After initial talks of scrapping the 2031 census sparked strong backlash, the government has now reversed course. It has officially commissioned the Office for National Statistics to conduct a mandatory whole-population census for England and Wales in 2031.
For over a decade, successive governments have debated the future of the traditional census after 2021, viewing the survey as an outdated method for collecting population data. Instead, they have suggested piecing together administrative data as a more modern, cost effective alternative.
However, following the UK Statistics Authority’s recommendation last month, ministers have now given the ONS the go ahead to begin designing the shape and structure of the next survey.
Looking for cheaper alternatives
Every ten years, the ONS’s questionnaire based census provides a snapshot of the population in England and Wales. Over its 224-year history, it has significantly evolved into the in-depth survey it is today, shaped by increasingly sophisticated methods and shifting priorities.
Discussions about discontinuing the census began in 2014. In a letter to the National Statistician, the Government welcomed the recommendation of an online census for 2021. However, they also suggested that, with modern technology, significant changes in how the census is delivered could offer a more effective and cheaper population snapshot.
While the 2021 census was carried out as expected, questions about the long-term future of the survey persisted.
In the years that followed, continued talks have highlighted potential alternatives. With vast amounts of public data now being collected daily by government departments, the NHS, and other public bodies, the perceived need for a decennial survey has begun to diminish. It has been argued that the census could, instead, be replaced by collating readily available administrative data. This approach could then help bridge the gap between frequent public data releases and an in-depth population snapshot.
The risks of abolition
While some statisticians have supported the move towards administrative data, most do not share the same view.
In 2023, 60 academics and leading statisticians published an open letter urging the Office for National Statistics to conduct the census in 2031, arguing that ‘an increased reliance on administrative data risks an increasingly fragmented and inaccurate data landscape’.
They argued that administrative data can lead to misrepresentation and that it is vital as a supplement tool, but cannot replace the decennial survey.
The letter went on to state that census data forms a foundational part of researchers’ data infrastructure, crucial for accurate monitoring and as a baseline for other survey data.
The signatories, including academics from UCL and the University of Oxford, contend that, before any plan to abolish the census, an alternative must be implemented and proven to deliver comprehensive, accurate results of equivalent quality.
Room for improvement
Despite the general consensus that the government is not yet in a position to replace the census with alternative processes, academics and statisticians agree that there is certainly room for improvement.
As highlighted in the open letter, the 2021 decennial census led to undesirable outcomes due to the phrasing of certain questions. For the first time, it included a question on gender identity, which led to results inconsistent with data from gender clinics and surveys, and showed spurious outcomes across different demographics. It appears that those with limited English proficiency may have misunderstood the question and submitted inaccurate responses.
Additionally, the question on sex was also met with confusion, as the concepts of sex and gender were misunderstood by a range of respondents.
The signatories of the letter have claimed that ‘The ONS has a weak track record of user-engagement. Greater willingness to listen to expert opinion would have avoided undesirable outcomes’.
Although the Office for National Statistics has announced that they will launch a public consultation later this autumn to collect feedback on what should be included in the survey, experts argue that further consultation with experts is desperately needed to avoid shortfalls in future.
Final considerations
Census data remains incredibly relevant for understanding the population of England and Wales, supporting a vast array of research from sampling for survey data, to providing the benchmark for equalities monitoring. While regular public data sources provide timely and in-depth information on a range of topics, comprehensive census data remains essential for complementing these insights.
As the government moves forward with a Census for 2031, researchers, academics and statisticians have welcomed the decision. Indeed, with its wide range of data insights on population numbers, employment, education and health, it remains invaluable for researchers and for understanding changes across the country.
Statistics bodies in Scotland and Northern Ireland are now expected to follow suit, as they advise respective ministers.
Alice Sullivan, professor of sociology at UCL’s social research institute, has commented: “I am absolutely delighted that the census has been saved. National census data is a foundational part of our data infrastructure, as it furnishes the benchmark against which we judge whether other data sources, such as surveys, are representative.”
“Without a trusted picture of the population, we would have moved closer to a post-truth world of untestable ‘alternative facts’.”