Occupation is a key variable in socio-economic research, used in a wide variety of studies, but its measurement is a major challenge. The national stocks of job titles are large with 10,000’s of job titles, they are unstructured with vague boundaries between job titles, and the stock has no fixed list but instead many entries and exits over time. Measuring occupations in a multi-country survey is even a larger challenge, because occupations with the same tasks have to be coded similarly across countries.

In 1958, the International Labour Organisation (ILO) of the United Nations developed the International Standard of Occupational Classification (ISCO) to harmonise the measurement of occupations, with revisions in 1968, 1988, and 2008. Today, ISCO has become the standard classification in many countries for their Labour Force Surveys or Censuses. ISCO-08, as was the case for its predecessors, defines a job as a set of work tasks and duties performed by one person. Jobs with the same set of main tasks and duties are aggregated into so-called 4-digit occupation units. On the basis of similarity in the tasks and duties performed, the units are grouped into 3- and 2-digit groups, which in turn on the basis of the skill level are grouped into 1-digit groups. ISCO distinguishes four skill levels, notably unskilled, semi-skilled, skilled and highly skilled, which are related to ISCED, the International Standard Classification of Education.

Most surveys use an open-ended survey question to measure occupations. The challenge relates to time-consuming and expensive office-coding. Alternatively, web surveys and CAPI surveys allow using a look-up database with occupational titles. Surveycodings provides a multilingual database of occupational titles that allow for self-identification of survey respondents, thereby tackling the challenge for  multi-country surveys  to classify job titles into ISCO-08 classification of occupations and to do so consistently across countries. The survey questions and answers for the occupation measurement, including their translations, can be found in the excel files below. The database is called the WISCO database, the World database of occupations, coded ISCO.

 

The PDF ‘D8.3 - Database of occupations’ explains the WISCO Database of ISCO Occupations and details the principles underlying the database.

The Excel file ‘WISCO database of occupations’ contains the source list of occupations and their codes, its translations, the structure of the search tree for online browsing of the database, a mapping table of the codes into the ISCO-08 classification codes, and a label set for all codes included in the database. The database is used in the live-search. Survey holders can use the database to prepare an API or search tree for the occupation question in their surveys.

Database versioning

For many years the WISCO database of occupations has been used in the WageIndicator web survey. The database will be updated continuously, as survey holders report new occupational titles, revised names for occupational titles, typing errors, and alike, and WageIndicator web visitors report occupational titles not yet included in the database. Database versioning can be traced by the date in the database name, referring to the date of updating. In 2019, SHARE - Survey of Health, Ageing and Retirement in Europe - has used the database for its wave 8 and SHARE team members have provided valuable suggestions. The WISCO database of occupations contains over 4,000 entries. Every effort has been made to ensure that the list of translated occupations is as accurate and complete as possible for each country and target language. However, users are advised to carry out checks on the database before use. These checks should be conducted for each country/language version for which the database will be used and involve the input of local survey teams wherever possible. Suggestions for improvements are highly appreciated.

Occupation > industry prediction

Many questionnaires have a question “Please write the main business activity of the organisation where you work”. The answer is commonly asked as an open text field, challenging the survey holder to code the response into an industry classification. Alternatively, in web surveys respondents can self-identify their industry from a look-up database. To facilitate self-identification from a database, an occupation>industry prediction has been developed, providing survey respondents with a limited set of industries, most likely for their occupation, followed by an option ‘other’, with the full look-up table shown in the next step. The prediction is based on the ISCO-08 and the NACEv2.0 classifications.

Based on pooled datasets from many surveys with both ISCO-08 and NACE2.0 variables, a set of linear probability models (LPM) could be estimated – one for each ISCO code.

 

The PDF ‘D8.11 - Occupation - industry prediction’ explains which datasets have been used and how the predictions have been calculated.

The Excel file ‘Occupation > industry prediction database’ consists of five tabs. RESULTS shows the coefficients and significance levels for all estimations. GROUPS details the grouping of the ISCO units. COUNTS ESTIMATES shows for which ISCO codes which level of detail was estimated. SURVEY MAPPING ISCO>NACE details the related NACE 4-digit codes per ISCO-08 4-digit occupation for routing in surveys. LABELS NACE 4dgt has the labels of industry codes.

ISCO 1958, 1968, 1988, 2008

We merged the available translations of ISCO 1958, 1968, 1988, 2008 into one file. ISCO 1958 and 1968 are available at 5-digit level and in the languages English, French, Spanish. ISCO 1988 is available at 4-digit level and in the languages English, French, Spanish, German. ISCO 2008 is available at 4-digit level and in 51 languages plus for another 2 languages not all entries have translations.

Measuring tasks in occupations

In analyses using the survey question “What is your occupation?”, it is assumed that the same occupational titles refer to the same work activities within and across countries. However, due to methodological limitations this assumption is hardly ever tested empirically. For this deliverable a TASKS DATABASE has been developed to measure work activities – called tasks - within occupations. The starting point for this database was the task descriptions provided by the ILO for occupations at the most detailed ISCO-08 4-digit occupational groups in its ISCO-08 coding index (2012). The database has been supplemented with task descriptions from national coding indexes and with translated task descriptions from the WageIndicator web survey on Salaries. All tasks have been checked for correspondence with the ILO list. If needed, translations have been added. The resulting database consists of 3,264 tasks for 427 ISCO-08 4-digit occupations in 16 languages: Albanian, Bosnian, Bulgarian, Czech, Dutch, Estonian, French, Indonesian, Norwegian, Russian, Spanish and Turkish, plus two language duplicates with minor differences, notably British and American English, and Portuguese for Portugal and for Brazil. Note that the Norwegian translation is of moderate quality.

 

The PDF ‘D8.6 - Job task collector’ explains how the database was compiled, based on the tasks descriptions in ILO's International Standard Classification of Occupations ISCO-08 Volume 1.

The Excel file ‘Job tasks database by occupation’ consists of four tabs. The tab CODESET consists of the codes, master labels and translations of all tasks. The tab STRUCTURE identifies the routing from the ISCO-08 4-digit occupational code to the related tasks. The tab MAPPING includes the cross-over tables of the tasks to the 4-digit ISCO-08 codes. The tab SURVEY_QUESTIONS has translations of the survey question “In your current job, how often do you carry out the following tasks?” and the answers on a five-point scale from never to daily.