Disclosure Control

**This information has been copied across from the SCROL website and some of it may be out-of-date. This will be updated in due course**

Disclosure Control

There is a legal obligation not to reveal information collected in confidence in the Census about individual people and households. In presenting very detailed results from the Census, protecting this data is of key importance. Disclosure of information in Census output is prevented by a combination of methods.

  1. Setting a target or average size for output areas OA's (50 households)

  2. Setting a minimum size of areas for key output (e.g. 20 households and 50 residents for CAS)

  3. Creating only one set of output areas (two sets of overlapping OAs could be "differenced" to create unintended below-threshold areas)

  4. Limiting the detail in classifications used in tables

  5. Record swapping before tabulation

  6. Small Cell Adjustment

Methods 1 to 4 are aimed at ensuring that there is only a limited number of cases in which all the households or persons in one of the categories of a variable in a table belong to a single category in another variable. When this happens, information can be disclosed from the table about those households or persons. For example, if there were only one Chinese person in an output area, a table for that output area tabulating ethnicity (with "Chinese" as a category) by employment status would reveal that person's employment status. Therefore, a further measure is needed so that no one can be certain that any such instance relates to actual individuals or households. That measure is Method 5 which completes the disclosure control package by swapping a small proportion of Census records. A swapped record is then tabulated in a different output area from where the data was collected but the aggregated statistics are not materially affected. This approach has been independently reviewed and endorsed by Dick Carter of Statistics Canada

A small number of tables have been subject to Method 6 whereby cells containing small numbers were adjusted randomly. These tables are those with OAs as place of workplace rather than place of residence. Method 5 is not effective for data on Workplace.

In addition to the above, one of the conditions of using Census data is that users will undertake not to attempt to obtain or derive information about a specific individual or household, nor to claim to have obtained or derived such information.

The means to protect confidentiality varies within the UK but all three UK Census Offices include the Method 5 in their package of measures.