Field entry, published 2026-11-03T00:00:00.000Z, updated 2026-11-03T00:00:00.000Z. Group: The four levels of ISCO-08
How do the four levels of ISCO-08 nest?
| Major group | General (all major groups) |
|---|---|
| Skill level | General |
| Code ladder | general |
Classification only, not a job decision
This guide explains the classification structure only. It does not place any reader in an occupation, assign a code to a person, or act as a finder tool.
The four levels nest like a set of boxes placed one inside another, so every unit group belongs to one minor group, every minor group belongs to one sub-major group, and every sub-major group belongs to one major group. Nothing at a lower level is shared between two parents, and nothing floats free outside the level above it. That single parent rule is what makes the structure reliable for counting: work counted at a detailed level can be added up to the broader level above it without double counting the same occupation.
What does the nest look like from the top down?
Start at the widest view and move inward. A major group gathers a broad family of work. Inside it, sub-major groups split that family into smaller families. Minor groups split those families further, and unit groups give the finest split used in ISCO-08. Reading downward therefore means narrowing the field of work step by step, while reading upward means gathering related detailed groups into a larger total.
- Major group: the broadest level, written with one digit
- Sub-major group: the next level inward, written with two digits
- Minor group: the third level inward, written with three digits
- Unit group: the finest level, written with four digits
A useful check follows from that list. The first digit of a longer code always matches the major group above it, and each added digit narrows rather than restarting the classification. The code grows by detail, not by jumping to a different family.
Why does each level sit inside only one parent?
A single parent keeps the boundaries testable. If a minor group could belong to two sub-major groups, a statistic built from it would be ambiguous, because the same workers could be counted in two places at once. ISCO-08 avoids that ambiguity by design, so a researcher, a statistician, or a reader comparing figures can trace any detailed group upward along one clear path. The path also helps when two detailed groups look similar at first glance. Their parents may differ, and that difference in parentage is often the first clue that the work belongs in separate branches even when daily tasks overlap in places.
How do the counts change as you go deeper?
The number of groups grows at every step down, because each broader group is divided into several narrower ones. ISCO-08 contains 10 major groups, 43 sub-major groups, 130 minor groups, and 436 unit groups. The jump from 10 to 43 is modest, while the jump from 130 to 436 is much larger, which reflects how much finer the final split is. Not every branch divides by the same amount. Some broad families split into many detailed groups because the work inside them is varied, while other families need fewer splits. Uneven splitting is expected, and it does not mean that one branch is more important than another. It means only that the classification needed more detail there to keep unlike work apart.
What goes wrong when nesting is read backwards?
The most common mistake is treating a detailed group as if it were a broad family, or treating a broad group as if it described one occupation. A major group, for example 2 Professionals, covers a wide range of professional work, and no single job sits at that level alone. The level describes a family, not a person. The opposite mistake is to pick a unit group first and then assume its neighbours must be close relatives in everyday terms. Neighbours inside the same minor group are related by the classification logic, yet their workplaces, tools, and employers can still differ a great deal. Nesting tells you where a group sits, not how similar two jobs feel on a given day.
How can you practise reading the nest without assigning a code?
Pick a published group name you already know, then trace it upward and downward in words rather than by labelling anyone. Name its level, name the level above it, and ask what narrower groups sit below it. Then reverse the exercise: start broad, name one division at each step, and notice how the description becomes more specific each time. Keep the exercise at the level of groups and descriptions, and resist turning it into a verdict about a particular reader. The point of the practice is to learn the shape of the classification, so that later, when a real statistical table or a group definition is in front of you, the path from broad to detailed feels familiar rather than mysterious.
That familiarity is the real payoff of understanding nesting. Once the shape is clear, the code digits, the group counts, and the explanatory notes all start to fit into the same picture, because each of them describes the same hierarchy from a different angle.
Sources checked
Back to the The four levels of ISCO-08 group or browse all occupations.