New Poverty Taxonomy Reveals Five Distinct Faces of the Poor in Pakistan

The World Bank’s report “Towards a Taxonomy of the Poor in Pakistan” categorizes the bottom 40% of households into five distinct groups based on socioeconomic and geographic traits. This data-driven segmentation aims to guide more targeted, effective poverty alleviation policies tailored to each group’s unique challenges.

In a groundbreaking initiative, the World Bank’s Poverty and Equity Global Practice, led by researchers Christina Wieser and Ibrahim Khan, has released “Towards a Taxonomy of the Poor in Pakistan”, a report that reframes how poverty is analyzed and understood. Drawing on the 2018–19 Household Integrated Economic Survey (HIES) from the Pakistan Bureau of Statistics, and supported by geospatial data from institutions like the World Bank’s Rural Accessibility Index and the Pakistan Agriculture Research Council, the report offers an empirically grounded segmentation of the country’s bottom 40 percent of consumption earners (B40). It breaks away from viewing poverty as a single, homogenous experience and instead introduces a classification system that reveals the diverse challenges faced by different poor households. This approach is designed not to identify causes of poverty, but to offer policy-relevant groupings that highlight which constraints are most pressing for specific population segments.

Growth Interrupted: A Slide Back into Vulnerability

Between 2001 and 2018, Pakistan made significant strides in poverty reduction, with real household consumption rising by 60 percent and poverty falling from 64.3 to 21.9 percent. But the country’s trajectory was disrupted by overlapping crises. The economic fallout from the COVID-19 pandemic, combined with macroeconomic instability, inflation, and the unprecedented 2022 floods, eroded prior gains. Poverty is now projected to have reached 25.3 percent by 2023. These developments revealed that poverty in Pakistan is not only multidimensional but also dynamic, shifting in response to climate shocks, global disruptions, and internal governance challenges. The persistence of regional disparities and weak labor market outcomes, especially among women and youth, further underscores the complexity of addressing poverty through generic interventions.

The Five Faces of Poverty in Pakistan

To navigate this complexity, the researchers applied hierarchical cluster analysis, a machine learning method that groups households based on shared traits without relying on pre-defined categories. Using 26 carefully selected indicators spanning education, employment, assets, infrastructure, and climate vulnerability, they classified the B40 into five distinct groups. Each of these clusters presents unique constraints and policy needs.

Group 1 represents the ultra-poor in remote rural areas who rely on unskilled sharecropping and receive public cash transfers, mainly through BISP. They live in some of the harshest conditions, with the lowest educational attainment and infrastructure, and face the greatest exposure to floods and droughts.

Group 2 includes smallholder farmers involved in self-cultivation. They own slightly more productive assets, motorcycles, sewing machines, and exhibit higher levels of education than Group 1. However, their productivity remains limited, and they remain vulnerable to agricultural shocks and market isolation.

Group 3 are remittance-dependent households with low employment levels. Often headed by women, these families depend on internal remittances sent by migrating members. Although they receive less public assistance, the income from remittances is insufficient to lift them out of poverty, particularly as many remain outside the formal labor market.

Group 4 is perhaps the most intriguing: urban, relatively well-educated households working in semi-skilled industrial and service sector jobs. They own productive assets and live in better conditions compared to others in the B40. While still officially classified as poor, their upward mobility is evident, making them strong candidates for exiting poverty if economic conditions improve.

Group 5 encompasses unskilled daily wage workers in construction and informal services. Spread across both urban and rural areas, they share poor housing conditions, minimal education, and receive limited but essential public transfers. This group is the urban equivalent of Group 1’s rural sharecroppers.

Assets, Employment, and the Climate Divide

The report uses an asset-based framework of shared prosperity to interpret the differences between groups, analyzing how variations in human capital, productive assets, income sources, and exposure to shocks shape household wellbeing. Access to education, for instance, strongly influences which sectors people work in and how resilient they are to economic changes. The urban Group 4, with better education and access to semi-skilled jobs, contrasts starkly with Group 1, where illiteracy is widespread and livelihood options are limited to unproductive farming. Similarly, climate vulnerability, particularly to floods and droughts, plays a decisive role. Groups 1 and 2, both rural and heavily reliant on agriculture, face significantly greater environmental risk than urban households. Public assistance, while vital for some, varies widely in reach, suggesting a need for refined targeting.

From Analysis to Action: Shaping Smarter Interventions

This taxonomy is more than an academic exercise, it’s a powerful tool for policy design. By illuminating the heterogeneity among Pakistan’s poor, it enables the government and development partners to tailor interventions to the distinct needs of each group. For example, rural sharecroppers in Group 1 need improved social protection and climate-resilient agriculture, while urban daily laborers in Group 5 could benefit from vocational training and better housing. Smallholder farmers in Group 2 require market access and improved technology, whereas remittance-receiving households in Group 3 might need childcare support and employment linkage services. Group 4’s constraints are macroeconomic, job creation and wage growth could help them graduate out of poverty altogether.

Ultimately, “Towards a Taxonomy of the Poor in Pakistan” is a call for evidence-based policymaking in a country where poverty remains a moving target. It urges planners to move beyond broad averages and focus on lived realities. As Pakistan grapples with climate change, demographic pressures, and economic volatility, this kind of analytical depth could be the key to unlocking more inclusive and sustainable development.