Stochastic modelling and projection of mortality improvements using a hybrid parametric/semi-parametric age–period–cohort model
Stochastic modelling and projection of mortality improvements using a hybrid parametric/semi-parametric age–period–cohort model
We propose a comprehensive and coherent approach for mortality projection using a max-imum likelihood method which benefits from full use of the substantial data available onmortality rates, their improvement rates, and the associated variability. Under this ap-proach we fit a negative binomial distribution to overcome one of the several limitations ofexisting approaches such as insufficiently robust mortality projections as a result of employ-ing a model (e.g. Poisson) which provides a poor fit to the data. We also impose smoothnessin parameter series which vary over age, cohort, and time in an integrated way. GeneralisedAdditive Models (GAMs), being a flexible class of semi-parametric statistical models, allowus to differentially smooth components, such as cohorts, more heavily in areas of sparse datafor the component concerned. While GAMs can provide a reasonable fit for the ages wherethere is adequate data, estimation and extrapolation of mortality rates using a GAM athigher ages is problematic due to high variation in crude rates. At these ages, parametricmodels can give a more robust fit, enabling a borrowing of strength across age groups. Ourprojection methodology assumes a smooth transition between a GAM at lower ages and afully parametric model at higher ages.
Age–period–cohort model, expert opinion, generalised additive model, overdispersed data, projection
134-155
Dodd, Erengul
b3faed76-f22b-4928-a922-7f0b8439030d
Forster, Jonathan J.
f3279b60-9eb4-43fc-bf21-945926529eb8
Bijak, Jakub
e33bf9d3-fca6-405f-844c-4b2decf93c66
Smith, Peter W.F.
961a01a3-bf4c-43ca-9599-5be4fd5d3940
16 September 2020
Dodd, Erengul
b3faed76-f22b-4928-a922-7f0b8439030d
Forster, Jonathan J.
f3279b60-9eb4-43fc-bf21-945926529eb8
Bijak, Jakub
e33bf9d3-fca6-405f-844c-4b2decf93c66
Smith, Peter W.F.
961a01a3-bf4c-43ca-9599-5be4fd5d3940
Dodd, Erengul, Forster, Jonathan J., Bijak, Jakub and Smith, Peter W.F.
(2020)
Stochastic modelling and projection of mortality improvements using a hybrid parametric/semi-parametric age–period–cohort model.
Scandinavian Actuarial Journal, 2021 (2), .
(doi:10.1080/03461238.2020.1815238).
Abstract
We propose a comprehensive and coherent approach for mortality projection using a max-imum likelihood method which benefits from full use of the substantial data available onmortality rates, their improvement rates, and the associated variability. Under this ap-proach we fit a negative binomial distribution to overcome one of the several limitations ofexisting approaches such as insufficiently robust mortality projections as a result of employ-ing a model (e.g. Poisson) which provides a poor fit to the data. We also impose smoothnessin parameter series which vary over age, cohort, and time in an integrated way. GeneralisedAdditive Models (GAMs), being a flexible class of semi-parametric statistical models, allowus to differentially smooth components, such as cohorts, more heavily in areas of sparse datafor the component concerned. While GAMs can provide a reasonable fit for the ages wherethere is adequate data, estimation and extrapolation of mortality rates using a GAM athigher ages is problematic due to high variation in crude rates. At these ages, parametricmodels can give a more robust fit, enabling a borrowing of strength across age groups. Ourprojection methodology assumes a smooth transition between a GAM at lower ages and afully parametric model at higher ages.
Text
SAJ_accepted
- Accepted Manuscript
More information
Accepted/In Press date: 23 August 2020
e-pub ahead of print date: 16 September 2020
Published date: 16 September 2020
Additional Information:
Publisher Copyright:
© 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
Keywords:
Age–period–cohort model, expert opinion, generalised additive model, overdispersed data, projection
Identifiers
Local EPrints ID: 443823
URI: http://eprints.soton.ac.uk/id/eprint/443823
ISSN: 0346-1238
PURE UUID: da08316e-985c-42ac-984b-a3d3e35e8417
Catalogue record
Date deposited: 14 Sep 2020 16:31
Last modified: 17 Mar 2024 05:53
Export record
Altmetrics
Contributors
Author:
Jonathan J. Forster
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
