A new poverty line shifted the World Bank's poverty data. What changed and why?

$3 a day: A new poverty line has shifted the World Bank’s data on extreme poverty. What changed, and why? In June 2025, the World Bank increased its extreme poverty estimates by 125 million people. This doesn’t mean the world has gotten poorer: it reflects a new, higher International Poverty Line of $3 a day, up from $2.15. To track progress towards ending extreme poverty by 2030, the United Nations relies on the World Bank to estimate the share of people living below a certain threshold, called the International Poverty Line (IPL). In June 2025, the World Bank announced a major change to this line, raising it significantly, from $2.15 to $3 per day. This increase partly reflects inflation — a consequence of the World Bank using international dollars at 2021 prices, updated from 2017 prices. However, the IPL has also increased substantially, even after inflation adjustments. The poverty line has increased in real terms. And with it, so have the World Bank’s estimates of extreme poverty. 125 million people who would not have been counted as extremely poor before June are now included.1 This recent rise in the IPL is due to an aspect of the World Bank’s approach that some users of the data may not be aware of. Although it’s an international threshold, it is set to reflect national definitions of poverty typical among low-income countries. Several low-income countries recently raised their own poverty lines, pushing up the IPL. The higher estimates of extreme poverty reflect a higher poverty threshold, not that the world is poorer. The updated data on global incomes accompanying the new IPL shows that incomes among the world’s poorest are higher than previously estimated. Higher incomes but higher extreme poverty — this apparent paradox is unpacked in this article. We’ll explain in more detail how the International Poverty Line is defined, what has changed in the latest World Bank data, and what this means for our understanding of global poverty. See the latest World Bank data You can find all the updated World Bank data on our Poverty topic page. How the World Bank measures global extreme poverty To understand the new data and what it tells us about global poverty, we must understand how the International Poverty Line is set and used.2 Estimating global extreme poverty involves two parts, illustrated in this figure and discussed in more detail below. First, the World Bank estimates the global distribution of income (shown on the left) — that’s an estimate of how many people are living at each income level, summed across all countries. Second, it defines an extreme poverty threshold, the International Poverty Line (shown on the right). Together, these yield the World Bank’s estimates of extreme poverty: the share of the world’s population with incomes falling below the International Poverty Line. Thinking separately about these two halves helps explain the paradox of the June 2025 update. As we’ll see, both sides of the calculation have changed simultaneously. Download image How the World Bank estimates the global income distribution First, we’ll look at the left side of the calculation. You can think of the World Bank’s approach to estimating global incomes as being broken into three steps: 1) Collecting the raw ingredient: data from household surveys The main way statisticians find out about people’s incomes is by asking them. Almost all countries run household surveys to gather this information. This is what the World Bank’s global data relies on. Household surveys are run nationally, and their results can’t be directly compared at first. Countries ask different questions, capturing somewhat different definitions of income — as discussed in this footnote.3 Coverage also varies, with some countries conducting surveys annually, and others at intervals of several years. The survey data are also measured in different local currencies: households in India report in rupees, while households in Argentina report in pesos. 2) Estimating comparable national distributions To get a global perspective on incomes, the World Bank takes a series of steps to make this national survey data as comparable as possible.4 One important step is converting the local currency data into international dollars. This hypothetical currency adjusts for differences in the cost of living between countries. You can read more about this step in our dedicated article: What are international dollars? International dollars are used to compare incomes and purchasing power across countries and over time. Here, we explain how they’re calculated and why they’re used. What is important to know here is that this conversion is based on “purchasing power parity” rates (PPPs) — a measure of how much local currency you need in each country to buy the same value of goods and services. PPPs are based on detailed price data collected in “rounds” every few years. The World Bank updated to a new round — the 2021 PPPs — which triggered them to revise the International Poverty Line. With this harmonized survey data, the World Bank can estimate national distributions that allow incomes to be compared within and between countries worldwide. 3) Lining up the data to a common year The available country data refers to a mix of different years and, in some cases, can be several years out of date. To account for how incomes may have changed, the World Bank “lines up” each country’s distribution to a reference year using growth rates observed in national accounts data.5 The lined-up, comparable national distributions can then be summed up to calculate the global income distribution for a given year. Download image Once set, the International Poverty Line is applied to the global income distribution to estimate the share of people living in extreme poverty. Download image The summary of the World Bank’s approach above helps us understand the June 2025 update. Three statements summarize what has changed in the new data. 1) Inflation explains part of the rise Even if nothing else had changed, the switch from 2017 to 2021 international-$ would have led to higher values — for both global incomes and the International Poverty Line. As we explain in our article on international dollars, the value, or “purchasing power” of one international-$ is benchmarked to what one US dollar can buy in the United States. Prices in the US rose on average between 2017 and 2021 — about 11% in total.10 This means that one 2021 international dollar (the World Bank’s new units) is worth less than one 2017 international dollar (the old units). You need more 2021 international-$ to buy the same amount of goods and services. We’ll look at how incomes and the IPL have changed before and after the June 2025 update. But here, we first plot this inflation as a backdrop to put those changes in context. To have stayed the same and kept the same purchasing power, global incomes and the IPL would have had to increase by 11% in nominal terms. That is shown in green in the chart. An increase beyond this level, falling in the beige area above, indicates a rise in real terms (an increase in the quantity or quality of goods and services it could buy). Download image 2) The new data shows the world’s poorest are slightly better off As part of the update, the World Bank has changed its estimates of global incomes. In the new data, incomes among the world’s poorest are higher. This can be seen in the chart. The three arrows show how estimated incomes have changed before and after the update, across richer and poorer households. They show the percentage change in the “nominal” figures — without any inflation adjustment to account for the data having shifted from being measured in international-$ in 2017 prices to international-$ in 2021 prices. The arrow on the left shows the change in the income level that places someone in the world’s poorest tenth in 2024. This rose by 28% in nominal terms, from $2.34 per day (measured in 2017 international-$) to $3.00 per day (in 2021 international-$). Comparing this change to the green background, we see that incomes among the world’s poorest are higher not just because of inflation, but beyond inflation. They are higher in real terms. According to the new data, those at the bottom 10% of the global distribution can buy and consume more — 16% more — than the old data showed.11 The two other arrows show the nominal change at the global median and the top 10%. In contrast with incomes at the bottom 10%, these rose roughly in line with inflation: the data before and after the update agree about what households positioned here in the global distribution can afford.12 You can compare the data before and after the update in this chart. This increase in incomes among the world’s poorest is due to two factors: new survey data and new price data. With the update, the World Bank has added several new household surveys to its dataset. In particular, new data for India plays a big role. An improvement in the survey methodology used in India has shown incomes to be substantially higher than data based on earlier methods.13 Incomes at the bottom of the global distribution are also higher due to the change from 2017 to 2021 international dollars. As mentioned, these units use cross-country price data to adjust for differences in the cost of living. The new 2021 price data revises our picture of living standards in low-income countries, showing, on average, lower costs of living and consequently higher incomes.14 Download image 3) The International Poverty Line increased even more The International Poverty Line rose by 40%, from $2.15 to $3, far beyond inflation and income growth. Had the IPL only risen in line with inflation — an 11% increase from $2.15 to $2.38 — the World Bank’s new data on global incomes would show roughly 540 million people living in extreme poverty in 2024. That’s around a fifth lower than the World Bank’s estimates before the update — the effect of the higher incomes just discussed. With the revised IPL, extreme poverty estimates for 2024 stand at 817 million — around 50% higher than had the IPL maintained its purchasing power and only risen in line with inflation. The calculations behind these numbers are given in the footnote here.15 This large real-terms rise in the IPL is the joint outcome of the many steps involved in setting the IPL, described above. In part, it reflects the use of the new 2021 PPPs and some changes in the group of countries classified as “low-income”.16 However, most of the increase is explained by changes in the raw ingredient: the national poverty lines on which the IPL is based. Since the last revision of the International Poverty Line, several low-income countries have increased their poverty line substantially. This includes several countries with poverty lines falling close to the median value to which the IPL is anchored.17 As we saw, it’s common for countries to raise their poverty line as incomes increase. In the case of this group of countries, however, the step change is mostly a consequence of their adoption of an improved household survey methodology. For many of these countries, this resulted in higher recorded incomes, and national poverty lines were raised in line with this shift.18 Download image Although it’s the result of a consistently applied methodology, the June 2025 update has seen a significant, real-terms jump in the International Poverty Line. This change has shifted the goalposts on the first, and best-known of the UN’s Sustainable Development Goals: the goal to eradicate extreme poverty by 2030. Around 125 million more people are now counted as living in extreme poverty under this new, higher threshold. This increase does not mean that the world is poorer, though. According to the World Bank’s latest data, incomes among the world’s poorest are somewhat higher. These revisions — to the IPL, the data on global incomes, and the estimates of extreme poverty they produce — are substantial. However, they need to be placed in context. This context is provided by the chart here, which shows the share of people living below a range of poverty thresholds since 1990. Firstly, we must consider the recent revisions in the context of income changes over the past generation. The prevalence of extreme poverty decreased massively over this time, from one in every two or three people to one in ten. This is a huge change for our world. The revisions discussed in this article are small in the context of this large change over time. They also need to be put in the context of the huge range of incomes we see globally today — the great inequality between rich and poor countries. Whether measured at $2.15 or $3, the International Poverty Line is an extremely low benchmark. In our home country, the United Kingdom, national poverty is measured using a threshold ten times higher than the IPL, close to $30 per day. This is similar to many other rich countries. As shown in the chart, if we apply this threshold globally, the vast majority — 80% — of the world’s population would be counted as poor. Poverty, very clearly, does not end at the IPL. Although nine out of ten people today fall above this extreme definition of poverty, many of these people face living conditions that are barely imaginable for most people in rich countries: lacking access to clean water close to home; having to cook on solid fuel fires that create dangerous levels of indoor pollution; being unable to afford a diet that provides sufficient nutrients for themselves or their children. As global incomes grow, it’s right that we lift our ambitions about what should count as a morally acceptable minimum standard of living. From that perspective, the rise in the IPL can be seen as a positive step. But at the same time, we also need to focus on the world’s poorest. This is crucial because of the huge number of people stuck on such extremely low incomes — perhaps the most important thing shown in the chart above. The progress seen in reducing extreme poverty over the past decades has ended. Incomes are stagnating in many countries where the world’s poorest live. Because of this, we are very clearly on track to fail in eradicating extreme poverty by 2030. That was clear in the data before the June 2025 update, and remains clear now. Among the technical discussions needed to interpret the World Bank’s data correctly, we mustn’t lose sight of what this data shows: should current trends continue, hundreds of millions will remain in dire poverty for decades. That is surely one of the most important insights we could know about our world. Acknowledgments We thank Marwa Boukarim for her work designing and producing the article's visualizations and Max Roser, Edouard Mathieu, and Esteban Ortiz-Ospina for their valuable comments and feedback. Cite this work Our articles and data visualizations rely on work from many different people and organizations. When citing this article, please also cite the underlying data sources. This article can be cited as: Joe Hasell, Bertha Rohenkohl, and Pablo Arriagada (2025) - “$3 a day: A new poverty line has shifted the World Bank’s data on extreme poverty. What changed, and why?” Published online at OurWorldinData.org. Retrieved from: 'https://ourworldindata.org/new-international-poverty-line-3-dollars-per-day' [Online Resource] BibTeX citation @article{owid-new-international-poverty-line-3-dollars-per-day, author = {Joe Hasell and Bertha Rohenkohl and Pablo Arriagada}, title = {$3 a day: A new poverty line has shifted the World Bank’s data on extreme poverty. What changed, and why?}, journal = {Our World in Data}, year = {2025}, note = {https://ourworldindata.org/new-international-poverty-line-3-dollars-per-day} }