Introducing the PIP Innovation Hub: A home for experimental work on poverty and inequality measurement

The PIP Innovation Hub is a new platform to showcase experimental work on poverty and inequality measurement undertaken at the World Bank. It enables researchers, policymakers, and the public to explore how alternative methodological approaches, drawn from peer-reviewed research, affect poverty and inequality estimates. It does so by placing these alternatives alongside the World Bank’s official numbers, making the implications of methodological choices transparent and accessible. 

The global poverty numbers are one of the most important statistics that the World Bank produces. The estimates in the Poverty and Inequality Platform (PIP) inform policy debates, guide resource allocation, and shape our collective understanding of who is poor and where. The underpinning data have strongly improved since these estimates were first published in the 1990s, and methodologies continue to evolve in light of new evidence and to adapt to changing lives. Different methodological choices – e.g., on whether to use income or consumption, how to account for inequality within the household, or whether (and how) to adjust for the gap between surveys and national accounts give rise to very different estimates, even if drawn from the same underlying household survey data. These differences matter – each decision can shift millions of people above or below the poverty line, or change a country’s inequality ranking significantly. A first step towards explaining these differences, is recognizing them and putting them on a comparable basis.

The Hub launches with two core components. The Deep Dives section presents estimates from four peer-reviewed papers, each of which modifies a different step of the standard PIP methodology, allowing users to compare the official poverty estimates to those produced using the alternative approach. Over time, additional methods will be added. All estimates are available for download. The Research Repository provides AI-powered semantic search over hundreds of World Bank Policy Research Working Papers, making it easier to discover relevant research on poverty measurement and related topics.

How methodological assumptions matter

Broadly speaking, the PIP methodology follows a five-step process. First, household survey data are obtained from relevant sources. Second, the survey data are used to create welfare aggregates – estimates of households’ income or consumption. Third, the welfare aggregates are adjusted for differences in price levels across countries and over time. Fourth, poverty and inequality are estimated for each country-year. Fifth, the estimates are extrapolated or interpolated to a common reference year and a population-weighted global poverty rate is calculated. 

We will examine various methodological assumptions, including the degree to which households share resources, the choice between income and consumption surveys, and the way subnational variations in poverty rates are addressed. To illustrate how these choices matter in practice, consider extreme poverty in Colombia, a country with relatively frequent household surveys.

It is important to note that while comparisons between the main PIP and alternative estimates within a given year and dataset are informative, comparisons across years or across methods should be made with caution. Each method uses a different data vintage, different survey years, and different underlying assumptions. The value of the Innovation Hub lies not in ranking methods but in making these choices – and their consequences – visible.

The Research Repository

The second part of the Innovation Hub is the Research Repository which provides access to hundreds of World Bank Policy Research Working Papers on poverty, inequality, and related topics. The repository is powered by a semantic search algorithm – users can describe what they are looking for in natural language, and the system retrieves the most relevant papers based on the meaning of the query, not just keyword matches. This capability builds on recent advances in natural language processing. 

The repository will be updated regularly as new papers are published. This ensures that researchers and policymakers have easy access to the latest methodological developments.

The Research Repository is part of the World Bank’s broader efforts to leverage cutting-edge AI and technology to improve knowledge dissemination and data discovery – making it easier for the global development community to find, understand, and build on the institution’s research output.

Looking ahead

The PIP Innovation Hub is designed to grow. As new research proposes new approaches to measuring global poverty and inequality, those methods can be incorporated into the Deep Dives framework – enabling ongoing, transparent comparison with official estimates. All underlying data are available for download, supporting further analysis and replication.

Measuring poverty well is not merely a technical exercise. The numbers we produce shape policy priorities, resource allocation, and our understanding of progress toward the first Sustainable Development Goal. By making the methodological frontier accessible and transparent, the PIP Innovation Hub aims to make the heterogeneity of methodological choices transparent and thus strengthen the evidence base that underpins these consequential decisions. 

Explore the PIP Innovation Hub.

Would you like to be updated with the latest news on PIP? Register to our newsletter here.

The authors gratefully acknowledge financial support from the UK Government through the Data and Evidence for Tackling Extreme Poverty (DEEP) Research Program.