Field entry, published 2026-11-03, updated 2026-11-03. Group: ISCO-08 in statistics and surveys

How do labour force surveys use ISCO-08?

A labour force survey uses ISCO-08 after people have described their work, when trained coders or coding systems place each described job into the occupation group whose tasks and duties match it most closely. The survey itself does not ask a respondent to pick a code from a list, and the published tables do not reproduce individual job titles. Instead, the survey collects words about work, and ISCO-08 supplies the shared structure that turns those words into comparable groups. That separation matters, because the quality of a labour force table depends as much on how answers were recorded and coded as on how many people were interviewed.

What does a survey ask before any code is chosen?

A labour force questionnaire usually begins with the respondent in plain language, asking what kind of work was done in a reference period, what the main tasks were, and sometimes what tools, materials or outputs were involved. Job title alone is rarely treated as enough, since the same title can cover different work in different workplaces, and different titles can cover similar work. Interviewers are therefore encouraged to capture a short description of duties, not just a label. A description that mentions supervising others, repairing equipment, preparing food, teaching a class or keeping records gives a coder far more to work with than a single fashionable title.

The order of questions also helps. Employment status, industry and hours are often collected alongside occupation, so that later tables can cross occupation with other characteristics without mixing the concepts together. Occupation describes the work done, industry describes the activity of the employer or business, and keeping those two ideas apart prevents a common reading error. A person can work in a hospital without holding a health occupation, for example, when the job itself is administrative or maintenance work, and ISCO-08 coding should follow the job, not the building.

How does coding turn a description into a group?

Once answers are collected, each job description is matched against ISCO-08 group descriptions. Coders look for the main tasks and duties, the level of skill implied by the complexity and range of those tasks, and the field of specialisation where relevant. A broad major group is the first sorting step, and then narrower sub-major, minor and unit groups refine the placement as the description allows. When a description is vague, a coder may only be able to place the job at a broader level, and statistical offices have rules for handling such cases rather than inventing detail that the answer did not contain.

Some offices code by hand, some use assisted systems that suggest groups based on text, and some combine both approaches, with human review for uncertain cases. Whichever route is used, consistency checks matter. Two coders given the same description should reach the same group wherever possible, and offices test that agreement by recoding samples and reviewing disagreements.

Why does the same job sometimes land in different groups over time?

Coding guidance can evolve, interview wording can change, and the work itself can change, so a labour force series needs documentation alongside its tables. If a survey changes the prompt that elicits job descriptions, coders may receive richer or thinner material, and the resulting groups may shift even when employment has not changed. Good survey documentation therefore records the questionnaire wording, the coding rules in force, and any training or system change that could affect comparability. Readers who skip that documentation risk mistaking a methods change for a labour market change.

Another source of movement is the boundary between groups. Many real jobs blend tasks from neighbouring groups, and a small change in emphasis, such as a new duty becoming the main task, can move a job across a boundary. It reflects the fact that classifications draw lines through continuous working life, and borderline cases will always require judgement. Transparency about those judgements, rather than pretending they do not exist, is what lets users judge a table sensibly.

What should a reader check before quoting a survey table?

A careful reader starts by checking which classification version the table uses, because a table labelled only with occupation groups leaves the edition unclear. Next, the reader checks the level of detail, since a major group table hides differences that a unit group table would reveal, and a detailed table may rest on thinner descriptions. The reader should also look for notes on coding, on how vague answers were handled, and on whether the survey covers the whole population or a defined subset. Each of those notes changes what a comparison can support.

It also helps to read occupation tables beside industry, hours and employment status tables, rather than treating an occupation figure as a complete portrait of work. ISCO-08 answers one question well, namely what kind of work people do, grouped by tasks and duties. It does not by itself explain pay, security, productivity or job quality, and a labour force survey usually needs other variables to address those questions. When occupation is combined thoughtfully with those other variables, the result is a clearer picture of a labour market. When it is quoted alone, without its coding notes, the picture can look more precise than the underlying descriptions allow.

Sources checked

1. ILO: ISCO-08

Back to the ISCO-08 in statistics and surveys group or browse all occupations.