The Global Jobs Index provides harmonized, internationally comparable estimates of the availability and quality of employment across countries and over time. This page describes the data sources, indicator definitions, modeling approach, and limitations.
The Authors


This methodology was developed by Impacc and the Kiel Institute for World Economy.
About the Kiel Institute for the World Economy
The Kiel Institute for the World Economy is Germany's leading research institute on global economic affairs and one of the most renowned institutions in Europe and worldwide in this field. It combines top-quality research comparable to that of leading universities with a visible and lasting influence on national and international economic policy. https://www.kielinstitut.de/
Tobias Heidland
Tobias Heidland is professor of economics and Research Director for international development at the Kiel Institute for the World Economy.
About Impacc
Impacc seeks to establish a new form of global aid – one that empowers rather than provides. The non-profit organization converts donations into investments in African startups, aiming to create sustainable jobs and increase local self-reliance. It invests in local founders with ideas for local markets, providing financial, technical, and operational support. The model combines elements of traditional development cooperation with the principles of venture capital. Founded by a team led by former Welthungerhilfe CEO Till Wahnbaeck, Impacc is based in Hamburg, Germany, with teams also operating in Kenya and Ethiopia. https://impacc.org
The Dataset:
The Global Jobs Index is an annual panel data set. It starts in 2000, tracks annual developments up to 2025, and projects them to 2060. There are three identifiers: country (country name, ISO3C), year, and scenario. The main scenario assumes a moderate effect of artificial intelligence; two further scenarios show how generative AI could change the picture (Optimistic, Pessimistic), and there is a baseline comparison without AI. All four are identical up to 2025.
Data Sources:
The Global Jobs Index combines several data sources. The most important are data from the International Labor Organization (ILO). National labor force surveys are the basis for the ILO’s estimates of how many workers live in households above or below given income levels, and for its data on hourly pay, which show how many workers earn less than two-thirds of the national median wage. Further ILO statistics give the size of the labor force and the number of unemployed. GDP per person comes from the World Bank’s World Development Indicators (2000–2025, adjusted for purchasing power). The World Bank’s country classification defines which countries are covered: low- and middle-income countries with at least 1 million people of working age, currently 97. For the years after 2025, the GJI uses the Shared Socio-Economic Pathways (SSP, 2025 release), established scenarios for population and economic growth that also underlie the IPCC climate reports. Population projections come from the Wittgenstein Centre. For the AI scenarios we build on Gmyrek, Viollaz, and Winkler (2026).
Definitions:
Job categories:
- Jobs below the poverty line: Workers whose household lives on less than $3.00 per person per day (adjusted for purchasing power). This is the World Bank’s international extreme poverty line. These people work, but they can barely afford the bare necessities.
- Gig jobs: Workers whose households are above the extreme poverty line, but who do not have a big job. Many of them work in informal, irregular, or low-paid jobs. Because the absolute poverty line is the same in every country, this allows comparisons across countries.
- Big jobs: Workers who earn at least two-thirds of the national median hourly wage in their main job and live in a household above $8.30 per person per day (adjusted for purchasing power). Employees and the self-employed can hold a big job whereas unpaid family workers cannot. The two-thirds threshold is the standard definition of low pay used by organizations such as ILO. It is a comparison within a country. The additional $8.30 line ensures that a big job also means a minimum standard of living that is comparable across countries.
The Jobs Gap (the employment gap):
- The Jobs Gap comprises members of the labor force who have neither a Big Job nor a Gig Job: people who are unemployed, and people who work but whose household lives on less than $3.00 per person per day, adjusted for purchasing power. People outside the labor force, including those who have given up looking for work, are not counted. Big Jobs, Gig Jobs and the Jobs Gap together make up the whole labor force.
The Model:
- The various data sources are integrated into a common database in panel data form (i.e., country - year - scenario). Each input data set is restructured to fit into the structure of the overall data set. Country names and other identifiers are standardized so that the data sets can be combined. The imports, preparation, and processing are written entirely in program code in the Stata statistical software, so that updates are easy to make and run largely automatically. This ensures complete replicability.
- This database can then be used to determine the average correlation between job growth, economic development, and population growth. For the past, we can use official survey data and estimates from the ILO. For the future, we must make assumptions about how key factors will develop. For this purpose, we use the SSP scenarios (2025 release), specifically SSP3. When projecting into the future, we use as few variables as possible because we have to make our own assumptions for each variable’s future development. The two core variables are GDP per person and the size of the male and female working-age population. Unemployment is held at each country’s 2025 rate. All other variables are implicitly assumed to remain unchanged, a necessary assumption to reduce complexity.
- The AI scenarios. Generative AI can raise growth, but it can also weaken the link between growth and good jobs. A country’s exposure depends on its mix of occupations, which shifts as it grows richer. In the “Moderate AI” scenario, AI adds modest extra growth, and three-quarters of the usual link between growth and good jobs survives in AI-exposed jobs. In the “Optimistic” scenario (previously “Brave New World”), AI adds strong extra growth, with the same link. In the “Pessimistic” scenario (previously “Doomsday”), AI adds modest growth, but only a tenth of the link between growth and jobs survives in AI-exposed jobs.
Data Availability
The Global Jobs Index builds on surveys by national statistical offices, which collect information on employment and wages. For 2000–2025, the figures are ILO estimates based on these surveys. Pay surveys are not conducted by every country every year. Between two surveys we interpolate. Before the first and after the last, the share of workers above two-thirds of the median wage moves with GDP per person based on parameters estimated from the 2000–2025 relationship. Some countries, including China, have never published a suitable pay survey. For them, this share is estimated from comparable countries in the same region, and the visualization marks them as having no ground-truth data. From 2026 on, all figures are projections.
The Visualizations:

Interactive Map: From extremely alarming to very good: the map shows the jobs gap as a percentage of the labor force. The "Play" button shows how it changes over time. Hovering over a country displays the figures. Clicking a country shows the trend from 2000 to 2060.

Leaderboard Who leads the list, who is last? And how do the positions change over time? The ranking list - sorted either by absolute numbers or by percentage values - uses the color classification from dark red (“extremely alarming”) to light green (“very good”).

Countries Compared Here you can compare countries and regions relative to size or in absolute numbers.

Individual Country While many indices ignore the growing gig sector or fail to adequately capture it, the Global Jobs Index explicitly integrates it as a separate category. Here, jobs are divided into big jobs (which offer relative financial security) and gig jobs (which are above the extreme poverty line).

AI Net Effect See how different AI scenarios—from optimistic to pessimistic—affect jobs and the jobs gap in different regions.
Download the dataset
Please cite the current dataset as: “Heidland, T. and Wahnbaeck, T. (2026). Global Jobs Index, 2026 Release.” The data are available under CC BY 4.0. We would be happy to hear how you use them. The 2025 dataset remains available separately as a historical release.