Global extreme poverty falls below 1%
Will the World Bank report that fewer than 1% of the world's population lived in extreme poverty in a calendar year?
Chance by Q4 2031
<1%
Cumulative probability
- Latest forecast
| Year | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| 2026 | <1% | |||
| 2027 | <1% | <1% | <1% | <1% |
| 2028 | <1% | <1% | <1% | <1% |
| 2029 | <1% | <1% | <1% | <1% |
| 2030 | <1% | <1% | <1% | <1% |
| 2031 | <1% | <1% | <1% | <1% |
Model reasoning
The World Bank's latest figures show about one person in ten worldwide in extreme poverty (9.8% projected for 2026). Its own outlook is for slow progress or even a rise. Getting below 1% by 2030 would need a 90% fall in four years, more than three times faster than the fastest drop ever recorded. Only a radical change in how the Bank measures poverty, or a data error, could plausibly change this.
Where things stand
The World Bank's latest figures put global extreme poverty (below $3.00 a day, 2021 PPP) at about 10%, roughly ten times the 1% threshold.
- September 2026 update to the Poverty and Inequality Platform (PIP) (blog dated 22 Sep 2026): 824 million people, or 10.1%, were in extreme poverty in 2024. The 2026 nowcast is 9.8%. https://blogs.worldbank.org/en/opendata/september-2026-global-poverty-update-from-the-world-bank--one-in
- PIP API, release 20260922, world aggregate at the $3 line: 2024 is 0.1012 (actual), 2025 is 0.0995 (nowcast), 2026 is 0.0979 (nowcast). https://api.worldbank.org/pip/v1/pip-grp?country=WLD&year=all&povline=3&group_by=wb
- World Development Indicators, SI.POV.DDAY for World, last updated 8 Oct 2026: 2024 is 10.1%. https://api.worldbank.org/v2/country/WLD/indicator/SI.POV.DDAY?format=json
- The Bank's own outlook is stagnation or a rise, not collapse.
- The Data360 Atlas (May 2026) says that on current paths the global rate rises from about 10% today to about 15% by 2050. https://data360.worldbank.org/en/atlas/extreme-poverty/
- The Bank's September 2026 blog "Why an extreme poverty line is still needed" warns the number of extreme poor could rise over the coming decade. It also defends keeping the line roughly where it is. https://blogs.worldbank.org/en/opendata/why-an-extreme-poverty-line-is-still-needed
- March 2026 update (26 Mar 2026): 10.4% for 2024. The June 2025 switch to 2021 PPPs raised the measured rate rather than lowering it.
What would be needed, and the base rate
Nowcasts for unfinished years don't count. So by the end of 2031, the latest eligible year is effectively 2030. Getting from 9.8% (2026) to below 1% (2030) means a fall of about 90% in four years, or about 44% a year.
I computed the historical pace from the full World series in the indicators API (1981–2024). The largest one-year proportional fall was 12.9% (2012–13), and the largest four-year fall was 32.9% (2009–13). The required pace is more than three times the best single year ever recorded, sustained for four years. That would have to happen while the poorest countries (Nigeria, Ethiopia, the Democratic Republic of Congo and others) are stagnating, and their populations are growing about 2.5% a year.
Paths to YES
- Real economic progress. Effectively nil. Even explosive AI-driven growth would not raise measured consumption of the poorest 800 million people more than tenfold in proportional terms by 2030. It would also have to show up in Bank nowcasts, which are built from Global Economic Prospects growth projections with partial pass-through, by end-2031.
- Method change. New 2024 PPPs from the International Comparison Program are expected around 2027 (ICP global overview, June 2026: https://www.cepal.org/sites/default/files/2026-07/dia-1-11.20_icp-2024-cycle-global-overview_2026-06-23.pdf). The Bank typically adopts new PPPs about a year after release, which points to a possible new line around 2028. Past PPP and line revisions moved the global rate by a few points, not by 90%. The line is anchored to the national poverty lines of the poorest countries, and the Bank is publicly defending that anchor.
- A data error or a misreading by the resolver (for example, a regional figure below 1% taken as global). This is the main source of residual probability.
Paths to NO and scheduling
Scheduled releases will all show about 9–10%:
- the Poverty, Prosperity, and Planet Report 2026, due October 2026 according to publisher listings;
- PIP updates each March/April and September;
- World Development Indicators refreshes.
Forecast construction
I used a small constant per-quarter hazard to cover glitches, misresolution and unforeseen method shocks:
- 0.05% in Q4 2026;
- 0.02% per quarter in 2028, to cover a possible new line;
- 0.015% per quarter otherwise.
Accumulated, this gives about 0.37% by the end of 2031.
Strongest case against: a sudden transformative-AI growth shock. Even then, measurement lags, how the nowcasts are built, and the 2030 cutoff make a sub-1% global figure implausible. I would take far better than 99:1 odds against.
Resolution criteria
Question. By the last day of each quarter, has the World Bank reported that fewer than 1% of the world's population lived in extreme poverty in a calendar year?
Definitions.
- Extreme poverty: income or consumption below the World Bank's headline international poverty line in use at publication. This is $3.00 per person per day in 2021 purchasing-power-parity dollars following the June 2025 update. Official successor international poverty lines and PPP bases count; user-selected thresholds and superseded lines do not.
- World's population: the World Bank's population-weighted global poverty headcount ratio, including its imputations for economies without surveys, not an average of national rates or an aggregate restricted to surveyed countries.
- Calendar year: a completed January–December year. The Observation period convention applies, so the earliest eligible year is 2026. Official estimates for completed years count even if provisional, extrapolated, or labelled nowcasts; forecasts for unfinished or future years and conditional scenarios do not.
- Fewer than 1%: a published point estimate strictly below 1.0 percent (below 0.01 when expressed as a proportion). Use the most precise official numerical value available in the same release. A rounded value of 1.0% alone does not qualify; an explicit statement that the rate is below 1% does. A confidence interval extending below 1% is insufficient.
Resolves YES if an eligible World Bank publication or data release reports the qualifying global annual rate. An official revision for an eligible year can first trigger YES. Subsequent revisions do not reverse YES under the Resolution window convention. No attribution to AI is required.
Resolves NO at a horizon if no qualifying publication has appeared. National or regional results, national poverty lines, higher middle-income poverty lines, societal or multidimensional poverty measures, aspirational targets, and statements that poverty has been 'eradicated' without a qualifying numerical rate do not count. Publications reporting only years ending before the opening date do not count.
Resolving source. World Bank publications and data releases only: the Poverty and Inequality Platform global aggregate at the headline international poverty line; World Development Indicators series SI.POV.DDAY, geography World (WLD); the Poverty, Prosperity, and Planet report and successor reports; and World Bank poverty-update technical notes, official blog updates, or press releases reporting these estimates. A third-party quotation alone is insufficient. This is a narrower source requirement than Public confirmation (A2). Apply Resolution date (A3) and the Resolution window convention: use the dated publication or data-release date, not the reference year. An undated live database value requires an identifiable dated World Bank release establishing when it was published.