Has India really ended poverty? What the World Bank's 2026 report shows, and why economists still argue

Has India really ended extreme poverty?

Almost. In 2024, about 2 in every 100 Indians lived below the World Bank's extreme poverty line, roughly Rs 68 a person a day. In 1990 it was about 49 in 100.

Raise the line, though, and the picture turns over. India is a lower-middle-income country, and the World Bank's line for that group is $4.20 a day, roughly Rs 95; about 15 in 100 Indians were below it. Its line for upper-middle-income countries, the group India hopes to join, is $8.30 a day, roughly Rs 187. About 4 in 5 Indians, 79 in 100, were below that in 2024. All these numbers come from the same household spending data. They are true at once.

The 2% is the World Bank's best estimate, not a settled fact. How far poverty has really fallen is disputed, for reasons part four explains.

Most of what follows explains how the low number and the high one can both be true. The World Bank's Poverty, Prosperity, and Planet Report, published in October 2026, is the latest round in a fight about Indian poverty that is decades old. One side reads the numbers as the end of mass poverty. The other reads them as a low bar cleared, with a new survey flattering the result. Both sides are looking at the chart below.

The lines fell at very different speeds. At $3 a day, the share went from 49% in 1990 to 29% in 2011 and 2.3% in 2024, about 3.3 crore people. At $4.20, the middle line, it went from 77% to 59% to 15%, about 21.6 crore people. At $8.30 it barely moved for two decades, from 97% in 1990 to 93% in 2011, before falling to 79% in 2024. That is about 114 crore people.

The figures for 2024 are estimates. India's last household survey was the Household Consumption Expenditure Survey (HCES) of 2023-24, run by the National Statistics Office, which found 2.6% below $3 a day, 16.1% below $4.20 and 80% below $8.30. The World Bank carries those survey readings forward a year using growth in the economy.

What does $3 a day mean in rupees?

About Rs 68 a person a day, or Rs 2,060 a month, at 2023-24 prices. For a family of five, that is about Rs 10,300 a month for everything: food, rent, school, bus fare, medicine. The $4.20 line works out to about Rs 95 a day, or Rs 2,880 a month. The $8.30 line is about Rs 187 a day, or Rs 5,700 a month.

These are not dollars converted at the exchange rate. The World Bank uses purchasing power parity, which asks what a dollar buys in America and how many rupees buy the same things in India. Because many goods and services are cheaper here, one of these dollars was worth about Rs 19.5 in 2021 prices.

The $3 line is the typical poverty line of the world's poorest countries. Clearing it means India is no longer poor by the standard of the poorest countries on earth. It does not mean an Indian family just above Rs 68 a day is eating well, can afford a hospital visit or has any savings.

There is one more way to see how low that bottom line is. India's last official poverty line, the Tendulkar line of 2011-12, was Rs 816 a person a month in rural India and Rs 1,000 in urban India. Carried forward with consumer prices to 2023-24, that is roughly Rs 1,660 and Rs 1,970 a month. The World Bank's "extreme" line, at about Rs 2,060, sits just above it. The two are measured differently, so they would not count exactly the same people. But when the World Bank calls its line extreme, it is drawing it at about the level India itself once called the poverty line.

For a sense of what household budgets look like, the 2023-24 survey is useful. Food is now less than half of what a family spends: about 47% in villages and 40% in cities. Average spending was Rs 4,122 a person a month in rural India and Rs 6,996 in urban India. The poorest 5% of rural Indians spent Rs 1,677 a month, about Rs 55 a day.

Did the whole world get here, or just India?

Just India, or close to it. Worldwide, the share of people below $3 a day fell from 43% in 1990 to about 11% by 2019, then stopped. The World Bank estimates it at about 10% in 2026, roughly 81 crore people. Most of them now live in Sub-Saharan Africa, where the share has stayed near 45% for a decade, and in countries the World Bank classes as fragile or in conflict.

India went the other way. Its rate fell through the same years in which the world's stalled. India now holds about 4% of the world's extremely poor, against about 18% of the world's people. The report is blunt about why India still appears near the top of the list.

By headcount, India ranks fifth, with about 3.3 crore people, behind Nigeria, the Democratic Republic of Congo, Pakistan and Ethiopia.

One finding deserves more attention than it has had. Between 2013 and 2024, poverty fell sharply among people across the world who never went to school. The report estimates that India alone accounts for about 70% of that fall within countries.

There is a gap in the record. From 2015 to 2021, India's numbers in the World Bank's database came from a private survey, the Consumer Pyramids Household Survey. In 2026 the World Bank removed them, over concerns about comparing them with the government's new survey. India therefore has no survey reading in the World Bank's database between 2011-12 and 2022-23. The line through those years is the World Bank's estimate.

If almost nobody is that poor, why does it feel like most people are still struggling?

Because most Indians live just above the lowest line, not far above it. In 2024, about 76 in 100 Indians lived between $3 and $8.30 a day, between Rs 68 and Rs 187. That band is where most of the country is.

Average spending per rural person, Rs 4,122 a month in 2023-24, is itself below the $8.30 line of about Rs 5,700. That comparison is rough, because village prices are lower than the national average the line uses. The point holds anyway. By the yardstick of upper-middle-income countries, the typical rural Indian is poor.

On the world stage, this changes India's role completely. India has 4% of the world's people below $3 a day. It has 15% of those below $4.20, the largest single group anywhere, at about 21.6 crore. And it has 30% of everyone below $8.30, about 114 crore people.

The World Bank also measures how far each country's people fall short of a modest standard of living: about $28 a day, the income of a typical person in a country just becoming high-income. In 1990, China accounted for 37.5% of the world's total shortfall and India for 15.7%. By 2026, China's share had fallen to 7.9%. India's had risen to 18.5%, the largest of any country. The two started in roughly the same place and went in opposite directions. The report's own verdict: "India still weighs heavily in the global gap."

How close to the edge are most Indians?

Close. The World Bank's data rank all Indians from poorest to richest and split them into a thousand groups of equal size. Take every group's spending down by 10%, as a bad harvest, an illness or a lost job might, and count again. The share below $4.20 a day rises from about 15% to 23%. The share below $3 roughly doubles, from 2.3% to 4.4%.

That is a thought experiment, not a forecast. It shows how many people sit just above each line.

If only 2% are poor, why do 80 crore people get free grain?

Because the free grain is part of why the number is so low.

Since January 2023, the government has given foodgrain free to more than 80 crore people under the National Food Security Act. In November 2023 the Cabinet extended the scheme for five years, to about 81.35 crore people, until the end of 2028. The government's own press release described them as "poor and poorest of poor people".

The World Bank counts that grain. In its method for India's recent surveys, it values free and subsidised ration grain at the local market price of similar grain, not at the zero rupees the family paid. So a family that would otherwise sit just under Rs 68 a day can be lifted over it by its ration. MoSPI's own survey shows the scale: counting free items adds about Rs 125 a month to the average rural person's spending, about 3%, and Rs 82 in cities.

The two numbers answer different questions. The 2% is how many people are below the line with the help they now get. The 80 crore is who the government decided should get that help. Nobody has published how many people the grain keeps above the line. What is known is its size: on average about Rs 125 a month for each rural person, about 3% of spending.

For someone living near Rs 68 a day, that can be the difference between falling below the line and staying above it, and a 10% fall in spending is enough to double the share below $3. So the 2% measures poverty with today's safety net in place, not poverty without it.

Who is still poor, and where?

Mostly in the east and centre of the country, in villages, and in families with small children.

At the $4.20 line, about 19% of rural Indians and 9% of urban Indians were poor in 2023-24. By state, about 47% of people in Jharkhand live below $4.20 a day, 40% in Chhattisgarh and 35% in Odisha. Bihar is at 27% and Uttar Pradesh at 26%. Kerala is at 5%, and Tamil Nadu, Delhi and Goa are lower still. The World Bank's India brief notes that a belt of states from Uttar Pradesh to West Bengal holds about 57% of India's poor.

Small children have it worst. In every state and union territory, children under six are more likely than the average resident to live in a household below $8.30 a day. In Kerala it is 75.8% of young children against 58.5% of everyone; in Maharashtra, 91% against 82%. A household with small children has more people to feed on the same earnings, so its spending per person is lower. The measure counts household spending per person. It cannot see what each child actually eats.

Are our neighbours doing better?

At the bottom line, India looks good. In 2024, 2.3% of Indians were below $3 a day, against about 5% in Bangladesh and Indonesia and 22% in Pakistan. At the top line, the order changes: 79% of Indians were below $8.30, against about 68% in Bangladesh and Indonesia. Viet Nam is in a different league, at 18%. India has a smaller share of its people below $3 than Bangladesh, Indonesia or Pakistan, and a larger share below $8.30 than Bangladesh, Indonesia or Viet Nam.

Poor Indians work as many hours as anyone else. Why are they still poor?

Not because they do not work. Paid employees in households below $3 a day worked about 44 hours a week in 2024, almost exactly as many as those above it. What differs is the work itself.

Workers in households below $8.30 a day are more often self-employed: 58% of them, against 45% of better-off workers. Self-employment in India usually means a small farm, a cart, a shop counter or a tailoring machine at home, without a salary. Nearly half of poor workers, 46%, are in farming, against a quarter of better-off workers. Over two decades, the share of poor households' workers who never finished secondary school fell from 87% to 56%.

That is a big change in schooling. It has not yet turned into a big change in jobs.

The report splits the growth in household welfare around the world between 2003 and 2024 into two parts: more people working, and more welfare per worker, meaning household spending or income divided by the number of workers. Almost all the gain came from the second. In poor Indian households, the share of working-age adults with a job slipped from 61% in 2003 to 58% in 2024. That leaves the question of pay, which these figures cannot answer directly. India's own wage data can.

Hasn't the economy's growth reached workers' pay?

Not much, for the workers most likely to be poor. Between 2014-15 and 2024-25, real GDP per person grew about 62%. Over the same ten years, the real daily wage of a male farm labourer grew about 9%. A rural non-farm labourer's wage grew about 2%, and a rural construction worker's fell slightly. Factory workers did a little better: about 5.5% in the nine years to 2023-24.

These come from two long-running government series. The Labour Bureau has recorded rural wage rates every month since 1998, and changed its occupation list in November 2013, so this chart starts in 2014-15, the first full year on the new list. The Annual Survey of Industries records what factories pay their workers. Each is adjusted for prices with a consumer price index on a single base, so no old and new series are spliced together. The economists Arindam Das and Jean Drèze, using the same rural series, found the same slow growth for farm labourers, about 0.8% a year, and called it "virtual stagnation".

How, then, did household spending rise? Partly because more people in each household started earning. The World Bank's India brief reports that employment grew by about 13.4 crore between 2017 and 2023, nearly twice as fast as the working-age population, mostly as rural women took up work, while real earnings stayed broadly flat. Free food and other transfers, and the change in the survey itself, are the other candidates. The data cannot say how much each contributed.

What would it take?

The World Bank ran a simple test. Suppose poor households had the jobs and the pay of better-off households. The share of Indians below $8.30 a day would fall from 79% to 57%.

That is a large drop. It is also a reminder of how far the line is: even then, more than half the country would remain below it. This is an accounting exercise, not a forecast or a policy.

Why doesn't India have an official poverty line any more?

Because the last attempts to draw one ended without agreement, and none has been made since.

India's poverty line began with calories. A 1979 task force set it at the spending level at which people typically ate 2,400 calories a day in villages and 2,100 in towns. The Tendulkar committee, reporting in 2009, did not build a new line from scratch: it took the existing urban line and changed how spending was measured against it. On that line, 21.9% of Indians were poor in 2011-12, down from 45.3% in 1993-94.

Then came September 2011. The Planning Commission told the Supreme Court that, at June 2011 prices, the poverty line was about Rs 32 a day in cities and Rs 26 in villages. The figure became shorthand for a line that seemed to describe destitution rather than poverty. A new expert group under C. Rangarajan, reporting in 2014, proposed a higher line, Rs 972 a month in villages and Rs 1,407 in cities, and found 29.5% of Indians poor in 2011-12.

That line was never adopted. A NITI Aayog task force reported in 2016 that it could not reach a consensus and suggested yet another committee.

None has been set since. So every Indian poverty number you read today, including the World Bank's, rests on somebody's choice of line.

Why did the World Bank's numbers jump in 2025?

Because the World Bank changed its yardstick. In June 2025 it switched its poverty lines to new international price data, collected in 2021. The old lines of $2.15, $3.65 and $6.85 became $3, $4.20 and $8.30.

For India the result looked contradictory. Measured on the same 2023-24 survey, extreme poverty rose from 1.0% to 2.6%, while poverty at the middle line fell from 19.9% to 16.1%. The share below the top line barely moved. Nobody's spending changed. The rulers did, and they did not all move by the same amount in rupees. That is worth remembering whenever a poverty number jumps overnight.

Did the new survey make poverty look smaller?

This is the heart of the current fight, and the honest answer is: probably somewhat, and nobody can yet say by how much.

The government did not publish its 2017-18 consumption survey, citing data-quality problems. Leaked findings reported in November 2019 suggested real spending per person had fallen 3.7% since 2011-12, with rural spending down 8.8%. When the next survey arrived, for 2022-23, it was redesigned. Households were visited three times instead of once, and asked about longer lists of items. Longer lists and a different visit schedule can change how much spending people recall. That makes the new numbers hard to compare with the old.

The World Bank's database records the break formally, grouping India's 2022-23 and 2023-24 surveys apart from 2011-12. Its October 2026 India brief calls the surveys "broadly comparable, despite several key improvements", then concedes the point that matters most.

There is a second change inside the World Bank's own numbers. Surveys ask people to remember what they spent, and the answer depends on the period they are asked about: the last week, the last month or the last year. In 2025 the World Bank recalculated its 2011-12 figures using the same recall periods as the new surveys. On the older price base, that alone lowered India's 2011-12 poverty rate from 22.9% to 16.2%. The old and new surveys became more alike, and the measured fall since 2011-12 became smaller than it had looked.

Then there are the camps. One says extreme poverty has all but ended. The economist Surjit Bhalla and two co-authors, in a 2022 International Monetary Fund working paper, put it at 0.8% in 2019, counting free food. NITI Aayog's chief executive said in 2024 that poverty was below 5%, and SBI Research gave similar figures.

The other camp says the fall is real but overstated. Himanshu, Peter Lanjouw and Philipp Schirmer, in a 2025 paper, try to correct for the survey change. On India's own Tendulkar and Rangarajan lines, they put 2022-23 poverty at about 11% to 26%, depending on the line and the model.

The World Bank has been on both sides. In 2022, two of its economists used the private Consumer Pyramids survey to estimate extreme poverty at 10.2% in 2019; Jean Drèze and Anmol Somanchi argued that survey left out many poor households, and the World Bank has since dropped those numbers. Even now its own documents disagree. The report estimates that about 10% of Indians will be below $4.20 in 2026, carrying the last survey forward with growth forecasts. Its India brief, using a different model, projects 14.2%.

Put together, the published estimates for 2022-23 run from about 2%, the World Bank's direct reading of the new survey at its old extreme line, to 26%. The spread mixes different lines and methods, so it is not a margin of error. Our read: the bridging estimates take the survey change more seriously, but they rest on models and on higher lines. What holds across all of them is that poverty fell a long way after 2011-12, though the withheld 2017-18 survey suggests it may have risen for a time in between. The exact size of the fall is not known.

Is India really one of the most equal countries in the world?

No, and the claim is a good example of a true number used wrongly.

In July 2025 a government press backgrounder called India "the fourth most equal country in the world", citing a Gini index of 25.5. The Gini runs from 0, everyone has the same, to 100, one person has everything. India's 25.5 came from spending surveys. Most countries it was ranked against measure income, which is always more unequal than spending, because the rich save much of theirs. Comparing the two is comparing different things. The World Bank's own report says India's figure rests on a spending survey that "may capture concentration at the top less fully".

To be fair to the claim, the World Bank has tried converting spending figures into what income figures would show. Even then, India does not join the group it calls highly unequal, with a Gini above 40. "Not among the most unequal" is defensible on these numbers. "The fourth most equal" is not. India's sheer size also flatters the world's figures: leave out India and China, and the share of the world's people living in highly unequal countries rises from about 19% to 30%.

Data on income and wealth tell a different story. Economists at the World Inequality Lab, combining national accounts, tax tabulations, rich lists and surveys, estimate that the richest 1% of Indians received 22.6% of national income in 2022-23.

They estimate that the same 1% held 40.1% of the country's wealth. Their method rests on assumptions of its own. But a survey that barely reaches the top cannot settle the question.

Within the two comparable new surveys, the spending of the poorest tenth of Indians grew about 13% in real terms in a year, the most of any group, while the richest surveyed tenth's fell about 3%. That is a measured gain for the poorest, though the World Bank says part of the jump between these two surveys may be better measurement. It is also a reminder of what the survey misses.

Since 1977, household spending in the national accounts, the government's own estimate of what the whole economy consumes, has grown about four and a half times in real terms per person. Average spending recorded in the household surveys has grown about two times. Some of that gap may be spending the surveys never see; the World Bank's India brief says the surveys may not fully capture the spending of wealthier households. Some is how the two are defined and valued. Nobody can split the gap precisely. But a survey that misses a growing share of the country's spending cannot be the last word on how equal India is.

Is being above the line the same as being well?

No. Spending is one measure of poverty, and the narrowest.

The World Bank's multidimensional measure, which adds schooling, drinking water, sanitation and electricity, counts 17.7% of Indians as poor on the 2022-23 survey, against 5% below $3 a day on the same survey. About 30% were deprived of adequate sanitation by the World Bank's measure. The fifth National Family Health Survey, in 2019-21, found 35.5% of children under five stunted, too short for their age. In 2022, about 31% of Indians faced financial hardship because of what they spent on health.

These are different measures for different years and cannot be added. But every one of them is larger than the share below the extreme poverty line.

The government's own count goes the same way. NITI Aayog estimates that 24.8 crore people escaped multidimensional poverty in the nine years to 2022-23. Both ends of that range are estimates, not survey readings.

What would settle the argument?

Two things, neither yet in place. The first is a consumption survey whose design stays the same long enough to compare like with like. The second is an official poverty line chosen openly and kept, so that the next fall can be measured against something agreed in advance. The 2011 Census still sets the survey's sampling weights, and the World Bank's brief warns that this probably understates how many Indians now live in cities.

Until then, the safest reading is the one that holds at every line. India has lifted its floor, fast. The floor is about Rs 68 a day.