Census microdata samples

In England and Wales, these are produced by the Office for National Statistics (ONS). As microdata samples contain individual-level data on census variables, they enable more detailed analyses than standard census output tables which report data at higher levels of aggregation. 

Microdata samples are available for seven successive censuses including the 1961, 1971, 1981, 1991, 2001, 2011, and 2021 Census. Person records in the microdata samples are not linked across censuses or with administrative records as is the case in the ONS LS, though they often contain information on a larger number of individuals in each census year. As such, when compared to the ONS LS, microdata samples are a more suitable choice for cross-sectional analyses of census data. 

What census microdata products are available?

There are several different census microdata products with each striking a balance between the detail of available data, access and security protocols.

Census microdata products are available at three levels of security:

• Public microdata samples

These are available to download directly via the UK Data Service (UKDS) and the ONS website and offer the lowest level of detail.

• Safeguarded microdata samples

These are available to download by registering with the UKDS and agreeing to an End User License agreement and offer a medium level of detail.

• Secure microdata samples

These can only be accessed by accredited researchers in a trusted research environment either via the UKDS or the Secure Research Service (SRS) and offer the highest level of detail. Access is subject to the submission and approval of a research project application.

Census microdata products with the same level of security can relate to one of two reference populations:

• Individual microdata samples

These include individual records sampled based on the population of individuals. They do not contain all person records within households represented in the sample.

• Household microdata samples

These include individual records sampled based on the population of households. They include all person records within sampled households.

Which census microdata products should I use for my research and how can I access them?

Depending on the level of security and the year of production, census microdata products can differ with respect to:

• The size of individual and household samples
• The sampling method
• The number of census variables available
• The detail of available categories within individual census variables
• The detail of geographical information available
• Geographical coverage

Prospective users need to evaluate these aspects of each microdata product against their research needs and objectives. The CeLSIUS Census Microdata Sample Finder tool displays the census microdata products that are available for the England and Wales population, relevant metadata, and information how these can be accessed.

Additional microdata products are available from the Scottish and Northern Irish censuses, which are respectively produced by National Records of Scotland (NRS) and the Northern Ireland Statistics and Research Agency (NISRA).

What types of analysis can census microdata samples be used for?

• Cross-sectional analysis of individual-level census data e.g. studies of associations between employment status and ethnicity; between long-term health conditions and housing tenure; or migration status and occupational class
• Variable combinations and multivariate analysis e.g. clustering analyses using combinations of census variables to explore intersections of multiple characteristics, and to identify socio-economic profiles or typologies of individuals or households.
• Sub-population and small-group analysis e.g. analysis of specific age-sex cohorts, occupational groups, religious groups or ethnic minorities that are underrepresented in other sources.
• Household context analysis e.g. studies of household composition, family structure, and of their associations with socio-economic outcomes
• Geographical comparisons e.g. analyses of geographical variation in socio-economic outcomes by region or local authority.