Providing faster and better estimates of the population

Image of queues at UK border terminal

The pandemic has highlighted the need for more frequent data about our population to understand the biggest challenges facing society. The ONS has been making great strides bringing together new sources of information to build the richest picture possible. As Pete Benton explains, this week we are publishing the improvements we have made, the limitations in the estimates, and our ambitions to refine the work in 2022.

The census is a brilliant source of data which gives us an in-depth understanding of our population like nothing else. The Census 2021 results, out next year, will deliver rich insights into all areas of the population at a time of enormous change as we lived through a pandemic and left the European Union.

But the census has one fundamental weakness – it only happens once every ten years. It provides a snapshot of a day in time but in this fast-paced modern world in which we live, we need a more timely and accurate understanding of our population on an on-going basis, as I outlined earlier this summer.

Alongside delivering a successful census collection in 2021, we have continued our work towards delivering better population and migration statistics. The pandemic has shown how important these are and this is a key priority to deliver.

Research into more timely population and migration statistics

Population statistics underpin a wide variety of other statistics, such as unemployment rates. They are vital in making a range of decisions and inform public debate. For example, the ability to forecast pensions, make decisions about local services (such as the number of school places or the provision of health services for an ageing population) and to support academic research.

We know users are interested in how migration patterns are changing and what this means for society and the economy. For example, understanding the impact on the UK labour market and public services, such as education and healthcare. This includes both the national picture and what is happening at a more detailed regional and local level.

Historically, our data sources have not allowed us to produce ‘snap’ estimates of the population. Estimates have traditionally been produced annually by using information from the census, and then rolling forward the estimates using information from registers for births and deaths, and migration surveys to produce the next year’s population estimate. But this means, the further away from each census you go, the more uncertain these estimates are.

This Friday we are also publishing a series of research articles which show our progress towards better population and migration estimates, using administrative data.

Using a range of sources from across government and the public sector, such as data from HM Revenue & Customs and the Higher Education Statistics Agency, we will be publishing an update on our admin-based population estimates (ABPEs), for the first time producing estimates for more than just a single year. Our time series will cover the years 2016 to 2020.

With each iteration of our ABPEs we are improving our methods. While these estimates are accurate at the national level, our work is focused on improving the accuracy of these estimates for certain population groups and at smaller geographic levels.

Some specific groups are more challenging to measure than others, for example students who are at an important life transition stage but where the admin data may struggle to keep up with rapid changes in their circumstances. Our work is also exploring the impact the pandemic has had on the data sources we are using to ensure continuity in the estimates we produce.

The benefits of administrative data to measure population and migration go beyond single estimates. They provide us with a unique opportunity to be more inclusive by measuring outcomes from migrants and all people. Next week we will share what insights we are learning from administrative data on how migrants transition from one visa to another and see if behaviours are changing since the pandemic.

Looking ahead

Over the next year we will continue to develop more timely and regular integrated admin-based population and migration estimates, while in late Spring 2022 the first results from Census 2021 will be out.

We will be making closer comparisons between census data and our ABPEs as we develop our estimates that relate to 2021, and as the census estimates become available. This will help us understand areas that could be improved.

Meanwhile, next month we’ll be publishing our feasibility research on Admin-based Labour Market Status (ABLMS).

There are obvious challenges for us to overcome as we continue on this journey towards ABPEs and transformed migration statistics, as we have highlighted, but also huge opportunities to meet the needs of our users better than ever before.

If you have feedback on any of our plans, please email

Pete Benton is Director of Transformation – Health, Population and Methods Group