Author(s) :

Andrew Kwon.

Volume/Issue :

Abstract :

This study models adult mortality in the United Kingdom using the Gompertz-Makeham (GM) law and tests if its parameters were altered by pandemic-era conditions, while preserving model validity. Life tables from the Office for National Statistics (ONS) in the years from 2017 to 2019 (pre-pandemic) and 2021 to 2023 (post-pandemic) were used to fit the Gompertz-Makeham models for males and females, separately, in the range of ages from 30 to 90 with nonlinear least squares, while assessing the goodness of fit with log scale residual diagnostics. Predictions were validated against empirical survival curves and life expectancy at age 65 (e65). This study seeks to answer whether the Gompertz Makeham parameters differed between pre- and post-pandemic periods, and whether the model remained valid in predicting survival and e65. This study hypothesized that there was no significant difference in the Gompertz-Makeham parameters for UK adults between pre- and post-pandemic periods. The age-independent (background) component was higher for both genders on average in the post-pandemic period, but this pattern was suggestive rather than conclusive based on statistical tests. However, the age-dependent risk level and rate remained unchanged. Observed vs. fitted hazards aligned well on the log scale, and residuals remained close to zero in the range of ages from 30 to 90. There was a slight dispersion at the oldest ages. The Gompertz-Makeham survival supported ONS survival closely, and the differences of e65 remained small. Background risk was directionally higher post-pandemic but not statistically significant at the 5% level (males, p=0.10; females p=0.38). Age dependent terms were stable, and the model validation was strong for both survival and e65. Taken together, the results in this study were consistent with the background mortality through a level-type shift rather than a structural change in age-related risk, but these inferences were not statistically conclusive.