Field entry, published 2026-11-03T00:00:00.000Z, updated 2026-11-03T00:00:00.000Z. Group: From ISCO-88 to ISCO-08, and revision

What changed from ISCO-88 to ISCO-08?

ISCO-08 replaced ISCO-88 as the ILO occupation classification, with adoption in December 2007 and endorsement in March 2008, and the change was a full revision rather than a light edit of labels. The hierarchy kept four levels, yet many groups were split, merged, moved, or rewritten so that tasks and skill profiles in the newer edition matched work as it had actually developed. A reader who treats an old label as identical to a new label will therefore miss real boundary changes, even where a familiar title survives.

Why did the older edition stop fitting the work it described?

By the late 1990s, several kinds of work had grown, shrunk, or changed character enough that the older boxes strained. Computing and communications work had become a large field in its own right, service work had diversified well beyond a few broad headings, and some production and clerical tasks had been reorganised by machines and software. Health and education work had also become more specialised, which made broad older groupings harder to use for workforce description. A revision offered a chance to redraw boundaries around tasks and duties as observed, instead of stretching old definitions to cover duties they were never written for.

What does the overall shape look like after the revision?

ISCO-08 organises all jobs into 436 unit groups, which sit inside 130 minor groups, 43 sub-major groups, and 10 major groups. Skill level and skill specialisation remain the organising ideas, with skill level describing complexity and range of tasks, and skill specialisation describing field of knowledge, tools, materials, and goods or services produced. The table below gives a quick sense of the scale, not a full map of every move.

Feature	Shape in ISCO-08
Levels	Four, from major group down to unit group
Major groups	Ten, including a separate armed forces group
Unit groups	436 detailed groups at the lowest level
Organising ideas	Skill level plus skill specialisation

Keeping four levels mattered for continuity, because statistical offices could still publish at broad levels while learning the finer detail. At the same time, the finer detail changed enough that a one to one match between old and new unit groups cannot be assumed.

Which broad moves matter most for a general reader?

One visible move was stronger recognition of professional and technical work linked to information, health, teaching, business, and culture. Groups in those areas gained clearer internal divisions, which helps a statistical release separate, for example, different kinds of specialist work instead of pooling them under a single heading. Managerial work was also reorganised with attention to the level at which decisions are made and the scope of responsibility, rather than treating every senior title as the same kind of management. Service and sales work gained more distinct treatment for personal service, personal care, and protective service, reflecting how varied that field had become. Craft, plant operation, and elementary boundaries also shifted where mechanisation changed tasks. None of these moves means that a job itself changed overnight. They mean that the classification lens changed, so the same set of tasks could be described more precisely.

What stayed recognisable, and what should not be assumed?

The idea of grouping jobs by similarity of tasks and duties survived intact, and many broad fields remain easy to recognise across the two editions. That continuity can mislead, though, because a shared word in a title does not prove a shared boundary. A group that kept its name may have gained duties, lost duties, or exchanged a subgroup with a neighbour. A group that changed its name may still cover much of the same ground, with a clearer label. For that reason, a careful reader checks the group definition and the list of tasks and duties, not the title alone, before drawing a comparison. ILO correspondence material maps the older structure to the newer one, recording splits and merges group by group.

How can a comparison avoid the most common mistake?

The most common mistake is to join two series at the revision year and read the jump as a change in employment. A sudden rise or fall at that point often reflects reclassification, not a sudden hiring wave or collapse. A safer approach is to state which edition each figure comes from, to use a published correspondence table for the groups in question, and to note where a clean match is impossible. Where a match is partial, a reader should describe the overlap in words and, if a chart is needed, mark the break clearly. At broad major group level, comparisons are usually less fragile than at unit group level, but even broad comparisons deserve a note about the edition change. Precision here is not pedantry. It protects every later claim about growth, decline, or shortage from resting on a boundary shift.

Reading ISCO-08 as a fresh map, rather than as ISCO-88 with new paint, gives the revision its proper weight. The older edition described the work of its own time, and the newer edition describes later work with finer tools. Used with that break in mind, the two editions together still tell a long story, as long as the join between them is handled openly.

Sources checked

1. ILO ISCO-08

Back to the From ISCO-88 to ISCO-08, and revision group or browse all occupations.