multiple Indicators Cluster Survey (MICS).

    Who compiles the data into indicators?

    Development Initiatives has developed and promoted this approach and analysed data at the global and national levels.

    Accessibility

    Development Initiatives published a global report, and identified the P20 country by country. More detailed analysis of the poorest 20% of people in Benin can be downloaded in English and French.

    Data use: strengths and weaknesses

    Strengths

    • The simplicity of an analysis by quintiles provides an entry point to identifying who is left behind and targeting efforts towards this population.
    • The method is flexible and adaptable; quintiles can be calculated based on any continuous indicator measuring a particular dimension of poverty (e.g., the 20% of the population with the lowest life expectancy).
    • Exploring the outcomes and characteristics of the P20 compared with the rest of the population can provide an important understanding of inequalities in a particular location.

    Weaknesses

    • Quintile analysis depends upon data that measures the whole population. As such, household surveys are the main data source for this analysis, which come with a range of limitations (see Box 1).
    • Quintile analysis is not designed to distinguish between people within each quintile, making it difficult to identify the furthest behind. Alternative thresholds and proportions for analysis may be more appropriate in different contexts.

    Data disaggregation

    The P20 approach proposes the use of five core variables to disaggregate the data, comparing the constitution of the P20 with the rest of the population with respect to wealth, gender, geography, age and disability. However, the flexible methodology means that any variable included in the underlying dataset can be used for disaggregated analysis.

  • 1.2 Composite indices

    Composite indices bring together indicators of different dimensions of poverty to generate a single summary statistic on who is living in multidimensional poverty. International indices use a standardised approach that can be used to compare multidimensional poverty between different countries. There are several different methodologies that can be used to combine the values from different indicators. Each methodology comes with its own implications of the choice of dimensions included and the weighting of these dimensions, which necessarily affects the findings for the scale and depth of multidimensional poverty.[2]

  • Overview

    The Multidimensional Poverty Index (MPI) measures 10 different non-economic indicators in over 100 low- and middle-income countries, covering 1.3 billion people. The MPI has been adopted by the United Nations Development Programme Human Development Report.

    In the global MPI, people are counted as multidimensionally poor if they live in a household that is deprived in one-third or more of 10 indicators. Each indicator is equally weighted within its dimension.

    Dimensions of poverty measured

    • Health
      • Nutrition
      • Child mortality
    • Education
      • Years of schooling
      • School attendance
    • Living standards
      • Cooking fuel
      • Sanitation
      • Drinking water
      • Electricity
      • Housing
      • Assets

    Visit the Oxford Poverty and Human Development Initiative website to see how the dimensions are weighted in this approach.

    Data provenance, governance and quality

    Where does the underlying data come from?

    This approach draws on existing microdata that covers the whole population and that includes data for the MPI indicators. For example, national census data, or nationally representative household survey sources, such as the World Bank’s PovcalNet, USAID’s Demographic and Health Survey (DHS), UNICEF’s Multiple Indicators Cluster Survey (MICS) and national survey sources.

    Who compiles the data into indicators?

    The methodology has been developed by Oxford Poverty and Human Development Initiative (OPHI). Since 2014, the methodology has been jointly developed with UNDP (United Nations Development Programme) to be more in line with Agenda 2030.

    Accessibility

    All MPI data is available to be downloaded from OPHI. Disaggregated data by region, ethnicity, race and class can also be downloaded for some countries. Analysis of the MPI is available as an official UNDP report. Calculations in R are now available for a selection of countries.

    Data use: strengths and weaknesses

    Strengths

    • Well-established index, with comprehensive country coverage and a growing time series to see change.
    • Summary measures for the scale and depth of multidimensional poverty make it easier to interpret and communicate progress over time.
    • The data and indicators follow international standards, allowing for comparisons between countries and a shared international understanding of a deprivations and capabilities approach to poverty alleviation.

    Weaknesses

    • The data draws on the most recent household survey data published since 2009. The differences in the timeliness of available data limits the extent to which cross-country comparisons can be made, on top of a range of other limitations discussed in Box 1.
    • Data is not available for every country for all MPI indicators, requiring proxy indicators to be used in some cases.
    • The standard MPI indicators may not be the most appropriate poverty measures for different contexts.

    Data disaggregation

    The MPI is primarily designed to provide estimates for multidimensional poverty at the national level, or lower geographic levels when the underlying data has sufficient sample sizes. MPI values can be disaggregated for different ethnic groups, urban/rural areas, region, age and other key household and community characteristics. However, as it is derived from household data it cannot be disaggregated by individual or within household characteristics, such as age and gender.

  • Overview

    The Individual Deprivation Measure (IDM) is a gender-sensitive, multidimensional measure of poverty. It uses 15 key dimensions of life – the dimensions which people with lived experience of poverty have deemed to be most important – to measure the poverty of individuals. The index is underpinned by a purpose-built survey tool that captures primary data on every adult within a selected household. This reveals important insights about differences that exist between members living in the same households. The index applies a gender-sensitive framework to identifying the dimensions of poverty that matter, such as unpaid work, time use, access to contraception and menstrual hygiene.

    The 15 dimensions of individual deprivation are composed of 23 themes and based on 34 individual indicators defined using responses to one or more survey questions. Combinations of responses to one or more survey questions are ranked on an interval scale from 0 (most deprived) to 3 (least deprived), reflecting individual experience of deprivation. An individual’s score for each theme is the mean average of all non-missing indicator scores within that theme, and the individual’s score for each dimension is the mean average of all non-missing theme scores within that dimension. The final IDM score for an individual is the weighted arithmetic mean of all dimension scores.

    Dimensions of poverty measured

    • Clothing
    • Education
    • Energy
    • Environment
    • Family planning
    • Food
    • Health
    • Relationships
    • Safety
    • Sanitation
    • Shelter
    • Time-use
    • Voice
    • Water
    • Work

    Data provenance, governance and quality

    Where does the underlying data come from?

    The Equality Insights survey tool was purpose-built for this index. It collects data at the individual level on 15 dimensions of poverty.

    Who compiles the data into indicators?

    Equality Insights, which is a flagship initiative of the International Women’s Development Agency (IWDA). Country reports are developed in partnership with local civil society organisations (CSOs).

    Accessibility

    Country reports available for six countries: Fiji, Indonesia, Myanmar, Nepal, Solomon Islands and South Africa. However, the underlying data is not publicly available.

    Data use: strengths and weaknesses

    Strengths

    Weaknesses

    • Currently data is available for a limited number of countries, which limits comparability.
    • The approach requires new primary data collection with the associated costs of collecting data in a timely and repeated way.
    • Sample sizes linked to cost and logistical challenges may be problematic for statistical power, which may be a particular challenge in small island development states for example, or for the granularity of disaggregated analysis desired.

    Data disaggregation

    This index has been explicitly developed to ensure that data can be disaggregated by individual characteristics by collecting data at the individual, rather than household, level. Information on the ethnicity, religion, age, gender of each individual is collected to highlight how these factors influence deprivation and inequality and interact to deepen disadvantage.

  • 1.3 Dimension-specific indicators

    A multidimensional approach can allow for a focus on a particular dimension, depending on the context, and for this to change and evolve over time. Internationally comparable and publicly accessible measures have been developed for some specific dimensions of poverty to enable comparisons between countries on deprivations of particular relevance to specific policy areas and sectors. These are measured by single indicators or a combination that measure different aspects of a single dimension.

  • Overview

    Access to digital and communications services and technologies, such as smartphones and the internet, are increasingly recognised not just as enablers to improve poverty outcomes, but as important dimensions of poverty in and of themselves. The internet is often required to access basic services such as banking. During the Covid-19 pandemic, for many the internet became a vital tool in accessing education, healthcare and work. Internet access can improve labour market outcomes as well as helping people stay entertained, informed and able to create connections with others.

    The World Data Lab defines internet poverty as being unable to afford a minimum package of connectivity. It is therefore determined by both a person’s total income and expenditure, as well as the cost of accessing a threshold of internet services where they live. By these definitions, the Internet Poverty Index (created by the World Data Lab) identified approximately 1.3 billion people living in internet poverty as of July 2022, around double the number of people living in extreme poverty.

    Dimensions of poverty measured by the internet poverty index

    A person is defined as Internet Poor if they cannot afford a minimum package of internet at a reasonable quality:

    • Affordability: maximum 10% of total personal expenditure
    • Quantity: 1GB monthly data – only mobile broadband (around 40 minutes/day)
    • Quality: 10MB per second download speed (allows use of most services).

    Data provenance, governance and quality

    Where does the underlying data come from?

    Internet access prices are based on data from the International Telecommunications Union (ITU), containing country-level information on the price of 1 or 1.5GB of mobile data.[3] These figures are combined with World Data Lab global spending data.

    Who compiles the data into indicators?

    World Data Lab, a data enterprise also responsible for the development of the world poverty clock.

    Accessibility

    The data from the World Data Lab is not open source, meaning a fee is required to access the original and raw data. It is therefore difficult to understand the foundational analysis for affordability and whether the data could be disaggregated any further, other than by gender and location.

    Data use: strengths and weaknesses

    Strengths

    • This index provides information about a dimension of poverty that is not generally included in existing multidimensional indices, or in household survey data.
    • This approach provides a definition of internet poverty that applies at the individual level in terms of affordability, as well as enabling comparisons between countries based on the current availability of infrastructure and support services.

    Weaknesses

    • The methodology uses data on people’s income and the market price of internet access to calculate what they are likely to be spending on the internet, rather than data on actual coverage and expenditure.
    • This provides a limited understanding of internet poverty. Useful internet access that enhances people’s quality of life depends not only on affordability, but also on ownership of technologies such as laptops and smartphones, education and digital skills, electricity, where someone lives and whether education or work is available online. These are not considered in internet access prices.[4]

    Data disaggregation

    The data is presented separately for men and women, but this relies on modelled data on the distribution of incomes. It is not clear whether further disaggregation is possible, as the underlying data is not publicly available.

  • Overview

    Nutrition is a critical dimension of poverty that goes beyond having enough to eat day to day. There are multiple aspects of nutrition that matter for health and wellbeing in the short and longer term. These can be measured as standalone indicators or combined to give an overall understanding of a person’s or country's nutritional status. The Global Nutrition Report provides a concise, data-focused update on the state of diets and nutrition across multiple dimensions of nutrition.

    Dimensions of poverty measured by the GNR

    The report sets out progress against internationally recognised global nutrition targets. This is then complemented by an evaluation of the impact of poor diets on our health and the planet, an assessment of the nutrition financing landscape, and a comprehensive overview of reporting on past Nutrition for Growth (N4G) commitments.[5] The 2021 report found that, despite some progress, unacceptable levels of malnutrition persist, diets are not getting healthier, and our food choices are making increasing demands on the environment.

    Global nutrition targets measured:

    • Childhood stunting
    • Childhood wasting
    • Anaemia
    • Low birth weight
    • Childhood overweight
    • Exclusive breastfeeding
    • Raised blood pressure, women and men
    • Sodium intake, women and men
    • Obesity, women and men
    • Diabetes, women and men

    Data provenance, governance and quality

    Where does the underlying data come from?

    Data comes from both publicly available and private sources. These include Global Burden of Disease, the Institute for Health Metrics, Evaluation NCD Risk Factor Collaboration, United Nations Children's Fund (UNICEF) and the World Health Organization (WHO).

    Who compiles the data into indicators?

    Numerous expert authors, including analysts from Development Initiatives.

    Accessibility

    All data for the for the Global Nutrition Report is available on the GNR website and Github.

    Data use: strengths and weaknesses

    Strengths

    • There is wide coverage of country data which allows for international comparisons over a range of nutritional issues.
    • The methodology and underlying data are available for each indicator, providing transparency on whether the data is modelled or projected and where it is sourced from.

    Weaknesses

    • For country-level progress by different nutritional indicators, there is a variation in data availability, with some countries having no data available.

    Data disaggregation

    Data included in the GNR includes national level measures, such as the environmental impacts of the food system. Due to their nature these cannot be disaggregated to the group or individual level. Where individual outcomes are measured, most indicators can be disaggregated by sex, age, education, location and wealth.

  • Share box

    Box 1

    Strengths and weaknesses of household survey data

    Many of the internationally comparable approaches to measuring multidimensional poverty, including the P20 and Multidimensional Poverty Index (MPI), use data from international household survey programmes. Their main sources are:

    • The Multiple Indicator Cluster Surveys (MICS), an international household survey programme developed by the United Nations Children’s Fund (UNICEF) in the 1990s. Over 300 surveys have been carried out in just under 100 countries.
    • The Demographic and Health Survey (DHS) is a nationally representative household survey developed by the United States Agency for International Development (USAID) in the 1980s. Over 300 surveys have been carried out in just under 100 countries.
    • The Living Standards Measurement Study (LSMS) is a household survey developed by the Development Research Group at the World Bank in the 1980s. Over 100 surveys have been carried out in just under 40 countries.

    Strengths

    • Comprehensive surveys that collect a wide variety of indicators across multiple dimensions of poverty.
    • Nationally representative with large sample sizes, which can allow for geographical disaggregation.
    • Repeated periodically using a standardised approach, which allows for comparisons over time and between countries.

    Weaknesses

    • Although designed to be nationally representative, some population groups are missed. This can include marginalised population groups such as people without a permanent address or migrant workers, as well as more privileged individuals who evade door-to-door surveyors.
    • As data is collected at household level, differences between individuals within that household cannot be identified. This limits the potential for disaggregation of outcomes by important variables such as gender and age.
    • Standardised questions and international management of the survey programme can mean that the dimensions of poverty included may not be the most pertinent to the local population – neglecting important dimensions of poverty such as voice and empowerment. The data itself may also not be accessible and understood by the population groups that it relates to, or the national and local governments and agencies that would use it to inform policies and services.
    • Available survey data is often several years out of date by the time it is published, due to the timeframe for conducting the surveys and analysing and publishing the data.

    Issues surrounding international household survey programmes are further discussed in Development Initiatives’ Key facts on household surveys factsheet.

    Share section

    Part 2: Context-specific multidimensional poverty measures and approaches

    A number of tools and approaches have been developed that seek to generate a multidimensional measure for poverty that is specific to the lived experience of people within a particular place. In contrast to the data and indices reviewed in Part 1, these are not designed to enable a comparative analysis between countries or provide a global perspective, but rather inform the policies and priorities of actors in a particular context and can include meaningful ways to include people's voices on the aspects of poverty that matter most to them.

    2.1 Localising global measures

    Guidance and resources have been developed for several of the main international multidimensional measures to be adapted to better suit the local context. Standardised methodological frameworks and indicators endorsed at the global level can be adjusted for different contexts including for example by removing, adding or adjusting the weighting of particular dimensions as appropriate.

  • Overview

    A national Multidimensional Poverty Index (MPI) is a country-specific adaptation of the Multidimensional Poverty Index. Such measures generally take the three core dimensions of health, education and living standards as their starting points, and supplement them with other dimensions measured by locally appropriate indicators.

    This localisation is supported by the Multidimensional Poverty Peer Network, creating a south-to-south dialogue and sharing of experiences and lessons learnt from measuring multidimensional poverty. The sharing of information helps the development of national MPI reports. OPHI has worked with the UN to develop a handbook, online training and other resources to enable the translation of the methodology to different contexts.

    Dimensions of poverty measured

    Structure of Malawi’s MPI report:

    • Health and population
      • Sanitation
      • Nutrition
      • Drinking water
      • Food security
    • Education
      • Literacy and schooling
      • School attendance
    • Environment
      • Electricity
      • Rubbish disposal
      • Housing
      • Asset ownership
    • Work
      • Unemployment
      • Job diversity
      • Child labour

    See Figure 1 (p.5) in Malawi’s MPI report for the weighting of these dimensions and indicators.

    Data provenance, governance and quality

    Where does the underlying data come from?

    The Malawi MPI is based on data from the Fourth Integrated Household Survey (IHS4) 2016/17. This report offers a baseline for findings on multidimensional poverty and will be revised using data from the IHS5 survey (2019/20) to show changes in the levels and composition of multidimensional poverty.

    Who compiles the data into indicators?

    This report was produced by the National Statistical Office (NSO) in collaboration with UNDP, the University of Malawi (Economics Department), the Centre for Social Research, the National Planning Commission and the Ministry of Economic Planning and Development and Public Sector Reforms, with technical support from the Oxford Poverty and Human Development Initiative.

    Accessibility

    The microdata used in the analysis from the household survey is available to download, including the headcount ratio and contribution of each indicator for each district. Other data points are less easily accessible.

    Data use: strengths and weaknesses

    Strengths

    • The choice of indicators included in the Malawi MPI is informed by national interests to directly inform policy priorities.
    • The index generates composite measures for the depth and prevalence of multidimensional poverty, which can be monitored over time.
    • The methodology is recognised internationally as a robust approach, with continuous learning and peer reviews strengthening its evolution over time.

    Weaknesses

    • The index uses household-level survey data, the disadvantages of which are discussed in Box 1. This also limits the variables available for analysis and the extent to which locally pertinent dimensions can be included.

    Data disaggregation

    Like the global MPI, the Malawi MPI is estimated at the national level. It can be disaggregated by region, district, place of residence and sex of household head, but due to the limitations of using household level data it is not possible to measure inequalities within households, such as gender.

  • 2.2 Local data collection

    A range of tools and approaches have been designed to ensure that, unlike in some standard household surveys, the most marginalised individuals are explicitly targeted for inclusion in the data collection process. Bespoke questionnaires can also be designed to include dimensions of poverty and variables for disaggregation that are not necessarily on the radar of international agencies, but are very important for local people.[6] Surveys can be administered by local groups, whilst using standardised methodologies to strengthen the quality and comparability of the data.

  • Overview

    Making Voices Heard and Count is a collaborative project of the Leave No One Behind Partnership, which itself is a coalition using community-driven data to fill knowledge gaps in local SDG monitoring, helping to better understand drivers of vulnerability and marginalisation. The ‘An Unequal Pandemic’ report is an outcome of the coalition and was produced by both the Leave No One Behind Partnership and the Inclusive Data Charter, with support from a steering group, including Development Initiatives. Through the use of non-official data, the report aimed to fill and highlight the unequal impacts of poverty and inequality for marginalised groups, which are often not highlighted when only official data is used. It found that the pandemic amplified inequalities relating to health care access, housing and sanitation, while official responses risked further discriminatory practices.

    Official data is categorised as data that is collected through a national statistical office, or through some form of government organisation. Non-official data is data that is collected through non-governmental departments, including the civil society organisations, research institutes and the private sector.

    The report drew on 38 reports that provided data and insights from 91 countries across Africa, the Americas, Asia, the Caribbean, Europe and Oceania. The CSOs in this collaborative effort represent and work with diverse groups, including ethnic minorities, Dalits,[7] indigenous peoples, internally displaced people (IDPs), LGBTQI+ individuals, migrants, older people, persons with disabilities, refugees, religious minorities, street-connected children, undocumented people, women and girls, and young people.

    Dimensions of poverty measured

    • Direct health risks from Covid-19
    • Access to health services
    • Economic impacts (income, employment, social protection)
    • Food insecurity
    • Education
    • Violence, abuse and discrimination.

    Data provenance, governance and quality

    Where does the underlying data come from?

    Online surveys during March and April 2021 which collated information from multiple CSOs to understand impacts of the pandemic. Subsequently, a desk review of resources containing further community data was analysed.

    Who compiles the data into indicators?

    A group of CSOs formed the Civil Society Collaborative on Inclusive COVID-19 Data, under the international Leave No One Behind Partnership.

    Accessibility

    The collated data is not available in one place, but the individual country reports where the data has been captured are referenced in the annex.

    Data use: strengths and weaknesses

    Strengths

    • The voices of marginalised communities and CSOs are used directly as a data source to measure outcomes.
    • Data on specific groups can be used to uncover community findings and trends that can be used at the local level to drive change.

    Weaknesses

    • The coverage of community and CSO data can vary widely. It is generally not representative of the whole population and cannot be used to compare directly between groups – surveyed and not surveyed – within a country.

    Data disaggregation

    Data has been collected in countries to specifically enable an understanding of the impacts of Covid-19 on particular marginalised groups, based on their migration status, caste, gender, sex, sexuality, disability, religion, gender and age. However, due to the non-representative nature of the data collection, it is not possible to directly compare between these groups and the national average using a single data source.

  • 2.3 Adaptive multidimensional poverty and resilience assessments

    Adaptive and flexible data collection tools that are directly informed by people and their lived experiences can provide a hyper-localised understanding of multidimensional poverty. Tools can collect qualitative and quantitative data on the nuance and specificity associated with a particular group or place that may not necessarily translate across to more generalised measures.

  • Overview

    Development Initiatives measured resilience[8] in Somalia by assessing shocks and stresses associated with responding to and surviving the 2011 famine and 2016 drought. The study sought to understand local community perspectives on vulnerability and how livelihood and coping strategies are changing as a result of changes in the frequency and severity of extreme weather events in Somalia.

    Following a literature review, primary (largely qualitative) data was collected in four locations. General open questions were used to elicit personal perspectives, but respondents were guided to consider potential inequalities. For example, questions asked how women/ female-headed households, or persons with disabilities and their families were affected by the 2011 and 2017 droughts. The data’s consistency was assessed before generating findings and implications for policy and practice. Three key drivers of marginalisation in Somalia were identified: ethnicity/kinship/clan identity, gender and livelihood/employment.

    Dimensions of resilience measured:

    • Exposure to hazards and stresses
    • Fragile livelihoods
    • Future uncertainty
    • Governance.

    These factors were chosen using the Vulnerability to Resilience Framework.

    Data provenance, governance and quality

    Where does the underlying data come from?

    Primary data was collected through participant observation, focus group discussions (FGDs), household dialogues and analyses of livelihood, wellbeing and gender. This information, from transcribed daily reports, was categorised by subject and topic based on predetermined research questions.

    Who compiles the data into indicators?

    Development Initiatives conducted the research.

    Accessibility

    Due to the largely qualitative nature of the information, raw data is not included in the report. A small amount of quantitative data (such as livestock ownership) was captured, but this is not publicly available.

    Data use: strengths and weaknesses

    Strengths

    • The data is able to demonstrate what resilience means specifically for the context of Somalia.
    • Local researchers were used to collect the data.

    Weaknesses

    • There were challenges with collecting quality data of this nature in a fragile context. Explaining technical terms in local languages and sensitivities among government officials, local militant group operatives and sympathisers caused difficulty for the research teams conducting investigations in communities.
    • The qualitative nature of the data may make it difficult to create comparisons over different time frames.

    Data disaggregation

    Development Initiatives was able to disaggregate the data and measure the degree of vulnerability to resilience, shock and stress experienced by each grouping based on numerous factors, including conflict, seasonal dynamics, urban or rural location and the predominant types of livelihoods. The study was to understand the changes in livelihoods for internally displaced people, women, children, youth and persons with disabilities, based on gender, age, location, migration status, and ethnicity, in relation to shocks and crises, and the resilience they had. The LGBTQI+ and Christian community were recognised, but due to contextual sensitivities and a short time frame, the research team were unable to speak or confirm with respondents from these communities.

  • Share section

    Useful resources

    This section collates data, approaches and processes related to the multidimensional poverty measures included in this guide, providing evidence and aiding further research.

    Data, approaches and processes supporting internationally comparable measures

    Data, approaches and processes supporting context-specific multidimensional poverty measures

  • National Multidimensional Poverty Reports

  • Making Voices Heard and Count, An Unequal Pandemic

  • Share section

    Appendix

    Table A1: Breakdown of multidimensional poverty indicators/indices

    Indicator/index Dimension of poverty being
    measured
    Data provenance Variables available for disaggregated
    analysis
    Who compiles the data into
    indicators?
    Accessibility Strengths Weaknesses
    SDGs 16 different dimensions, including poverty, work, sustainability National statistical offices Limited disaggregation National statistical offices are responsible for reporting the data to
    custodians of the specific indicator, who compile the data from numerous
    countries. UNSD then uploads it into the global database.
    The data can be downloaded from numerous dashboards and data portals. International comparability of data that strictly follows
    international standards. The summary measure makes it easier to interpret
    basic progress.
    Due to the large number of SDG indicators, it can be difficult to prioritise
    the measures and the equal weighting of the indicators means it can also be
    difficult to interpret the overall message. International measures may not be
    well suited to the local context.
    P20 Income, nutrition and Civil Registration and Vital Statistics (CRVS) Existing data sources such as the World Bank’s PovcalNet, USAID’s
    Demographic and Health Survey, and UNICEF’s Multiple Indicators Cluster
    Survey
    Wealth quintile, gender, geography, age and disability. Development Initiatives The data can be downloaded in English and French. The method is flexible and adaptable, focusing on a few
    indicators. It aims to
    provide a flexible approach to promote the importance of disaggregating data.
    The use of household data, see Box 1.
    The Multidimensional Poverty Index Health, education, living standards Using microdata, from
    censuses, household surveys and civil registration data (births and deaths),
    including the Demographic and Health Surveys (DHS), and Multiple Indicator Cluster
    Surveys (MICS).
    Ethnic group, urban/rural areas, region, age and other key household
    and community characteristics.
    The Oxford Poverty and Human Development Initiative (OPHI) All published MPI data is available to be downloaded. Disaggregated
    data can also be downloaded.
    International comparability of data that strictly follows
    international standards. The summary measure makes it easier to interpret
    basic progress.
    International measures may not be well suited to the local context.
    Limitations of using household survey data, as discussed further in Box 1.
    Equality Insights The 15 dimensions of individual deprivation are composed of 23 themes
    and based on 34 individual indicators that are defined using responses to one
    or more survey questions.
    The Equality Insights
    survey tool was purpose-built for this index. It collects data at the
    individual level on 15 dimensions of poverty.
    Numerous, including
    ethnicity, religion, age, gender
    Equality Insights Country reports, but not data, available for six countries: Fiji,
    Indonesia, Myanmar, Nepal, Solomon Islands, South Africa.
    It provides data on many indicators beyond what is currently offered
    from traditional household surveys and highlights how these dimensions
    overlap. The intersectional approach means deprivations that exacerbate each
    other (such as gender, age, disability) are able to be identified.
    Requires new primary data collection with the associated costs of
    collecting that data in a timely and repeated way. Data available for a
    limited number of countries, limiting comparability.
    Sample sizes linked to cost and logistical challenges may be
    problematic for statistical power, which may be a particular challenge in
    small island development states.
    The Internet Poverty Index Internet affordability, quantity and quality Internet access prices
    are based on data from the International
    Telecommunications Union (ITU), containing country-level information on
    the price of 1 or 1.5 gigabytes of mobile data. These numbers are
    combined with World Data Lab global spending data.
    Gender, country World Data Lab The data from the World Data Lab is not open source, meaning a fee is
    required to access the original and raw data. This means it is difficult to
    understand the foundational analysis for affordability and whether the data
    could be disaggregated any further, other than by gender and location.
    - Internet access also
    depends on ownership of technologies such as laptops, education and digital
    skills, electricity, or where someone lives, something the ITU doesn’t
    consider in its internet access prices. Factoring in these categories will
    then determines for example who participates in online education or work.*
    There is also a need for basic infrastructure when analysing internet
    poverty, such as electricity and mobiles, so in this particular context of
    internet poverty, it is more a measure of monetary poverty.
    Global Nutrition Report Nutrition Data comes from both
    publicly available and private sources. These include Global Burden of
    Disease, the Institute for Health Metrics, Evaluation NCD Risk Factor
    Collaboration, United Nations Children's Fund (UNICEF) and the World Health
    Organisation (WHO).
    Sex, age, education, location and wealth, alongside many nutritional
    indicators.
    Numerous expert authors, including analysts from Development
    Initiatives.
    All data for the Global Nutrition Report is available on the GNR
    website and Github.
    There is wide coverage of country data which allows for international
    comparisons. There are many different indicators of nutritional wellbeing are
    measured, which are easily able to be accessed and downloaded. The methodology
    is available for each indicator, providing clarity on whether the data is
    modelled or projected and where it is sourced from.
    For country-level
    progress by different nutritional indicators, there is a variation in
    data availability, with some countries having no data available.
    National Multidimensional Poverty Reports (Malawi) Health, education, living standards The Malawi MPI is based
    on data from the Fourth
    Integrated Household Survey (IHS4) 2016/17. This report offers a baseline
    for findings on multidimensional poverty and will be revised using data from
    the IHS5 survey (2019/20) to show changes in the levels and composition of
    multidimensional poverty.
    Like the global MPI, the Malawi MPI is estimated at the national
    level, and then it is disaggregated by region, district, place of residence,
    sex of household head and age groups to identify some of the poorest
    population subgroups, as well as specific locations with multiple overlapping
    deprivations.
    National Statistical
    Office (NSO) in collaboration with the United Nations Development Programme
    (UNDP), the University of Malawi (Economics Department), the Centre for
    Social Research, the National Planning Commission and the Ministry of Economic
    Planning and Development and Public Sector Reforms, with technical support
    from the Oxford Poverty and Human Development Initiative.
    The microdata used in the analysis from the household survey is
    available to be downloaded,
    but a large amount of analysis in the report isn’t. The headcount ratio and
    contribution of each indicator for each district is available.
    The use of context specific data analysis meant a better understanding
    of poverty is understood for the Malawi population. It can monitor the key
    and simultaneous inequalities that affect those living in poverty.
    Limitations of household data.
    Making Voices Heard and Count: An Unequal Pandemic The impacts of the pandemic on:
    Direct health risks from Covid-19
    Access to health services
    Economic impacts (income, employment, social protection)
    Food insecurity
    Education
    Violence, abuse and discrimination
    Online surveys during March and April 2021, collating information from
    multiple CSOs to understand impacts of the pandemic. Subsequently, a desk
    review of resources containing further community data was analysed.
    Migration status, caste, gender, sex, sexuality, disability, religion,
    gender and age.
    A group of CSOs formed the Civil Society Collaborative on Inclusive
    COVID-19 Data, who put the data for the report together.
    The collated data is not available in one place, but the individual
    reports where the data has been captured are referenced in the report’s annex.
    Uses insights from the community and CSO data to highlight who has
    been most marginalised by the impact of Covid-19 and associated policies and
    programmes. The data on specific groups can be used to uncover community
    findings and trends.
    Good coverage of numerous countries, with input from marginalised
    communities.
    The cover of community and CSO data can vary widely depending on the
    methods used and overall most community and CSO data is not comprehensive
    enough to be representative of the whole population.
    Resilience in Somalia Resilience to shocks, crises and stresses. Primary data collected through participant observation, focus group
    discussions (FGDs), household dialogues, livelihood analysis, wellbeing
    analysis and gender analysis.
    Internally displaced people, women, children, youth and persons with
    disabilities, based on gender, age, location, migration status, and
    ethnicity, in relation to shocks and crises, and the resilience they had.
    Development Initiatives The qualitative data is available in the report. Context-specific understanding of what resilience means specifically
    for different groups in Somalia.
    There was difficulty in explaining technical terms in local languages
    and sensitivities among government officials, and local militant group
    operatives and sympathisers found difficulty with the research teams
    conducting investigations in communities. This made the quality of the data
    collected variable in different study locations.
    The qualitative nature of the data may make it difficult to create
    comparisons over different time frames.

    * Access to Internet-capable or Internet-connected personal computational devices is not a sufficient precondition for continued Internet use. Rather, Internet Use Continuance is a function of broader socioeconomic factors among them socioeconomic status, communal influence, and government influence https://www.tandfonline.com/doi/abs/10.1080/02681102.2021.1962234?journalCode=titd20

    Download this report

    Download now

    Notes

    • 1
      Holy See and Palestine have also committed to supporting and achieving the SDGs.
      Return to source text
    • 5
      Nutrition for Growth brings together country governments, donors, NGOs, and beyond to adopt evidence-based nutritional policies and the global, regional and country levels. It also encourages pledging to increase financing for proven nutrition-specific and nutrition-sensitive interventions, and alignment of actions across sectors and stakeholders.
      Return to source text