Modi Govt's Poverty Reduction Claims: What Changed in the World Bank's Numbers?

Official celebration of falling poverty rests on estimates, imputation, method changes, writes Subhash Chandra Garg.

The World Bank has become India’s biggest cheerleader. Curiously, the Modi government chose not to issue any press release for celebrating the reduction of poverty by 11.3 crore/3.9 crore in one year (was it too egregious for it?) The job was left to the media and Indian missions.

(Photo: Kamran Akhter/The Quint)

The World Bank has joined the Modi government in claiming that India has achieved a massive poverty reduction.

The World Bank India Poverty and Equity Brief, released in October 2026, asserted that the lower middle-income poverty rate ($4.20 per day in 2021 purchasing power parity or PPP) dropped from 57.7 percent in 2011 to 16.1 percent in 2023, lifting 49.6 crore Indians out of poverty. 

In terms of another poverty benchmark ($3 per day in 2021 PPP), the brief claimed that the poverty in India had declined from 27.1 percent in 2011 to 2.6 percent in 2023. This meant that only about 3.8 crore Indians are living in extreme poverty.

This claim has another very substantial hidden achievement.  

The World Bank had, in its poverty update released in June 2025 (page 13), had estimated India’s poverty at 57.68 percent ($4.20 in 2021 PPP) in 2011 and 23.89 percent in 2022. At $3 PPP per day, India’s poverty was estimated at 27.12 percent in 2011 and 5.25 percent in 2022. 

If you read the two World Bank briefs (October 2026 and June 2025) together, the poverty in India, in one year (2023), reduced by 7.8 percent at $4.20 per day and 2.7 percent at $3 PPP. In 2023 alone, India lifted, taking its population at 145 crore, 11.3 crore and 3.9 crore poor people out of poverty! Wow!

The World Bank has become India’s biggest cheerleader.

Curiously, the Modi government chose not to issue any press release for celebrating the reduction of poverty by 11.3 crore/3.9 crore in one year (was it too egregious for it?) The job was left to the media and Indian missions.

Is the World Bank poverty reduction claim credible? What is the secret potion which the World Bank has used to make such outrageous claims? 

World Bank Poverty Method

The World Bank makes global and country-specific poverty estimates using two key data and methods. 

First, the World Bank takes into account a country’s estimated consumption income, in local currency, derived from its national consumption surveys. In the case of India, for 2011, the World Bank considered the 68th Round of the quinquennial Household Consumer Survey. For 2022/2023, the Household Consumption Expenditure Survey (HCES) 2022-23 data have been taken into account.

To recall, the Modi government had rejected the quinquennial survey (the 75th round, July 2017 to June 2018) on quality considerations as its findings did not suit it. There are no national consumption/income survey estimates for India between 2011 and 2022.

Second, the World Bank establishes the PPP exchange rates from time to time to convert the national survey incomes/expenditures collected as part of international comparison programme (ICP, last 2021) to establish the PPP rates (India PPP rate Rs 20.728). Year-to-year PPP rates are determined by adjusting established PPP rate relative inflation differential (GDP deflator) between the US and comparator country. 

The percentage of population in a country with income/consumption expenditure estimates lower than the benchmark (there are three World Bank poverty benchmarks currently: $3.00, $4.20, and $8.30 PPP 2021, adjusted for inflation upto the year of consumption survey), yields the poverty rate for the country concerned. 

The World Bank is using changes/adjustments made in the HCES 2022-23 data, after first rejecting it, to artificially raise income/consumption expenditures in India. It's ridiculously low PPP exchange rate for India completes the story of the great poverty reduction illusion. 

World Bank’s Capitulation to India 

The World Bank publishes its regular Global Poverty Stocktake every seven years. Its last regular report Poverty, Prosperity and Planet Report 2024: Pathways Out of Polycrisis was published in October 2024. For this regular report, India’s HCES 2022-23 was examined by the World Bank but not considered good enough for working out India’s poverty estimates. 

The World Bank said (Box 1.3, page 56): "Before the India 2022/23 HCES Survey can be included in the Poverty and Inequality Platform (PIP), more work is necessary to understand the impact of various methodological changes in the 2022/23 survey, as well as the implications for the historical poverty series in India. One key element that has changed in the 2022/23 survey from previous rounds is the design of the questionnaire that collects information on household consumption. Analysis using previous surveys shows that this is an important change with important implications for the poverty rate." 

According to the 2024 report, the World Bank had estimated at $2.15 per day (2017 PPP), that India had 129 million poor, out of total 692 million global poor in 2022, or about 9 percent of India’s population.

Something mysterious happened in the next few months. 

Making an about-turn, the World Bank came out with a special 2025 Update on Poverty (unusual, off-cycle) in April 2025, accepting the HCES 2022-23 and drastically revised India's poverty estimates. 

The World Bank adopted all new definitions and methods of the HCES 2022/23, making considerable adjustments to its 2011 poverty estimates to make it 'comparable'. The standard methodology, called uniform recall period (URP), was replaced with a modified mixed recall period (MMRP), which the HCES had used. The World Bank also agreed to count free foodgrains, provided by the government, at market value for estimating consumption expenditures which made 2022 estimates quite high in comparison to 2011 when the food security programme did not exist. 

The World Bank did sound some cautionary notes. At page 12 of its Update, amongst others, it said that changes introduced in 2022/23 HCES "pose challenges for comparability" and "increased number" of consumption items surveyed (from 347 to 405) "potentially inflated reported consumption". 

The World Bank recalculated and revised its poverty estimates for 2011, at $2.15 per day 2017 PPP, from 28.94 crore people (22.86 percent of India’s population, at $2.15 a day, 2017 PPP) to 20.60 crore (16.2 percent). As many as 8.34 crore Indians got lifted out of poverty retrospectively in 2011 with the World Bank changing its poverty estimation formula. 

At the same time, the World Bank estimated that, at $3.00 per day (2021 PPP), India’s poverty rate in 2011 at 27.12 percent (34.44 crore people), which is being now used to determine poverty reduction during the Modi era, as according to this threshold, India’s poverty rate stands reduced to 2.6 percent in 2023, almost non-existent.

The World Bank’s magic trick made poverty disappear in India.

Strong PPP Rupee

The Indian rupee is getting beaten black and blue against the US dollar at the market exchange rate. It is, however, getting strengthened in PPP terms against the US dollar. What is going on?

Let us look at the rupee in PPP terms since 1990, when the PPP exchange rate system was introduced. India’s PPP exchange rate was Rs 5.5 in 1990. It nearly doubled to Rs 10.16 in PPP dollars in 2003. It further depreciated to Rs 17.29 to a PPP dollar in 2013, recording annual depreciation of as high as 5.46 percent during the Manmohan Singh era. 

The rupee did depreciate in PPP terms somewhat during the Modi era, until 2017 when the conversion rate reached Rs 20.46, yielding an annual inflation of 4.30 percent in the first four years of Modi 1.0. The rupee became rock solid in PPP terms thereafter. 

It was at Rs 20.47 in 2018 (almost unchanged, end of Modi 1.0). It turned out to be Rs 20.46 to a dollar in 2023 (end of Modi 2.0). Not a paisa depreciation during the entire five years of Modi 2.0. Magically, the rupee began appreciating in Modi 3.0. It rose to Rs 20.45 in 2024 and further to Rs 20.09 in 2025. 

The rupee, in PPP dollars, at Rs 20.09 in 2025 is stronger than Rs 20.46 in 2017. For eight years, the rupee in PPP has seen appreciation (indicating no inflation/depreciation). The rupee in market exchange terms has been subjected to a continuous depreciation onslaught all through Modi era. During Modi 1.0 and 2.0, it depreciated by 4.11 percent per year (from Rs 54.78 in 2013 to Rs 81.94 in 2023). In 2024, it depreciated by 3.53 percent (Rs 84.83), and, in 2025, by 4.59 percent (Rs 88.72).

The movement of the rupee against the US dollars in the opposite direction in two markets is totally inexplicable. The rupee strengthens against the US dollar in the World Bank PPP system on the ground that Indian inflation is lower than the US inflation. The rupee weakens against the inflation-adjusted US dollar in the market despite "better inflation record". 

There is something seriously wrong with the World Bank PPP methodology.

Poverty Reduction Illusion

The result of these two factors—the World Bank capitulating on accepting a liberal HCES 2022/23 consumption survey methodology which resulted in higher consumption (income) expenditure for India, and an artificially appreciating rupee in PPP terms—delivered great poverty reduction for India. 

The situation changes dramatically if one uses the market exchange rate to convert the consumption expenditure/ income into US dollar for determining how India stacks up according to the three World Bank poverty thresholds.

The $3.00 per day PPP poverty threshold is for low income countries (LICs). India is not an LIC and is currently classified as a lower middle income country (LMIC). Let us use the $4.20 poverty threshold which the World Bank uses for LMICs. 

For 2023, at PPP rate of Rs 20.46 to a dollar, $4.20 per capita per day threshold implies that an Indian person’s income/consumption value of Rs 85.9 a day or Rs 2,578 for a month is enough to consider him non-poor. For a family of four, consumption/income of Rs 10,312 will be considered enough for a living standard not amounting to poverty. It is easy to visualise that a family of four with income of about Rs 10,000 a month will still be living in poverty in India. 

If we take current market exchange rate of Rs 95 to a dollar, a family’s consumption/income would have to be Rs 47,880 per month for being considered above the poverty line at $4.20 per capita per day. This income can be reasonably considered satisfactory to afford good nutrition, clothes, housing, education for children, and taking care of health expenditures for a family in India.

The World Bank October 2026 release says that 16.1 percent of Indians (23.35 crore, considering 145 crore population) were poor at $4.20 per day at PPP. This is still quite a lot of people. The World Bank uses another poverty benchmark ($8.30 per capita per day in 2021 PPP) suitable for upper middle income countries (UMICs). According to this benchmark (Rs 20,378 income/ consumption in 2023 for a family of four), India had 80.75 percent of its population living below this poverty rate (World 46.6 percent, China 21.87 percent). 

The Indian family of four with income/consumption of Rs 20,378 per month is still way down income/consumption of Rs 47,880 per month required to meet $4.20 per day benchmark at market rate. As 80.75 percent of India’s population is poor at $8.30 per day PPP threshold, almost entire India is poor at $4.20 per day at market exchange rate. 

The World Bank poverty estimates, using defective PPP methodology and extra-liberal consumption estimates of HCES, are nothing short of an illusion. Poverty in India is still deeply entrenched. Let us address the real issue instead of celebrating illusory achievements.  

(Subhash Chandra Garg is the Chief Policy Advisor, SUBHANJALI, and Former Finance and Economic Affairs Secretary, Government of India. He's the author of books such as 'The $10 Trillion Dream Dented, 'We Also Make Policy', and 'Explanation and Commentary on Budget 2025-26'. This is an opinion piece, and the views expressed above are the author’s own. The Quint neither endorses nor is responsible for the same.)