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 When there is national data available, it may be out of date due to the methods used to collect it, or other issues such as difficulties with reporting via statistical systems.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.

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

• IDM provides data on dimensions of poverty that are generally not included in traditional household surveys, but that are particularly important for women and girls.
• Individual level data allows for disaggregation by multiple different variables individually, as well as looking at the intersectionality of these variables.

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.

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.

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.

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.

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.

Sustainable Development Goals

• Leave No One Behind: Data Disaggregation for SDGs, Open Data Watch.
• Practical Guidebook on Data Disaggregation for the Sustainable Development Goals, Asian Development Bank, 2021.
• Sustainable Development Goals Helpdesk: Data Portals

P20

• Factsheet: How the P20 are doing in your country, Development Initiatives, 2018.
• P20 Initiative: Baseline report, Development Initiatives, 2017.
• The P20 in Benin: From consultation to consensus, Development Initiatives, 2020. Report available in English and French.

The Multidimensional Poverty Index

• Disaggregated MPI data
• Discussion on the different methodologies used to create composite indices and their limitations, M. Ravallion.
• Global Multidimensional Poverty Index
• Oxford Poverty and Human Development Initiative (OPHI)
• UNDP analysis of the MPI

Individual Deprivation Measure

• Assessing Contributions of the Individual Deprivation Measure, Development Initiatives, 2020.
• Individual Deprivation Measure (IDM)

The Internet Poverty Index

• Factsheet: Poverty trends: global, regional and national, Development Initiatives, 2021.
• Internet Poverty Index methodology
• ITU price data collection rules
• World Data Lab

Global Nutrition Report

• Data for the Global Nutrition Report
• Evaluation NCD Risk Factor Collaboration
• Global Burden of Disease
• Global Nutrition Report
• Institute for Health Metrics
• United Nations Children's Fund (UNICEF)
• World Health Organization (WHO)

National Multidimensional Poverty Reports

• Fourth Integrated Household Survey 2016-2017 (Malawi), Microdata Library.
• The Global Multidimensional Poverty Peer Network (MPPN), Oxford Poverty & Human Development Initiative (OPHI).
• Handbook: How to Build a National Multidimensional Poverty Index (MPI), United Nations Development Programme (UNDP) and the Oxford Poverty & Human Development Initiative (OPHI), 2019.

Making Voices Heard and Count, An Unequal Pandemic

• 'An Unequal Pandemic’ report, Civil Society Collaborative on Inclusive COVID-19 Data, 2021.
• Leave No One Behind project, International Civil Society Centre.

Improved understanding of resilience in Somalia

• Towards an improved understanding of vulnerability and resilience in Somalia, Development Initiatives, 2019.
• From Vulnerability to Resilience: A framework for analysis and action to build community resilience, Katherine Pasteur, 2011.