Prospects for human longevity

From AcaWiki
Jump to: navigation, search

Citation: Olshansky, S. Jay, Carnes, Bruce A., Desesquelles, Aline (2001) Prospects for human longevity. Science (Volume 291) (RSS)
Internet Archive Scholar (search for fulltext): Prospects for human longevity
Tagged: uw-madison (RSS), wisconsin (RSS), sociology (RSS), demography (RSS), prelim (RSS), qual (RSS), WisconsinDemographyPrelimAugust2009 (RSS)

Summary

As life expectancy rises, it is less sensitive to changes in death rates. This is called ENTROPY in the life table. Example: If e0=50 then you need a 4.1% decrease in mortality at all ages for a gain of 1 year in e0. If e0=80, than increases to a 9.1% reduction. In 1990, Olshansky asserted that e0 for males and females would be unlikely to surpass 85 year. He updates with newer data in this article. For the U.S., mortality would have to decline by 50% at all ages to reach e0=85. In reality, huge declines in mortality would have to occur at older ages b/c there isn't much room for decrease before age 50. Rapid increases in e0 will not happen. Entropy causes Olshansky to conclude that e0 is no longer reliable as a population health indicator. The U.S. Social Security administration has ignored entropy and predicted a rosy e0 or 83 and 87 for males and females in 2070. This has serious policy implications. Responses: Ronald Lee responds that Olshansky and colleagues should have used more than 10 years of data for their projections, and that the Social Security Administration was in fact prudent in its life expectancy projections because their predictions were less than long-term historical trends would suggest. The authors counter that recent data must be used because of shifting demographic regimes where improvements in mortality must be made among the oldest in the population to increase life expectancy.