Professor David Martin, Deputy Director of the UK Data Service, as part of our ongoing census series, discusses how the UK censuses have adapted to change over the past sixty years and some of the challenges we face in 2021/22.
All change
The process of constructing a census is one of adapting to continuous change, from deciding which questions are going to be asked to what technology will be used to collect, produce, and disseminate the data.
There’s a continuously changing set of tools and no modern census has been exactly the same as the one that went before: each has represented some significant technological advance.
Alongside the technological advances have been differences in societal attitudes and expectations of what should be asked in a census.
To take one example, in 1971 respondents were asked whether they had an inside toilet, as a lack of one was a signifier of household deprivation. In the 2021/22 censuses, this question would have been meaningless.
However, census questions can still help researchers understand the underlying issue of deprivation.
In the most recent censuses, questions about whether people have central heating and about levels of overcrowding of their housing address a similar concept, while recognising that what is perceived as deprivation has changed over time.
Developing census questions is thus a complex process, balancing continuity with previous censuses, against developing questions which measure current realities.
The 2021/22 UK censuses have variously included new questions on gender identity, sexual orientation and past service in the armed forces. These new questions also reflect the way that acceptability of asking about particular topics changes over time.
The census is always a trade-off between informing contemporary policy questions and the ability to make historical comparisons. This is one of those aspects that makes census-taking an endlessly fascinating exercise.
Geographical challenges
There are multiple factors which change in UK geography over time and need to be reflected in censuses, including:
- Reorganisation of local and regional government areas
- The changing allocation of land for residential use, such as greenfield and brownfield sites being redeveloped as housing and, less frequently, the clearance of former residential areas
- Changing patterns of where people choose to live, such as the huge increases in city centre living across the UK in the last two decades (albeit temporarily reversed by the global Covid-19 pandemic).
For studying how places change over the long term, small area census data are one of the best sources we have.
That said, the small areas themselves can be tricky: if they are updated to best match contemporary patterns of housing and demand for services, they will no longer match those from the previous census. For 2021, changes to output areas in England and Wales are generally being minimised, while Northern Ireland, by contrast, will have entirely new small areas.
Some of the changes we observe between censuses will therefore be down to changes in boundaries and others to actual changes in society. Managing this process of balancing continuity versus change is fundamental to many census design decisions.
No census is perfect (but most are very good!)
No matter how much thought, effort and testing has gone into question design and delivery, issues can still arise – it’s inevitable with the size of the exercise and the many functions for which it is used.
One of the fascinating dilemmas about the UK censuses is that most users want to use the data to understand society objectively, while society itself has a huge impact on shaping the data collection process and the questions.
Some issues with previous censuses have included particular kinds of under-enumeration or challenges related to public policy debate at the time, impacting on the ways people respond and answer questions.
The unpopularity of the ‘poll tax’ affected census response levels in 1991, while an outbreak of foot and mouth disease greatly increased the complexity of enumeration in 2001.
However, no recent census has had to face societal disruption on the scale of the pandemic and we are seeing marked differences in response rates between Scotland’s delayed enumeration in 2022 and the excellent rates seen in the rest of the UK in 2021.
While core demographic data should be of very high quality, the pandemic will also have differentially affected data about population groups such as students and topics such as economic activity and travel to work.
The UK Data Service uniquely offers access to data from all the UK censuses from 1961 to 2011 (and 2021/2022 once released) – a treasure trove for those interested in researching both continuity and change in British society.
Upcoming posts in the census series
- Changing definitions in the census in the past fifty years
- Looking to the future of the UK censuses
About the author
David Martin is a Professor of Geography at the University of Southampton and a Deputy Director of the UK Data Service.
He has been involved in several major data initiatives, including as Director of ESRC’s former Census Programme. His research in geographic information science led to development of the system of census output areas currently used in England, Wales and Northern Ireland. David was awarded an OBE in 2019 for services to Geography and Population Studies.