Measuring multidimensional poverty

In the opening speeches, the organizers stressed on the importance of calculating multidimensional poverty, especially after the economic and social impacts of the COVID-19 pandemic, which has resulted in an additional 16 million people falling into poverty in the Arab region. They pointed out that the workshop aims to present a new, broad and comprehensive concept of poverty that does not depend only on the material perspective, but includes many dimensions, including health, education, employment, assets, services, standard of living, and others. It is also necessary to take the peculiarities and needs of each country when calculating multi-dimensional national poverty.

The organizers mentioned that measuring poverty alone is not sufficient to solve the problem, but it is an important first step that must be taken to develop effective policies that change the situation of the poor, raise them from the category of poverty to other non-deprived and middle-class categories.

The objectives of the workshop was also set forth: Enhance the capability of participants in understanding and measuring multidimensional poverty; empower participants to design MPIs; evaluate the state of multidimensional poverty; understand the simulation functionality of MAT; and finally empower participants to use the MAT.

In this session, development challenges in the Arab region were presented. The Human Development Index recorded slow progress between 1970 and 2000 compared to global averages. It also becomes clear to us that after the year 2000, the rate of development in the Arab region has recorded modest rates, which are generally lower than global averages.

On the other hand, multidimensional poverty in the Arab region seems to be declining when measured over time, although it is slow compared to the decline in global rates. Results also show how the region has seen success in the reduction in inequality and improvement in gender disparity

Money metric poverty was introduced in this session, as results show that the Arab region has the second highest poverty rate in the world, using the approved global poverty lines (1.9 and 3.2 dollars per day). This confirms that the fiscal space constitutes a major challenge for the future of this region, where debt seems to be rising dramatically. In addition to this, social expenditures in this region receive the least percentage of financial support from governments, and this affects the quality of some services in these countries.

In summary, and in the near future, this region will face a major problem, which is the limited fiscal space for social spending due to the high debt and deficit.

The session presented the Alkire-Foster methodology from OPHI used to calculate multidimensional measures that are used to produce the MPI which shows the breadth and depth of poverty. The session covered the flexibility of this method and how it can adapt to different contexts and how it may be tailored to specific countries.

The session then moved to more technical aspects of the methodology in terms of the (1) logic of selecting indictors, (2) setting deprivation cut offs, (3) computing the deprivation matrix, (4) selecting weights, (5) building the weighted deprivation matrix, (6) computing deprivation scores, (7) selecting the poverty cut off, (8) creating the censored matrix and finally (9) computing the MPI.

The session also covered the various redundancy and robustness testing methodologies. It assessed how robust our MPI measures are when subject to variation in normative assumptions.  

In conclusion, the global framework of the multidimensional poverty index with its three dimensions and its ten indicators was explained, and it was found that more than 1.2 billion people (one billion and three hundred million people) are poor, especially since the majority of them live in homes where women are deprived of education.

The session began by briefly discussing the first Arab report on multidimensional poverty, which was approved by the Arab Ministerial Social Council in 2017. It showed that the Arab region was exposed to major setbacks and challenges, starting with displacement due to conflict and with economic crises rising especially due the COVID-19 pandemic, which further exacerbated multidimensional poverty rates.

The Arab framework was also revised in 2019 in order to obtain better accuracy in capturing poverty levels and taking into account the increasing cases of economic difficulties in many Arab countries, as well as to facilitate comparison between countries over time and compatibility with surveys.

This session concluded by going through country cases and lessons learned on building a national MPI in Egypt, Palestine and Iraq, in addition to a discussion of the time frame for these studies and challenges faced.

The following sessions presented the advanced tool created by ESCWA: The Multidimensional Poverty Index Assist Tool. The participants were divided into small working groups and each group built a conceptual framework used to compute a sample MPI using survey data from the State of Iraq. Participants then presented their discussions and findings.

In this session, three simulation methods of the Multidimensional Poverty Index were presented. The different methods help answer the following questions:

• What to do when facing technical challenges related to limited survey data?
• How to predict now casting methods / how to predict and simulate shocks to certain dimensions and indicators based on real data.
• How to predict for the future
• How to guide a country’s public policy that helps reduce MPI in the future

In the eighth session, the entry points for evidence of multidimensional poverty within the life cycle of the "social protection strategy" were addressed. Using social protection strategies as a tool to identify additional components of cash transfer programs (in relation to health or education needs) or to identify implementation priorities in social protection strategies when faced with a limited budget. There has been a transformation in social assistance programs that further help households break out of cycles of poverty. The session also highlighted pathways leading out of poverty such as referral mechanisms to case management and cash transfer programmes.

Such referral mechanisms aim to connect suitable beneficiaries either with TVET programs and thus provide optimal access to formal employment or - alternatively - reduce intergenerational poverty while increasing productivity and human capital of households through providing assets such as access to land, livestock, machinery to produce etc. for those who live in areas with limited formal employment opportunities.

The session began by suggesting ways to develop an effective communication plan that will help launch the MPI in a country.

This session continued forward to discuss how countries may use the national MPI to inform policy and its usefulness for analytical purposes. Below you will find the most prominent uses of the Multidimensional Poverty Index in this regard:

(1) It complements money-metric measures (2) helps track poverty over time (3) can help allocate recourses effectively by sector and region (4) assists in targeting marginalized groups (5) promotes multi-sectorial policy coordination (6) helps adjust policies to ‘what works’ (7) trend analysis and targeting strategies and (8) transparency and accountability so that all stakeholders are involved.

The session concluded on looking at policies and strategies and re-prioritizing public spending and delivery of services which ensure access, quality and equity of basic services.

A study was presented to develop two innovative approaches that assess the impact of reforms in the social protection system using the case of Morocco and the Revised Arab Framework for measuring policy effectiveness in relation to multidimensional poverty. The difference between the two approaches is the targeting mechanism for individuals benefiting from the social protection system.

In the first approach, beneficiaries are identified by random selection from disadvantaged households on each of the indicators in the MPI.

In the second approach, beneficiaries are identified using a model based on the probability theory to determine which households would change their status from deprivation to not.

The session concluded by analyzing challenges to the approaches using confidence interval, simulation and distributional analysis.

In developing countries, household income or expenditure surveys are usually conducted every three or five years, but the periods vary between countries. Household survey questionnaires can vary and similar surveys may not be completely comparable due to differences in survey sampling methods. These problems have become less prevalent, as survey methods are improving and have become more standardized than before, but it is still difficult to achieve accurate comparability. To overcome some of these problems, ESCWA works on:

• Internationally comparable poverty lines are useful for comparison. However, the poverty line in for analyzing poverty in a particular country must reflect its economic and social conditions, and it may also need to be modified over time to suit a different reality, if there is a difference in prices or access to goods and services. Therefore, ESCWA has developed a method to address these gaps (Both parametric and non-parametric models)

• Growth in national accounts (GDP, PCE) is one of the best measures for projecting growth in average income in surveys. How much growth of national accounts is reflected in surveys? What is the ratio? Usually for developed countries and countries using expenditure surveys, the ratio can be just under 1. For developing and non-developed countries, this ratio is lower as surveys are more prone to measurement errors. ESCWA has developed several regression and clustering models to solve this problem and calculate the ratio for each country, thus predicting the average income of each country in the future.

• Finally, in order to answer the question of how the distribution of income has changed from one country to another and over time, ESCWA has developed the tools necessary to account for these changes

ESCWA is currently working on developing a tool that provides an answer to all the above-mentioned questions and enables the user to predict measures of monetary poverty. The user of this tool will benefit:

• Better understanding of theoretical concepts

• Better understanding of results and sensitivity of each of the contributing factors to money-metric poverty

• Nowcasting and forecasts beyond 2020

• Several methodological frameworks will be presented, and they will be able to pick whichever they prefer

Poverty can be calculated based only on aggregate data and there is no need for data at the micro level

The concluding session congratulated participants for the fruitful participation and completion of three prolonged days of technical training on multidimensional poverty, and briefly on money metric poverty. In hopes that next year a similar workshop will be prepared and that the work on poverty is continuous.

This is the beginning of a great cooperation between the LAS, DESA, ESCWA and member state representatives to work together towards a prosperous region.

Suggestions and recommendations:

1 - To persevere in holding similar sessions annually, as it enhances:

• Exchange of experiences between countries in the region
• Informing the countries of the region about the regional challenges and working on ways to solve them
• Keep the Member states informed about the latest studies done by ESCWA

In this regard, the League of Arab States and the Arab Poverty Center announced during the conclusion of the session that they would propose a follow-up training workshop next year with ESCWA and DESA .

2- ESCWA will have to support any Member state country that is willing to compute its own National MPI

3- Organize similar workshop to estimate money-metric poverty

4 - Studying the option and possibility extending the duration of the workshop for two additional days to allow the participants to work on practical exercises. This will enable them to have sufficient expertise going forward.