The UNECE Steering Group on Population and Housing Censuses, a group of 20 census experts representing 17 countries and international organizations, has released a study, New Frontiers for Censuses Beyond 2020, exploring the new challenges and opportunities for censuses in the region beyond the 2020 round.
The unprecedented circumstances of 2020 due to the global COVID-19 pandemic have accelerated a transformation in the demand for data that was already underway. The days of waiting a year or even two for cleaned, organized, tabulated census data to be released may be coming to an end, as researchers and policymakers grow accustomed to near-time data availability.
Information gathered through a census forms the foundation of many statistics that help us make informed decisions affecting all aspects of our lives: families, businesses, schools, hospitals, transport systems, neighbourhoods and cities.
A census is not just about counting how many people live in a country. Information collected in a census includes age, sex, and other key characteristics that allow countries to paint a picture of how different groups, with different needs, are spread out across the country.
In line with UNECE Recommendations which guide countries in what information to collect and how, most countries conduct a full census every ten years, although some conduct them more or less frequently, and others, such as Sweden, Finland and the Netherlands are already using alternative methods based on population registers that permit ongoing production of census information without requiring traditional data collection. With the aim of facilitating international comparisons, most censuses are conducted at a broadly similar point in time—the current round, known as the ‘2020 round’, includes censuses conducted in 2020 or 2021 in most UNECE countries, and 2022 in a few.
But what will happen after this round is completed? By 2030 our world may look very different. The ability to collect information from a wide array of sources, very rapidly, with minimal human intervention, is increasing all the time. Knowing this, citizens may rightly wonder why so much time and money should be spent on asking them to complete a questionnaire. If all the information they provide is already available somewhere else, why, people might ask, should they provide it again in the census? And why should governments authorize national statistical offices to spend what can amount to many millions of dollars on such an endeavour?
In some countries the level of public trust and willingness to complete the census, even though mandatory, is low and declining, presenting a significant challenge to those trying to produce accurate figures; while in others it is hard to reach those living in remote rural areas, those without a fixed address or who migrate regularly, and those with limited literacy, making finding and counting the entire population a long, slow and expensive process.
Methods employed for population and housing censuses have been evolving constantly over the past decades to deal with these increasing pressures. Adapting to changes in society and in information requirements; leveraging opportunities with new or different data sources such as registers and administrative files; and capitalizing on opportunities offered by innovations in enabling technologies, censuses have already begun a process of transformation from the traditional house-to-house enumeration that prevailed in the past. Italy and Israel are developing ways to derive their census data from systems of population registers, while the United Kingdom and Canada, where there are no population registers, are investing in ways to combine information from a wide range of existing administrative data sources to produce census data.
Recognizing the growing pace of change, the new analysis proposes that a careful review is required of some of the fundamental, even defining features of a census.
Traditional censuses rely on the idea of ‘simultaneity’ – meaning that everyone provides their information at, or referring to, one particular point in time, such as midnight on the census reference day, so that no-one is counted twice in two places nor missed out. The growing use of multiple alternative sources rather than direct questioning may make such simultaneity impossible. Tax files and border crossing records are collected for purposes other than statistics and have their own reference dates. Instead of attempting to adjust figures to arrive at a common reference date using complex models, the new analysis suggests that modern censuses may do better to accept this range of dates and relinquish the idea of simultaneity.
Increasing trends towards transnational migration and residence patterns, such as repeated cross-border movements, make the international coordination of censuses both more challenging and more important, to avoid either duplicating or skipping the parts of a population that move regularly between countries. A growing trend towards people having two or even more places that they call home brings into question another of the core ideas of censuses as we know them—that of ‘usual residence’. The Steering Group’s analysis suggests that future censuses, rather than striving to develop complex rules to determine which of a person’s many locations should be recorded as their usual residence, may instead need to apportion people in multiple places to reflect more accurately the way they divide their time and their use of services.
Other topics covered in the analysis include advances in the way that data quality is assessed; improved use of geospatial information in all stages, from census operations to data collection to dissemination of results; and renewed attention to relationships with the public to ensure their trust, such as communicating about the methods used to ensure privacy and confidentiality.
With country examples to illustrate the shifts in methods and perspectives, the analysis shows how the international coordination provided by UNECE can serve to aid countries in learning from one-another. An example from Poland demonstrates how geospatial information has been deeply integrated into census processes, offering a path for others to follow to increase their efficiency and improve the utility of census results. Another example, from the Netherlands, shows how information from a range of administrative registers (data collected for administrative purposes such as municipal population and real estate registers) can be integrated and cross-checked to provide a full census without having to conduct a survey.
We cannot know what the future will bring. But we can say with certainty that both data needs and methods for meeting those needs will change. Censuses will cross new frontiers, the Steering Group argues, but however and whenever they are conducted, censuses will remain at the core of informed decision-making into the 2030 round and beyond.