What Do Most People Do for Work? The Biggest Job Group in Each Country

What we looked at

Every country counts its workers by occupation using the same international list, ISCO-08. It has nine main groups: managers, professionals, technicians, clerical support, service and sales workers, skilled farm workers, craft and trades workers, plant and machine operators, and elementary occupations (basic jobs such as cleaners, labourers, street vendors and helpers). For each country we found which of the nine is the biggest, meaning the group with the most workers. AI exposure is our own score from 0 to 10 for how much of a country's work could be done or helped by AI. It is model-estimated, not measured.

Which countries are included. We use the 172 countries that report all nine groups, so the groups are comparable. Several large countries report only seven or eight groups and are left out: China, Japan, Canada, Ukraine, Saudi Arabia and Venezuela. Ghana is also left out because its stored employment figure looks wrong (see the biggest workforces). Percentages of workers below are shares of the workers in the 172 included countries.

The biggest job group, country by country

Three groups cover almost every country. Farming is biggest in 54 countries. In those, it is typically 41% of the workforce. Service and sales is biggest in 52 (typically 23%): shopkeepers, waiters, cashiers, hairdressers and carers. Professionals are biggest in 36, typically 25%. Elementary occupations are biggest in 21 countries, spread across several regions, such as Vietnam, the Philippines, Colombia, Peru, Morocco and Botswana. Only 9 countries have craft and trades, technicians or managers as the biggest group.

How the biggest group shapes AI exposure

The Average AI exposure column of the table above is the point. Countries where farming comes first average 3.42 on our exposure score. Where elementary jobs come first it is 3.57. Where service and sales comes first it is 4.13. And where professionals are the biggest group it is 4.99. That is a gap of 1.6 points between farming-led and professional-led countries.

This is the same story as our analyses of development and informal work, seen from a simpler angle: what the biggest slice of the workforce does.

One-job countries and mixed ones

Where one group dominates

In the most concentrated countries a single group employs most of the workforce, and it is farming in each case. Burundi has 82% of workers in farming. In these economies one kind of work, which our scores rate low, makes up most of the workforce.

Where no group dominates

At the other end, a country's biggest group is under a fifth of its workforce. Italy, Tuvalu and United Arab Emirates are examples. In these countries the job mix is spread across many groups, so no single sector sets the national picture.

What this means

When someone says "AI will change work in your country", ask what the biggest group of workers there does. In a farming-led country, the largest group is the one AI reaches least. In a professional-led country, the largest group is the one it reaches most. Both can be true of one planet at once, and they require different responses: one is about access to useful tools, the other about how fast office tasks change.

Limits to keep in mind

• Coverage. Only 172 of 205 countries with data are included. The 33 left out report fewer groups and include very large countries such as China and Japan, so worker shares are shares of the included countries only.
• Different years. Employment data are the latest available per country, from different years and sources, and some are modelled ILO estimates.
• Biggest is not most. A group can be biggest with under a fifth of the workforce. See the table of mixed countries above.
• Scores are model-estimated. They say how much of a job could be done with AI, not how much will be.

Spotted an error or know a better source? Report it and we will check it against the source data.