EVERY few years, Pakistan’s poverty line moves, and the poor multiply or vanish accordingly. Sometimes 22 per cent of Pakistanis are poor, sometimes 39pc, sometimes 44pc. Each new estimate arrives with a World Bank press release or a government statement claiming progress or crisis. Yet few ask why the numbers differ, what they really mean, or whether the differences are even statistically meaningful. Poverty has become a ritual, not a revelation, a statistical performance that signals concern without demanding reform.

Pakistan currently lives under multiple definitions of poverty. The national ‘cost of basic needs’ line places about 21.9pc of the population below the poverty threshold, based on 2018-2019 survey data. The World Bank, using 2021 purchasing power parities, offers its own global poverty lines, $3.00, $4.20, and $8.30 per day, designed for international comparison.

On the $4.20 line, roughly 44pc to 45pc of Pakistanis qualify as poor. Under the older 2017 PPP series, the figure was around 39pc. Then there is the multidimensional poverty index, or MPI, which measures deprivation across health, education, and living standards: according to the Oxford Poverty and Human Development Initiative, about 38pc of Pakistanis are multidimensionally poor, while another 13pc are on the edge. The World Bank’s nowcasts add another layer of confusion, projecting, not measuring, poverty at 25.3pc for 2023-24 through simulation rather than survey.

These numbers differ not because poverty changed overnight but because the yardsticks did. The problem is not that we have too little data, it is that we have too much of the wrong kind, deployed without clarity or context.

Then some talk of ‘deprivation’. It is not simply low income but the denial of essentials, which include clean water, electricity, housing, education, nutrition, security, voice, and dignity. It is no longer about how much people earn, instead, it is about whether they can live with choice and dignity.

We are now trapped in a world where counting poverty has replaced combating it.

We are now trapped in a world where counting poverty has replaced combating it. Every new dataset becomes an opportunity for institutional signalling. Governments trumpet falling poverty as a proof of success; oppositions invoke rising poverty as the evidence of failure. Donors use new headcounts to justify fresh loans or social-protection projects. International agencies build entire careers around ‘poverty diagnostics’. None of this tells us how to generate jobs, reform education, or modernise cities — the things that actually lift people out of poverty.

There is also a statistical sleight of hand. Poverty figures are rarely reported with confidence intervals. In surveys such as the Household Integrated Economic Survey, the sampling error is typically plus/minus two to four percentage points. A change from 22pc to 24pc may not be statistically distinguishable, yet it is presented as gospel truth.

Worse, modelled projections, like the World Bank’s nowcasts, are treated as hard data when they are, in fact, simulations based on inflation and consumption trends. Without confidence intervals, without transparency on method, these numbers are little more than educated guesses dressed in mathematical clothing.

Meanwhile, the poverty-measurement complex has become an economy in itself. Economists recalibrate PPPs and issue new global benchmarks. Consultants write reports and regressions. Governments set numerical ‘targets’ aligned with the Sustainable Development Goals. International agencies publish ‘poverty assessments’ that do not alter policy. Everyone stays busy counting the poor, while few take responsibility for why they remain poor.

Only in limited contexts does measurement serve real policy purpose. The Benazir Income Support Programme is one such case, where data determines who receives cash transfers. But they hardly ever graduate from poverty. Poverty measurement has been harnessed to administer relief, not to trigger reform. The state uses data to stabilise poverty, not to end it. The poor are categorised, catalogued, and queued, but rarely are they liberated.

What do we actually learn from all these numbers? Absolute lines tell us who cannot afford essentials. Relative lines tell us who is left behind. Multidimensional indices tell us what they lack. But without integrating these insights, publishing margins of error, and linking measurement to action, they remain statistical décor. They are more about moral satisfaction than material change.

If poverty measurement is to serve any real purpose, we need clarity, confidence, and consistency. Every estimate should specify which line, which PPP vintage, and whether it is survey-based or modelled. Every headcount should include its confidence interval. The national poverty line should remain stable, updated only for inflation.

Most important of all, we need to focus on social mobility to unlock the hurdles to people bettering their lot. Poor don’t need state patronage; they need the opportunity to progress. Until then, poverty measurement will remain what it has become: a convenient currency for donors, politicians, and consultants, a moral fig leaf for systems that perpetuate dependence. We have multiple poverty lines but few escape routes. We debate decimals, not dignity. The real poverty in Pakistan today is not of income or data. It is a poverty of imagination, of courage, and of reform.

The writer is former VC of the Pakistan Institute of Development Economics and deputy chair of the Planning Commission. He is currently director at the think tank Socioeconomic Insights and Analytics.

Published in Dawn, November 12th, 2025