附註:Includes bibliographical references.
Contributions presented at a workshop of the same title, organized by the Netherlands Interdisciplinary Demographic Institute on 5 September 1997--Foreword.
Cover -- Table of Contents -- Foreword -- Preface -- List of Authors -- List of Figures -- List of Tables -- Part 1. Introduction -- 1. A Review of Demographic Forecasting Models for Mortality -- 1.1. Most Common Classifications of Forecasting Models for Mortality -- 1.2. Parameterisation Functions -- 13 Statistical Association Models: the Lee and Carter Approach -- 1.4 Age-Period-Cohort Models -- 1.5 Mortality Forecasts/Projections/Scenarios in International Agency Practice -- 1.6 Forecast Errors -- 1.7 Prospects for Modelling and Forecasting of Mortality -- 2. A Review of Epidemiological Approaches to Forecasting Mortality and Morbidity -- 2.1 Introduction -- 2.2 Statistical Regression Models -- 2.3 Dynamic Multistate Models -- 2.4. Discussion -- Part 2. Theoretical Perspectives on Forecasting Mortality -- 3. A Regression Model of Mortality, with Application to the Netherlands -- 3.1 Introduction -- 3.2 A Regression Model of Mortality -- 3.3 Mortality in the Netherlands: Formulating a Model -- 3.4 A Descriptive Model for Mortality in the Netherlands -- 4. Forecasting Mortality from Regression Models: the Case of the Netherlands -- 4.1 Descriptive and Predictive Models -- 4.2 Forecasting Dutch Mortality from a Descriptive Model -- 4.3 Forecasting Dutch Mortality from a Predictive Model -- 4.4 Discussion -- 5. Gompertz in Context: the Gompertz and Related Distributions -- 5.1 Introduction -- 5.2 The Basic Gompertz Model -- 5.3 The Gompertz model as a model of survival and duration data -- 5.4. The Gompertz Distribution as an Extreme Value Distribution -- 5.5. Conclusion -- 6. Comparing Theoretical Age Patterns of Mortality Beyond the Age of 80 -- 6.1 Introduction -- 6.2 Improved Data, New Models -- 6.3 Mortality at Age 80 to 109 Years in Four Countries -- 6.4 Fitting Models to Data for Ages 80-109 -- 6.5 Extrapolation of the Age Pattern seen from the Perspective of 14 Models -- 6.6 Discussion -- Part 3. From Theory to Practice -- 7. Predicting Mortality from Peri
摘要:Information on future mortality trends is essential for population forecasts, public health policy, actuarial studies, and many other purposes. Realising the importance of such needs, this volume contains contributions to the theory and practice of forecasting mortality in the relatively favourable circumstances in developed countries of Western Europe. In this context techniques from mathematical statistics and econometrics can provide useful descriptions of past mortality. The naive forecast obtained by extrapolating a fitted model may give as good a forecast as any but forecasting by extrapolation requires careful justification since it assumes the prolongation of historical conditions. On the other hand, whilst it is generally accepted that scientific and other advances will continue to impact on mortality, perhaps dramatically so, it is impossible to quantify more than the outline of future consequences with a strong degree of confidence. The decision to modify an extrapolation of a model fitted to historical data (or conversely choosing not to modify it) in order to obtain a forecast is therefore strongly influenced by subjective and judgmental elements, with the quality of the latter dependent on demographic, epidemiological and indeed perhaps more general considerations. The thread running through the book reflects therefore the necessity of integrating demographic, epidemiological, and statistical factors to obtain an improvement in the prediction of mortality.
