The relationship between Standard Occupational Classification 2010 and Standard Occupational Classification 2020

This guide describes the correspondences between the Standard Occupational Classifications (SOCs) 2010 and 2020 by presenting the analysis of two datasets dual-coded to SOC 2010 and SOC 2020.

For more information on the changes between SOC 2010 and SOC 2020, see Main areas of revision from SOC 2010 to SOC 2020.

The datasets used are the Labour Force Survey (LFS) for all quarters between January 2014 and September 2016 and the economically active 1% sample of the 2011 Census of Population for England and Wales.

Only those cases where a code has been recorded for both SOC 2010 and SOC 2020 are included in the analyses. Those cases where occupational information was deemed insufficient for coding to either of the classifications have been excluded.

The analysis is presented using eight sets of tables, showing the relationship between Standard Occupational Classifications (SOCs) 2010 and 2020. These are shown at the Major Group, Sub-major Group, Minor Group and Unit Group levels. The tables are split by gender and display the results separately for the Labour Force Survey (LFS) and 2011 Census data.

These tables show the percentage of the cases from a SOC 2010 group found in different groups using the SOC 2020 codes. So, for example, at the Unit Group level, the tables show what percentage of cases coded to a Unit Group in SOC 2010 have been coded to the same group at SOC 2020 and what proportion to other groups. This reveals the relationship between the new and old classifications.

These tables also show the percentage of the cases from a SOC 2020 group found in different groups using the SOC 2010 codes.

Perfect correspondences between the classificatory schemes exist only for those SOC codes for which 100% of cases have a single destination. For example, in Table 1, the SOC 2010 Unit Group 1136 (Information technology and telecommunications directors) has had 100% of its cases coded to SOC 2020 Unit Group 1137 (Information technology directors). So, in this Unit Group there is a perfect correspondence between SOC 2010 and SOC 2020.

When some cases from a SOC 2010 group are coded elsewhere using SOC 2020 there is no longer perfect correspondence. This can be seen in Table 2 below, where SOC2010 Unit Group 1132 (Marketing and sales directors) has been split into two groups in SOC 2020. Looking at the LFS data for men the tables shows that 99.6 per cent of people coded to 1132 (Marketing, sales and advertising directors), while 0.2 per cent were coded to 1133 (Public relations and communications directors).

Note that the percentages in Table 2 do not always sum to exactly 100%. This is because a small number of cases have been coded to other groups. These are not shown because where less than 1% of people were coded to a Unit Group, we have not presented them in the tables.

The tables for Minor, Sub-major and Major groups can be interpreted in exactly the same way. For instance, Table 3 below shows that among people coded to SOC 2010 Major Group 3, Associate Professional and Technical Occupations, 11 per cent were coded to Major Group 2 and 88 per cent were coded to Major Group 3 in SOC2020.

This user guide is based on the methodology adopted for the analysis of the relationship between SOC 2000 and SOC 1990, the OOSS User Guide 2000:22 by Jane Birch (ZIP, 883KB).

Similar health warnings to those in Section 4 of the document will apply to this user guide.

From the relationship and frequency tables, users should be able to:

• compare coding to the same classification from the two different sources and over time
• explore the relationship between Standard Occupational Classifications (SOCs) 2010 and 2020 within the same dataset
• compare the relationship between SOC 2010 and SOC 2020 across the datasets and over time

The information contained in this user guide may prove useful in the transition from Standard Occupational Classification (SOC) 2010 to SOC 2020. It gives an indication of the likely composition of a SOC 2020 code in terms of the coding of its cases in SOC 2010 and vice versa. However, the data are limited by various factors including diverse dual-coding methodologies across datasets and different sources of error. Users should be aware that any relationships shown for a code are provided as a guide only to the likely relationships they will find for their own data.

Contact

Classifications Unit Office for National Statistics Segensworth Road Fareham
HANTS
PO15 5RR
occupation.information@ons.gov.uk