Click the icon to view and/or download the PDF file.
Once you are in the PDF file, use your browser back button to return to this page.
Abstract
We propose an alternative way of dealing with mortality tempo. Bongaarts and Feeney have developed a model that assumes a fixed delay postponing each death. Our model, however, assumes that changes take place with the removal of a given cause of mortality. Cross-sectional risks of mortality by age and expectations of life therefore are not biased, contrary to the model of the two authors.
Treating the two approaches as two particular cases of a more general process, we demonstrate that these two particular cases are the only ones that have general properties: The only model enjoying a decomposable expression is the removal model and the only model enjoying the proportionality property is the fixed delay model.
Author's affiliation Hervé Le Bras École des hautes études en sciences sociales, France
