Abstract
This paper takes a detailed look at socio-economic variation in mortality across England. Generic metrics such as the index of multiple deprivation (IMD) can be effective at the aggregate level of national deciles, but we demonstrate that they perform much less well at the regional, urban–rural, and neighbourhood level. We use local linear regression to develop a new, customized index for neighbourhood mortality that addresses the mortality-specific shortcomings of the IMD based on socio-economic and related, non-spatial predictive variables. We find that old-age income deprivation and employment deprivation are key determinants of mortality, but also that urban–rural class and the presence of care homes in a neighbourhood have an important role to play in assessing underlying mortality rates relative to national mortality. Residual spatial/regional variation in mortality is found to be much less significant than socio-economic variation and much lower than the residual regional variation that results from the use of the IMD.
Original language | English |
---|---|
Pages (from-to) | 989-1012 |
Number of pages | 24 |
Journal | Journal of the Royal Statistical Society Series A: Statistics in Society |
Volume | 187 |
Issue number | 4 |
Early online date | 27 Feb 2024 |
DOIs | |
Publication status | Published - Oct 2024 |
Keywords
- Social Sciences (miscellaneous)
- Statistics and Probability
- Economics and Econometrics
- Statistics, Probability and Uncertainty