Projecting the possible future size and composition of the UK population is of importance for national level decision-making and policy. In this blog James Robards explains some of the considerations and challenges in developing national population projections, particularly at a point in time when the demographic impacts arising from the Covid-19 (Coronavirus) pandemic continue to unfold.
National population projections (NPPs) published today tell us about the potential future size of the population of the UK and its constituent countries. This is an interim set of projections to reflect the short interval between the 2020-based principal projection and subsequent projections which will incorporate data from Census 2021.
Using our most up-to-date population estimates and assumptions about future fertility, migration and mortality we have been able to give an indication of the future size and composition of the population. This shows projected UK population growth is slower than in the 2018-based projections. Alongside this we have used these projections to update our past and projected period and cohort lifetables which show boys and girls born in 2045 are projected to live longer than 90 years on average.
Ongoing and extensive user feedback has continued to highlight the wide use and importance of 2020-based NPPs for a number of long-term planning and policy making areas including pensions and government spending. It is important that projections are not viewed as predictions or forecasts, but as an indication of the future if past demographic trends continue. For this reason, new sets of projections are normally published every two years when the underlying assumptions are reviewed in light of more recent evidence.
A time of uncertainty in demographic trends
Our previous blog referenced increased uncertainty in our population estimates. Furthermore, there has probably never been a more difficult time to develop the assumptions for such long-term projections.
For our NPPs we use the most up-to-date mid-year population estimates as the base population from which we project forwards for up to 100 years. In our current population statistics system, the mid-year estimates continue to be rolled forward from one census to the next (with births, deaths and migration factored in). This means that whilst we are processing and quality assuring census outputs from Census 2021 ahead of publishing our estimates from late spring 2022, our current mid-year estimates have a higher uncertainty compared to earlier points in the last ten years. This is because they are based on the 2011 Census and we are at our furthest point away from this before new census data becomes available.
We plan for our next set of national population projections to be 2021-based and published in 2023. These will include Census 2021 data and a range of variant projections to give an indication of the outcomes which would come from different demographic scenarios for fertility, migration and mortality. Once again, we plan to update our demographic assumptions and benefit from revised population estimates for the decade from 2011.
For our 2020-based interim NPPs we have updated all our assumptions and used the latest data to give the best possible insights on future population change and what this means for the overall population.
Engagement with users of these statistics highlighted that this is a time of uncertainty as possible impacts from the Covid-19 pandemic continue to unfold and it is challenging to distinguish between short- and long-term demographic changes with some demographic measures. This is especially the case where some data sources were interrupted in the early part of the pandemic in 2020.
However, it is important to remember that we use long-term, observed demographic trends to guide our assumptions on likely future demographic trends. For example, our migration assumptions are based on a 25-year period in which migration has varied considerably.
No variant projections
We usually produce a range of variant projections to convey possible different outcomes based on different future scenarios. We know that our users value these. Variant projections are based on higher or lower assumptions about the future trajectories of fertility, mortality and migration. Because of the heightened uncertainty at the moment we have only produced one set of projections (principal projections), as it would be very difficult to assess the plausibility of a range of variants. We plan for our future releases to include a range of variant projections and understand that our users value and benefit from making comparisons between them.
Other developments in projections
As announced last spring, we plan for our next subnational population projections and household projections (for England) to follow in the middle of 2023, again using Census 2021 data. Since outlining our future plans for research on population estimates and projections in July 2021, where we addressed the recommendations made in the Office for Statistics Regulation (OSR) review of ONS population estimates and projections, we have progressed many areas of work, including the publication of guidance on use of variant population projections and household projections.
We have been reviewing our user engagement activities which are completed ahead of each set of subnational population and household projections. In 2022 we will conclude this and are endeavouring to build local and user feedback into our next set of projections outputs in 2023. At the same time, we continue to work on the transformation of our population and migration statistics system and where possible will seek to realise the benefits of this in the short-term for our projections.
We always welcome feedback (to email@example.com) and as we plan for our next set of projections (national, subnational and household) will continue to gather views to ensure our outputs continue to meet the needs of wider government and the public.