▶️ Back to Poverty & Inequality CoP
How to innovate in the policy space?
This webinar launched “How to Use National MPIs as a Policy Tool: From Metrics to Policy,” a collaborative report by the United Nations Development Programme and the Oxford Poverty and Human Development Initiative.
This handbook offers innovative strategies for integrating the Multidimensional Poverty Index (MPI) into public policy, providing governments and practitioners with a roadmap to address poverty in all its dimensions.
We were joined by an esteemed panel of speakers to explore how to redefine the fight against poverty and building a more equitable world.
Recording
Learning objectives
- Gain insights into how the MPI can enhance poverty reduction strategies and SDG reporting.
- Learn from global case studies where MPIs have been successfully implemented.
- Explore practical tools for tailoring policies to address multidimensional poverty.
Speakers
- Sabina Alkire, Professor, University of Oxford
- Gonzalo Hernández Licona (author), Director, Multidimensional Poverty Peer Network (MPPN-OPHI)
- Nyendi Moloyi, Director of the Department of Enterprise Development, Ministry of Entrepreneurship, Botswana
- Angela Lusigi, UNDP Resident Representative in India
- George Gray Molina, UNDP Chief Economist and Director Inclusive Growth
- Yanchun Zhang, UNDP Human Development Report Office Chief Statistician
Moderator
- Francine Pickup, Deputy Director, Bureau for Policy and Programme Support, UNDP
Key takeaways
1. Using National MPIs as a Policy Tool
- The handbook provides a roadmap for policymakers to move from metrics to policy implementation.
- MPI should not just be seen as a measure of poverty, but as a strategic tool to shape evidence-based and targeted policymaking, ensuring that no one is left behind.
2. The Need for Real-Time, Dynamic Data
- Advances in technology allow for real-time data collection through digital tools, mobile apps, and geospatial analytics.
- MPI data should be complemented with dynamic and local-level insights to capture the evolving nature of poverty and improve policy responsiveness.
3. Political Opportunities Must Be Leveraged
- MPI adoption requires political champions who can engage decision-makers and create space for policy dialogue.
- Opportunities such as the upcoming World Social Summit in Qatar provide critical moments to position MPI in global discussions and advocate for its role in poverty reduction strategies.
4. Integration with Other Policy Areas is Essential
- The discussion emphasized the importance of embedding MPI into broader policy frameworks, linking it with:
- Fiscal policies to ensure sustainable financing.
- Social protection and economic policies to address overlapping deprivations.
- Climate change, digital transformation, and gender equality to ensure holistic development approaches.
- Cross-ministerial collaboration is key to making MPI a tool for multi-sectoral policy action, rather than just a statistical measure.
5. MPI as a Storytelling and Communication Tool
- Translating MPI metrics into compelling stories is critical for ensuring policy uptake.
- Using poverty maps, human-centered data visualization, and real-life case studies makes MPI more accessible to decision-makers and the public.
- This approach enhances public engagement, political will, and targeted policy action.
6. Flexibility and Future Adaptation of MPI
- MPI must remain adaptive to capture emerging forms of deprivation (e.g., digital exclusion, climate risks).
- Countries should continuously refine MPI indicators to ensure their relevance in rapidly changing economic and social landscapes.
7. Building Institutional Capacity and Political Will
- Sustained engagement is needed to build institutional knowledge and national ownership of MPI.
- Capacity-building efforts should empower local policymakers, statistical offices, and civil society to actively use MPI for decision-making.
- Countries must invest in training, technical expertise, and cross-sector dialogue to ensure MPI remains embedded in governance structures beyond political cycles.
8. Using MPI to Track Localized and Multi-Dimensional Deprivations
- Botswana and India have successfully used MPI to identify and track localized poverty hotspots.
- Botswana’s MPI work includes using real-time data from mobile apps and digital tools to monitor services in informal settlements.
- India’s MPI tracking allows for monitoring poverty trends over time, especially in response to economic shocks and climate events.
9. MPI as a Tool for Budget Optimization and Policy Targeting
- India has successfully used MPI to optimize fiscal allocations and improve budget targeting for social programs.
- MPI insights help identify which deprivations require urgent intervention, ensuring that limited resources are used effectively.
- India’s case highlights how cross-sectoral coordination—between government, private sector, and communities—is key to successful MPI implementation.
10. Strengthening Local Ownership and Sustainability
- Botswana’s experience demonstrates the importance of building local expertise and institutionalizing MPI within national planning frameworks.
- The country has invested in training policymakers and officials on how to use MPI data for decision-making.
- Sustained engagement across political transitions is key, ensuring that MPI remains a long-term policy tool rather than a short-term project.
About
➡️ Learn more about Multidimensional Poverty Index (MPI)
➡️ Access the Handbook: How to Use National MPIs as a Policy Tool: From Metrics to Policy
Thanks for the recording. I was hoping the gaps in MPI for the Caribbean might have been addressed but i see the region still not included. One suggestions for next time, If they wanted the world to use the MPI they should have started with smaller countries to show proof of concept and success and then gone global. Take up would be easier that way. Anyways, thanks for the webinar
Q: Can you share examples of Colombia and Chile mentioned?
A: Available with more details in the Handbook: Colombia has used its national MPI to prioritize regions and municipalities with high levels of multidimensional poverty to assign social protection beneficiaries. Chile has used its Social Household Register based on its national MPI to identify more than 25,000 children who do not attend school and design effectively targeted programs for school enrolment.
Q: Does the handbook refer only on the use of OPHI-UNDP MPI? Or does it include countries with their own Multidimensional Poverty measurements?
A: The handbook is mainly about national MPIs, meaning MPIs created and adopted by each country.
Examples used are based on the experience of countries supported by UNDP and OPHI (Costa Rica, Colombia, Chile, Panama, Botswana, Cuba, Ghana, Nepal, Mexico, Pakistan, Vietnam, Dominican Rep). There are over 40 countries that developed national MPIs and about 70+ that have reported under the SDG 1.2.2 on multidimensional poverty measures (nationally determined).
Q: Greetings from Montenegro, and congratulations on creating such a valuable resource with this handbook! I was wondering if you have had a chance to review the Survey on Income and Living Conditions (EU-SILC), which is widely used in EU candidate countries. In this survey, adopted by national statistical offices, the calculation of the poverty rate is replaced by the at-risk-of-poverty rate. Instead of setting a minimal threshold to determine absolute poverty (and consequently the poverty rate), EU-SILC defines 60% of the national median income as the benchmark, assuming that anyone below this amount is at risk of poverty.
However, being at risk of poverty does not necessarily mean falling into poverty. For example, someone categorized as at risk may never actually experience poverty. As a result, these statistics may not be particularly useful when the goal is to identify and support the most vulnerable populations.
Would you say that the handbook provides any approaches to overcome the limitations of EU-SILC without conducting new household survey, particularly in identifying those who are most in need based on MPI?
A: Warm greetings. Europe also uses, and 34 countries report against SDG 1.2.2 indicator in the global database, an indicator of 'At Risk of Poverty and Exclusion' (AROPE). This indicator identifies as deprived anyone who is a) at risk of income poverty or b) with quasi-joblessness or c) with severe material deprivation.
In fact we have worked with the EU-SILC dataset as part of the Net-SILC 2 and Net-SILC 3 processes of Eurostat, and built MPIs from them. However there are some challenges because a) the level of education is not necessarily comparable across EU countries and b) the health indicators are subjective or self-report. These create some challenges in building a stronger measure for the EU. (our papers are published and are online as OPHI working papers - the first co-authored with Mauricio Apablaza, and the second also with Anne-Catherine Guio).
Often national MPIs are built using existing datasets - perhaps Demograhic and Household Survey, or Multiple Indicator Cluster Survey, or an LSMS or Household Income and Expenditure Survey. However often governments seek to add a few questions - for example health or employment - if these are lacking in an existing survey. However it may be possible to start with whatever data exist, and improve datasets later.
Q: Has the hand book covered the exteranl dimmesions to the poverty index or it only deals with the indegneous political econmy i.e. the strings attached usually to Foreign Aid meant for social protection..?
A: The handbook deals with how we can address key stakeholders for them to implement and use the MPI. This means trying to reach people within countries and also externals. This starts with a mapping of key stakeholders.
Q: We had successfully advocated for the use of the MPI in country X and saw it being used as a KPI in at National Development Plan. However, a new government was ushered and literally discontinued the use of the index. Interested more on how to ensure continuity on the use of MPI when there is a change in government?
A: That is why the institutionalization of the MPI is key. A good institutionalization implies going beyond a current government. In Colombia this happens by institutionalization the MPI in the national statistical office. In the Mexican case it happens through Congress. It doesn't mean that these steps are 100% effective for continuity, but they help
Q: What were the most common political barriers encountered in adopting MPI as a policy tool, and how were they overcome?
A: In many occasions, poverty is the objective of one ministry (for example the ministry of social development), and therefore other ministries may not see poverty as their responsibility. The MPI can help to address this issue by showing that those ministries also include dimensions affecting poverty. The Mexican government implemented this coordination strategy 2014-2018
Q: In the region of Europe and Central Asia only 3 countries have NMPI embedded into their National accounts with support from UNDP and included this measure into national SDG1.2.2 statistical framework. Other 3 countries were supported by UNICEF, UNECE and WB accordingly with NMPI design and UNDP has little influence over promoting their MP poverty measure. How under such circumstances can NMPI metrics be translated unto policy change among other 12 UNDP programmic countries in the region? Shall we continue building the critical mass of NMPI countries or use 1-2 country examples as champions (agents of change). Any other options or suggestions?
A: One way is to have peer exchanges between countries. A very good way of learning is by listening the advances and challenges of similar countries. We can have 3 champions but then we can include more countries through these exchanges. The Multidimensional Poverty Peer Network actually does this: help the exchange of learning between countries.
Q: The Istanbul Regional Hub (IRH) is implementing a multi-country survey on the socio-economic vulnerabilities of Roma in Georgia, Moldova, and Ukraine. We were pleased to see that the majority of the Multidimensional Poverty Index (MPI) indicators are integrated into our questionnaire modules, while others have been incorporated through additional refinements.
Furthermore, we conducted additional consultations with the National Bureau of Statistics of Moldova to integrate some of the national MPI indicators into our questionnaires.
Incorporating national MPIs is a significant step toward institutionalizing monitoring, evaluation, and reporting on progress beyond individual projects. It also supports the continuous revision and improvement of policies.
Nevertheless, would be great to have more indicators from National Governments for Roma groups
A: Thanks, please also refer to the Handbook on How to Design a National MPI, available here: https://www.undp.org/publications/how-build-national-multidimensional-p…
Q: Is there an example of using/ adapting MPI for conflict or post-conflict contexts to inform recovery & reconstruction?
A: Yes many countries with conflict have MPIs - for example iwth UN agencies there is a national MPI for Yemen; also Somalia recently launched; Afghanistan launched its national MPI earlier and UN agencies are updating it. The State of Palestine has an official national MPI in teh global SDG database, but this has not been updated due to the present circumstances. Quite a number of other national MPIs are in progress in conflict-affected situations.
Q: Hello from UNDP India! How can the MPI be made more gender-responsive, given that it measures poverty at the household level—masking intra-household gender disparities—and lacks gender-specific indicators like unpaid care work, reproductive health access, and data disaggregation to capture intersecting inequalities faced by women?
A1: Please refer to the analysis done in LAC across 10 countries: https://www.undp.org/es/latin-america/publicaciones/indice-de-pobreza-m…
A2: Thank you - and perfect question! OPHI have a methodology for gender analysing any 'individual' indicators like education, school attendance nutrition, employment, and so on, to identify gender disparities and intra-household differences. UNDP and OPHI used some of this in our 2021 global MPI report, which found that 2/3 of the poor people in that year did not have any female in their household who had completed 6 years of schooling. Later this year we will be releasing figures for all global MPI countries related also to school attendance and undernutrition.
Q: The challenge with the computation of MPI is largely based on the availability of consistent data
A: That is very true. We are so grateful to national statistics offices and to those funding household surveys as these are the bedrock of poverty measures. Thus far other data cannot replace these in any country. OPHI also have now a compendium of all dimensions and indicators of national MPIs so we learn from one another.
Q:1. What are the key dimensions used to measure the Multidimensional Poverty Index (MPI)? What specific indicators are used to assess deprivation in each dimension of the MPI?
A: Please refer to our first Handbook on How to Design a National MPI - which refers to the basket and selection of indicators/dimensions to be measured, and how the methodology is applied at national/local level: https://www.undp.org/publications/how-build-national-multidimensional-p…
Q: How can our NMPI-related work be better utilised for achieving UNDP Poverty Moonshot?
A: The MPI, along with income based measures, vulnerability assessments (as well as MVI), have been used by many of the UNDP Country Offices to design their country relevant moonshots.
Q: MPI isn't just about identifying poverty, it looks at many dimensions and takes a truly human focused approach. How can it be strategically leveraged to push for human-centered design in development policy-making, projects and budgeting ? How can MPI measurements be used to shift development approaches to encourage more participatory governance or local economic ecosystems for example? I am particularly interested in citizen engagement and moving away from hand-holding approaches towards community independence for sustainable development. How are people TRULY engaged and their stories taken into account thanks to the MPI. Thank you!
A: I guess this would depend on the size of the survey (or the census) to have representation at the local level. When the survey is large, the MPI can be computed at state level. With census data, we can compute MPI at the municipal level or even further.
Q: What happens when countries add their own (advantageous) national indicators to the MPI method, and dilute the index and misrepresent national data? how do we, as UNDP prevent/deal with that?
A: This is a tough one. On one hand we would like countries to develop their own MPIs according to their goals, their national plans, their statistical strenghts. On the other hand we need "objective" measurements of poverty. For citizens, they need to know what is happening to poverty and its dimensions. I guess we need a lot of dialogue with the country to include their ideas but also recognize the importance of the dimensions and indicators "we" think should be there. Dialogue is, I guess, is an important option. The other one is the comparison with similar countries. If country X has a completely odd national MPI, in comparison to other countries, there might me a pressure to correct. I guess.
Q: Can MPI be used (or adapted/adopted) to assess the effectiveness of an initiative or Nature-based Solutions program in eradicating poverty?
A: Yes to both questions. On effectiveness, it requires a longer answer, but there are more 'accounting' based approaches for measuring impact that do not require RCT data only administrative or monitoring data, but can be linked to MPI.
On Nature based solutions you might look at UNDP-supported work in Sri Lanka, in which they developed a multidimensional vulnerability index (MVI) that added a dimension of environmental hazards to the national MPI."
Q: Do you have any suggestions on how to design and use MPIs for social registries? Measuring and targeting have different implications for policy makers and households. Are there any considerations that should be taken when applying an MPI approach to the deployment of social registries?
A: The use of MPI for household targeting using data from social registries is one of the fastest growing areas.
Ideally there is an official national MPI, and then the social registry-based MPI seeks to 'proxy' it. When these are aligned, then if the targeted activities are successful, the national MPI will also recognise and profile their success.
However social registries are far larger, so need to have simpler questions. Often the indicators need to be more 'verifiable' (or triangulated using home visits or other techniques) so that there are not allegations that respondents misrepresent their poverty to gain benefits.
The value-added of the MPI is cost-saving. Either a programme is universal, but the MPI identifies people who have 'fallen through the net' so they can be reconnected to existing programming. Another is to identify beneficiaries of targeted programmes - but then the identification is precise: who exact requires a scholarship or other benefit - metrics increases efficiency.
Q: What are the potential benefits and challenges of developing a gender-specific MPI? Has a gender-specific MPI been developed and implemented in any country before, and if so, what were the key findings and policy impacts?
A: Please have a look at the work done in LAC across 10 countries - on a gender-sensitive MPI measure - see it here (available in Spanish, please use internet tool for translation): https://www.undp.org/es/latin-america/publicaciones/indice-de-pobreza-m…
Q: For the Botswana case: you mentioned that you use the MPI to target your resources to the right group/ activities. Can you expand on that? Are you e.g. using the MPI to do results-based budgeting?
A: The Handbook uses the following examples on budgeting: Angola, Bhutan, Costa Rica, Mozambique. (please see here: https://www.undp.org/publications/how-use-national-mpis-policy-tool-met…)
Q: Based on the MPI when the schemes are designed for intervention in key areas, do these schemes imbed the graduation strategies? would you comment in this regard while refering to some of the graduation strategies?
A: Many household-targeting schemes do 'graduate' people when they are neither MPI nor monetary poor. However there are then wise schemes that create incentives (so poor people are better off if they graduate) and minimise people falling back into poverty. So there may be a period of transition, to ensure that poverty reduction is durable.
Q: Related to Nyendi's intervention and as part of the process of developing an MPI in any country, is there a way to also empower the citizen (in this case the poor) in the process?
A: This is really an important question. Many countries already draw on participatory exercises to identify the dimensions and indicators of poverty - for example UNDP supported one in El Salvador. The next step is to share information back to communities. That requires a lot of visuals, stories, local languages or videos / apps (when will UNDP do a national MPI app? I can't wait!). What local people often want to see is how they compare with their neighbours - so sharing that usually generates a buzz, but also clarity about where to focus. Sometimes there is also a healthy competition to reduce poverty fastest.
Q: Thanks to the panelists and moderator. Sierra Leone has produced 2 national MPIs (2017&2019), which are being actively used in government and development partners programming. In 2023 UNDP assisted the government to leverage the National MPI methodology to produce a Local MPI targeting remote areas in the borderland districts towards Liberia. The results confirmed the huge deprivations in the rural areas in energy, education, food security, housing and living standards. The report was launched on Dec 17 2024, and a panel discussion was organized to discuss the key policy areas including energy, human capital development, social protection targeting, and statistical harmonization.
My question is: how is the issue of data and data harmonization addressed in the book, given the different MPI measures now being computed?
A: For harmonization of data issues, please also refer to the UNDP-OPHI Handbook on How to Build a National MPI: https://www.undp.org/publications/how-build-national-multidimensional-p…
Q: In your experience, what are the ways rising authoritarianism affects the use of the MPI—both by restricting data transparency and limiting its role in shaping inclusive, evidence-based policies to address multidimensional poverty—and how can these challenges be circumvented to ensure the MPI remains an effective tool for policymaking?
A: Another tough one! When governments move towards less transparency (alternative or "other" data), we need to push back, in the middle of hard barriers. Academia can help, by showing the "true" poverty results. Civil society should keep demanding proper information. International institution could also help by insisting on good evidence. Regarding the strength of governments we need to keep pushing for transparency.
Q: While MPI combines two key components (MPI = H X A), i.e. ‘the percentage of the population who are multidimensionally poor & intensity of poverty, how programs to the sectors most in need (sectoral programme – e.g. Health, Education etc), can be addressed by the use of MPI especially, while resources are much limited compared to programme need, how the resources can be allocated to the most urgent needs.
A: Please refer to the country experiences using the MPI on budgeting allocation (country examples: Angola, Bhutan, Costa Rica, Mozambique) - more at: https://www.undp.org/publications/how-use-national-mpis-policy-tool-met…
Q: Given the persistent structural challenges in Central Asia, such as weak regional trade integration, a high reliance on extractive industries, rural poverty, and others, how can the MPI be adapted to better capture and address these region-specific barriers to poverty reduction? Are there specific MPI indicators that could be adjusted or introduced to reflect the realities of economic diversification and labor market vulnerabilities, as well as and climate resilience in Central Asia?
A: We have been developing a family of multi-dimensional indicators through our work at the Human Development Office, but also our programmatic work on inclusive growth. MPI + other contextual indicators provide a strong basis for addressing trade, and structural exclusions and SME involvement.
Q: Thank you for preparing the handbook with many examples from different countries. MPI is calculated based on household surveys, which are representative but only make up a few percent of the total population. What could be the mechanism for identifying the remaining multidimensionally poor households that did not participate in the surveys?
A: An increasing number of countries - from Angola to Mexico and Colombia to Viet Nam and Bhutan - make a version of their national MPI fro their housing and population census. This enables the to have a high-res poverty map and measure, with appropriate privacy gains. This often needs some adjustments, but is very useful. You could look at Ghana whose statistics office issued 261 reports of poverty based on their census, so local actors had the information they needed to act.
The other data source many countries use are the social registries that are used to target conditional or unconditional cash transfers. Often using these same datasets (sometimes with small changes to the questions), countries target the MPI poor using a proxy MPI.
Q: Can IoT or/and Open data be used to make the MPI, or even just supply and expand it for more transparent data and a bit less reliance on surveys? For health and nutrition for example?
A: We have used satellite data to incorporate some environmental conditions like fire, forest loss, air precipitate quality, cyclones etc. We have looked at health and nutrition data quite deeply but it is not able to be used in most countries at this time - perhaps in the near future, we hope.
Q: In resource rich contexts marked by political crisis and division, has there been experiences of MPI being implemented in such contexts and in ways that helps inform policy making and budget allocation but also prevents potential politicization and hence resulting in exacerbating the division? Any relevant experience or learnings on this would be appreciated.
A: Thanks - please refer to the experience of Afghanistan, Northern Uganda, Jordan (with refugee communities), Somalia, etc. Kindly refer also to the UNDP-OPHI Handbook on How to Build a National MPI, available here: https://www.undp.org/publications/how-build-national-multidimensional-p…
Please log in or sign up to comment.