Population estimation and forecasting

During the transition period, the CPC modelling strand will focus on (a) extending our work on demographic estimation and forecasts through further improvements in the way mortality, fertility and migration are estimated, and (b) widening the scope of our research to encompass population forecasts for small domains.

Our work on mortality has identified challenges as to the best methods for estimating and predicting mortality at the oldest ages. This line of research will be further explored by designing and testing approximate and fully Bayesian models for mortality forecasting, engaging directly with the actuarial profession. Similarly, our research on fertility forecasts has identified a gap in the current state of the art with respect to the use of information on birth order. Here, we will develop models that take into account the multidimensional associations between age and parity-specific fertility rates, such as those based on couples.

The strand's work on migration has highlighted that long-term forecasts of migration are problematic given the high degree of uncertainty underlying the processes. Therefore, population projection methods will be developed with fully probabilistic treatment of mortality and fertility, but with stochastic migration scenarios based on elicited expert judgement. The team will also continue engaging with the Office for National Statistics and the European Asylum Support Office to facilitate implementation of the advanced estimation, forecasting, and early warning methods developed as part of CPC II.

Reflecting the shift towards increasingly devolved governance and the focus on regional development, the strand will explore new methods for the estimation and forecasts of populations in small domains, accounting for different components of demographic change, as well as their uncertainty. The research on small-domain estimation and forecasting will be motivated by the needs of local authorities, and other stakeholders, who are primary users of such outputs. This new thread of work could be of relevance to the UK Government's Industrial Strategy, providing a firm basis for robust forecasts of the local labour force and the demand for public services, including school places and the need for health and social care.

This research is co-ordinated by Professor Peter W.F. Smith and Professor Jakub Bijak.