FOI reference: FOI-2026-3818

You asked

In February 2025, in response to request FOI-2025-2589, ONS explained that estimates of the number of women in individual construction trades are suppressed because the numbers are small enough to risk identifying individuals. I would like to establish the current position and the rule being applied.

Please provide the following.

1. For the most recent period available, estimates of the number of people employed in each SOC 2020 unit group within sub-major group 53, skilled construction and building trades, broken down by sex, from the Annual Population Survey or whichever source ONS considers most appropriate.

2. Where any cell at point 1 is suppressed, the suppression rule applied, including the threshold value and whether the threshold is applied to the unweighted sample count, to the weighted estimate, or to both.

3. The number of unit groups within sub-major group 53 for which a female estimate can be published, and the number for which it cannot, for the most recent period.

4. Whether ONS holds, published or unpublished, an estimate of the total number of women employed across sub-major group 53 as a whole, which may sit above the suppression threshold even where the individual trades do not.

Point 4 is the part I am most interested in. If the detail cannot be released, an aggregate that can be published would still be of use.

I would like to receive the response by email to this address, and my contact details are below in case you need to discuss the request.

We said

Thank you for your request regarding estimates for women in skilled construction and building trades and the suppression of Annual Population Survey (APS) estimates. 

For the APS, all statistics based on an unweighted count of less than 3 are suppressed for disclosure reasons. If the unweighted count is 3 or more, we would not suppress the numbers on disclosure grounds.

In the previous request, for each individual four-digit Standard Occupational Classification (SOC) code that corresponded to the occupations requested, we would have only interviewed 2, 1 or 0 women who were employed in each occupation. 

Each quarter, we publish a report on our Nomis website showing APS estimates of the number of people employed in each SOC 2020 occupation, broken down by sex. The report presents the full SOC hierarchy, including estimates at the four-, three-, two- and one-digit levels where they meet the relevant disclosure and quality criteria. The link will update to show the latest available data and provides access to historical and regional breakdowns.

Looking at the APS data for the 12-month period July 2025 to June 2026, for many of the four-digit SOC codes for construction occupations, it is still the case that the estimates for the number of women are suppressed on disclosure grounds. We are unable to release these suppressed counts under FOI, as section 39 of the Statistics and Registration Service Act 2007 strictly prohibits the disclosure of information that identifies individuals. As we are prohibited by law from publishing statistics in which individuals or individual businesses can be identified, Section 44(1) of the Freedom of Information Act 2000 (FOIA) applies.

In addition, Nomis applies a further quality suppression to small cells where there are insufficient observations to calculate a meaningful quality measure. This typically affects estimates based on fewer than 10 responses, although the precise position may vary according to the characteristics of the sample. In the case of July 2025 to June 2026, the only statistic that Nomis has suppressed on quality rather than disclosure grounds is the estimate for women in occupation 5330 -- Construction and building trades supervisors. While Nomis suppresses this on quality grounds, we can provide this information as a response to a FOI request. The estimated figure for this group is 2,800; however, caution should be taken when using this figure because of the concern around the quality of the estimate. 

As shown in the linked Nomis table, in addition to estimates of the number of women in each four-digit occupation, estimates are provided at both the three-digit and two-digit level of classification, with an overall estimate of 18,700 for women across all the Skilled construction and building trades. 

In summary, APS estimates based on fewer than three unweighted observations are suppressed for disclosure reasons. For July 2025 to June 2026, female estimates for three of the 12 SOC 2020 unit groups within sub-major group 53 are published on Nomis. Eight are suppressed because they are based on fewer than three unweighted observations, while the estimate for SOC 5330 is suppressed on Nomis for quality reasons but has been provided in this response with an appropriate caution. At the aggregate two-digit level, the APS estimates for women employed in Skilled construction and building trades is 18,700.