To monitor global poverty, the World Bank compiles and harmonizes household survey data from around the world. When reporting aggregates, a population coverage threshold is used to determine whether the available survey data represents a sufficient share of the regional population. For a regional aggregate, survey data collected within three years of the reporting year need to account for more than half of the regional population. For reporting a global aggregate, the same population coverage rule applies together with an additional coverage requirement of more than half of the population in low- and lower-middle-income countries, which are home to most of the poor globally. When these thresholds are not met, trend-lines are either not shown, or shown as dotted lines.
With the June 2025 update, there is sufficient population coverage to report estimates until 2023 for the world and all regions, except Sub-Saharan Africa. The availability of survey data for Egypt helped achieve coverage for the Middle East and North Africa, which was not covered in the previous update (September 2024). Within Sub-Saharan Africa, coverage is particularly limited in West Africa due to the absence of recent data for Nigeria.
The new international poverty line is set at $3.00 using 2021 international dollars. Anyone living on less than $3.00 a day is considered to be living in extreme poverty. In 2022, about 838 million people lived in extreme poverty using this measure, which is an upward revision of approximately 125 million people from the previously reported estimate of 713 million people (from September 2024 using the $2.15 line and the 2017 PPPs).
With the June 2025 update, three main factors explain the revisions to the previously published poverty estimates. First, the adoption of the 2021 PPPs. Second, the updating of the international poverty line, which is explained both by changes in PPPs, but also updates to the underlying national poverty lines. Third, updates to the underlying household survey data, in particular new data for India.
The overall impact is a combination of these counteracting forces: The new PPPs, new poverty line, and new survey data together leads to an upward revision in extreme poverty by approximately 125 million people in 2022. 45 million fewer people are estimated to live in extreme poverty in South Asia, but all other regions experience upward revisions in the number of people below the international poverty line. The largest revision is in Sub-Saharan Africa, where an additional 111 million people are now estimated to be living in extreme poverty.
The revisions to countries’ poverty rates can come from several factors. The revision can be the direct result of the new price information implicit in the PPPs. For example, with the new PPPs, the price level might be higher than previously thought (so local currency has lower purchasing power than before), resulting in a higher poverty rate all else equal. In addition, with new PPPs, the global poverty lines are revised, which depends on both PPPs, but also the underlying national poverty lines. For example, with the update to 2021 PPPs published in June 2025, the international poverty line capturing extreme poverty is revised up by some 40 percent, which is significantly more than explained by the pure price movements.
Changes in PPPs and associated poverty lines can lead to revisions in countries’ poverty levels, but the trend is mostly unaffected. In other words, these kinds of revisions typically shift the entire time series up or down.
The other factor that leads to revisions in poverty estimates are new survey data that replace prior estimates. For example, the June 2025 update incorporates new survey data for India that leads to revisions in the poverty rate for the country.
Changes in the poverty lines do not change the day-to-day realities of people living in poverty. Nonetheless, these updated lines will better capture them and thereby enable more effective and impactful policy solutions to tackle poverty. A good analogy would be a doctor who has better information and tools to diagnose, and treat, a patient’s condition.
Finally, it should be highlighted that, when analyzing trends for a single country, the national poverty line is the appropriate standard to use. It captures the definition of poverty that is most relevant for this context, as well as how it should be updated over time to reflect changes in survey methodologies and evolving needs.
The revised international poverty line affects the current level of extreme poverty, but it generally does not affect recent and projected global trends. Therefore, the main conclusions from the Poverty, Prosperity, and Planet Report are still valid. If anything, some of the findings and recommendations are even more pertinent: global poverty is at an even higher level than previously estimated; poverty reduction has slowed to a near-standstill; the 2020s are set to be a lost decade; and an even larger share of extreme poverty persists in Sub-Saharan Africa.
With the revised data, the World Bank projects 9 percent of the global population will still be living in extreme poverty by 2030. Thus, the SDG target of eradicating extreme poverty, as well as the World Bank goal of 3 percent or less by 2030, are now even further out of reach.
Over the last several decades, India has changed its method of collecting consumption data. Historically, the surveys collected consumption data using the Uniform Reference Period (URP), in which survey respondents are asked for their consumption during the preceding 30-day period. With the 2011–12 round of the National Sample Survey (NSS), the Modified Mixed Reference Period (MMRP) was introduced (in addition to the URP instrument), which asks survey respondents for their consumption over the preceding 7 days for perishable items, 365 days for five low-frequency items, and 30 days for the remaining items.
To maintain comparability with historical data, the World Bank’s poverty estimates for India in the Poverty and Inequality Portal have thus far been based on consumption measures derived using the URP instrument. However, the new Household Consumption and Expenditure Survey (HCES) microdata for 2022–23, which is incorporated in the June 2025 update to the global poverty estimates, only collects data using the MMRP instrument. Therefore, with this update, data for India over the entire available time series has undergone adjustments.
The 2011–12 survey has both the URP and the MMRP aggregates and allows for a comparison of poverty based on the two approaches. To ensure comparability, the PIP estimates of 2011/12 have been revised to the MMRP, along with a number of methodological improvements. Nominal expenditure aggregates have been replaced by welfare aggregates that incorporate several methodological improvements, including an adjustment for differences in costs of living. These provide a revised trend between 2011-12 and 2022-23.
With these adjustments, extreme poverty, using an MMRP-based welfare aggregate, is estimated at 16.2 percent in 2011-12, which falls to 2.3 percent in 2022-23. This is a big change in the poverty rate estimated for India, but it is not unprecedented. Many countries around the world have improved their measurement of consumption, leading to higher measured levels of consumption, similar to India’s case (see Appendix to Chapter 1 in World Bank, 2024).
Pre-2011 estimates have been adjusted to reflect both of these changes. Noncomparable estimates from 2015–16 to 2021–22 previously available in thePoverty and Inequality Portal have been removed. More information on the changes to India's data can be found here and in thisbackground methodological paper.
