The World Bank Uses System Dynamics to Identify Root Causes of Poverty

The Problem

Poverty in Madagascar has remained persistently high and somehow complex, despite many attempts to reduce it through anti-poverty policies and interventions. As shown in the Macro Poverty Oulook, the share of the population living in poverty (US$2.15 in 2017 PPP) was 81.7% in 2020, 81% in 2021 and 80.6% in 2022.

One of the reasons why past anti-poverty policies have had less than desired success in reducing poverty is the limited understanding of its determinants. Empirical investigations of the determinants of poverty often use linear specifications that regress objective measures of poverty such as income or consumption against observable indicators theoretically justified to explain poverty.

To uncover the root-causes this analysis used a systems thinking approach to characterize a high-level, dynamic, and multi-dimensional view of the poverty trap with Causal Loop Diagrams (CLD) generated from focus group discussions in 6 provinces across the country.

The Solution

The World Bank organized a series of focus groups across a variety of diverse communities within judiciously selected communes in Madagascar. The aim was to identify, understand, and address the unique poverty challenges they face, and to enrich these findings with qualitative data. These focus groups were intentionally structured to spark discussions about the primary obstacles encountered by the communes. Insights from participants across different ages and genders were used to deconstruct these challenges into several root causes. Each root cause was further distilled into simpler, more comprehensible sub-causes.

The causes and their corresponding sub-causes were reanalyzed after deconstructing the poverty issues faced by each commune. This allowed for diligently examining the intricate interconnections that define these challenges. Special emphasis was given to understanding the interactions among various sub-causes and the overall impact on the system. Specifically, we identified reinforcing patterns or loops that sustain poverty and negate efforts by both individuals and policy interventions to alleviate it. In the end, insights shed light not only into the economic determinants of poverty but also the cultural, political, and behavioral intricacies across different demographics, including gender, age, occupation, and community type (rural/urban).

Analyzing poverty by amplifying the voices of the impoverished and evaluating qualitative data was influenced by a seminal World Bank book titled “Voices of the Poor: Crying out for Change.” This report harnessed focus groups across 23 countries to relay the experiences of the impoverished in their own words. In doing so, it underscored significant multidimensional aspects of well-being that had not been previously explored in conventional poverty analyses.

In addition to highlighting the voices of the impoverished and the crucial subtleties pertaining to their circumstances, this project incorporated systems thinking tools. This innovative approach enabled us to evaluate the qualitative data from the focus groups and discern the pivotal narratives and relationships explaining the perpetual cycle of poverty experienced by many households.

The Causal Loop Diagram is a partial representation of the multi-dimensional poverty trap from the perspective of urban households. The CLD identifies eight feedback loops representing the urban poverty trap that inhibits the group’s ability to:

• secure high income livelihood opportunities (R1)

• alleviate impact of cost of living (R2)

• uplift socio-economic status (R3)

• participate in decisions (R4)

• develop resilience to shocks (R5)

• mitigate increasing cost of business (R6)

• mitigate shocks to business (R7)

• enhance skills (R8)

Outcomes

Clearly addressing one or few sub-causes will not generate an overall change required in poverty trap situation.  The systems thinking exercise proposes a portfolio of upstream and downstream entry points and pathways to improve the ability of communities to move out of poverty over time. This will inform design of projects by the World Bank to activate the identified entry points.

In particular, the project resulted on a list of potential interventions that are directly linked to the feedback loops driving the system. For example, one of the potential interventions could be to promote business and entrepreneurship. Stakeholders in the focus groups thought that training, mentoring, infrastructure, and financing support could help circumvent the current low levels of business and economic activities (R1), and to diversify employment opportunities (R8).

Do you want to know more?

Connect with Amitaksha Nag, Francis Muamba Mulangu