Census 2021 takes place in England and Wales on Sunday 21 March, with the data collected providing a unique resource for understanding population characteristics and distributions across England and Wales. A census is also taking place in Northern Ireland on the same day, but this article focusses on England and Wales.
The census enables organisations to make decisions on planning and funding public services across the country, including transport, education and healthcare. For the first time, Census 2021 is designed to be ‘digital first’, however paper forms are also available, and census field officers will visit households who have not completed their census questionnaire after Census Day. They will not enter anyone’s house.
We caught up with Donna Clarke and Donna Viney at the Office for National Statistics (ONS) to discuss their roles in making spatial sense of the data collected, how geography helps unlock value from the census and the challenges that the Census 2021 data will help to address.
What is your role within ONS and how does this relate to Census 2021?
Donna Clarke (DC): My role is to manage a team to update the statistical output areas for census. This is the lowest geographical level at which census data are generated. These output areas and their boundaries are based on thresholds of population and households, a maximum target of 625 people and 250 households per output area. Obviously, there have been changes in population over the past 10 years since the output areas were done for the 2011 census, so some will need to be updated to reflect the underlying demographic changes.
Donna Viney (DV): I manage a team that are responsible for providing all the geography support to the census field operations and the Census Coverage Survey. We create bespoke boundary sets, bespoke mapping and calculate the amount of work we think would be involved in completing fieldwork in a particular part of the country. This depends on factors such as whether it is urban or rural and how easy it will be to get responses related to the demographics of the population that live there (for example are there blocks of flats with intercom systems, gated communities etc?). We factor in all of these things and then generate a boundary set that represents the work areas for each of the members of the field staff, trying to make it as equal as we can.
How does geography help unlock value from the census?
DV: Geography comes into all parts of the census cycle: collection, processing and outputs. In terms of field collection, it’s all about putting people into the right work areas and geography is the key component of that.
DC: By factoring in geography, we’re ensuring that the basic denominators, that is population and number of households, for each output area are the same, which allows for comparable and stable geographies to support statistical analysis and allows any outputs to be comparable to each other. Having baseline data to compare to is so important. The outputs of the census impact on just about everything we do; for example, they will feed into economics, COVID-19 response, and resource planning of catchment areas, plus other statistics we produce.
DV: On the processing side, it is about maintenance and updates to the small area geographies (output areas, super output areas and workplace zones). As we process the census information, we add geographical variables onto every single record so that later on it can be aggregated in whatever way it needs to be.
This helps us to identify areas around the country with similar characteristics, such as large student populations or elderly populations, and compare different parts of the country to see which have similar characteristics. Even the most basic statistics like producing a population count for a local authority has geography involvement!
Have your team(s) done anything innovative to improve or change the process of collecting and processing data compared to the 2011 census?
DV: This time round we’re using network analysis to calculate more accurate travel routes for field collection. In the past we’ve calculated this using ‘as the crow flies’ distances, but this doesn’t take into account where the roads actually are. This approach, using road networks and in some cases walking times should be more efficient.
We’ve also switched to using interactive online maps which can be used for some of the field operations on mobile devices. Previously, we’ve produced hundreds of thousands of paper maps, but now we have the technology to do these online, we’re saving on cost (e.g. paper and printing) and it’s a lot easier for us to make changes to the maps – pretty much right up to census day.
DC: Although we’re using the same tools to update the output areas, my team are trying to automate a lot of the process with R code and FME to try and streamline the work flow and remove the manual part of processing the data.
The census data comes into my team after it has been imputed and quality assured. We will receive data in small packets called Delivery Groups. It is broken down into 101 groups – so we have to do the processing 101 times. As the data comes in, we have roughly four to 10 days to turn it around and get it back to census, and given there are currently 181,408 output areas, there is an awful lot of processing to do. This is why the automation is so important so my team can work in parallel with each other and not delay the whole census process.
What challenges will the Census 2021, and geography, help to address?
DC: Understanding the geography and the people within whichever geographical area you’re looking at, at whatever resolution, is incredibly important – at a range of scales. Locally, it enables communities and local authorities to have funding to provide appropriate resources and forecast what services might be important to an area over the next decade. For example, the number of children in a catchment area will impact decisions around funding for schools in that area, and the size of the population in an area is also important to inform decisions around NHS services and electoral boundaries. But the data also has a lot of value at larger scales. For example, it can help us to understand how well we are meeting Sustainable Development Goals in certain areas and what we need to do to improve. We can’t do that without the census or the geographical understanding that accompanies it.
DV: Obviously up to date demographic data will be relevant for the current coronavirus pandemic and how we track our recovery from it. The data that is currently being used to inform the COVID-19 response is either old census data from 10 years ago or estimates people have calculated since then. So, the census will allow us to have a more accurate benchmark for the demographics of the country at the time of the pandemic.
It will also allow authorities and organisations, like the ONS, to judge how accurate the estimates and projections that they have been using, for example those related to the COVID-19 response, actually are, and it will inform their estimations over the next decade too.
Basic demographic data such as this also feeds into understanding the economy whether that be denominators for employment and unemployment rates but also in understanding how businesses may adapt over the coming years following the UK’s exit from the EU.
Find out more about Census 2021.
Donna Clarke is a Senior Spatial Researcher at ONS, supporting Census 2021 by updating the statistical output geographies for England and Wales. She has worked as a spatial researcher since 2000 using her knowledge for conservation ecology, bioenergy, international development, population modelling and UK geography. Donna has a PhD in quantitative ecology from Deakin University, Australia.
Donna Viney is a Senior Geospatial Researcher at ONS. She uses geography data and GIS to support field operations, analyse data, and provide mapping to social and business surveys (e.g. Census, Labour Force Survey, and Consumer Price Index) and other aspects of official statistics. At present she is working on network analysis, zoning methods, and using mobile devices in the field. Donna has a Masters in GIS from UNIGIS.