In November 2009, I wrote an article, (Updated life tables: Is the tide shifting for Australian life expectancies?), reporting that average life expectancies have increased, except for women in their eighties and nineties, based on the latest life tables released by the Australian Government Actuary.
I invited Bernau Hadyn, an Actuary from the Australian Government Actuary, to comment on this shift, and whether the life expectancy for older women had peaked, or whether it was merely a statistical blip.
He referred me to page 6 of the Australian Life Tables 2005-07 (which is page 12 in the PDF document), where the report explains the changes to the mortality rates since the 2000-02 Life Tables were released five years ago.
Note: The section refers to mortality rates, which is the opposite of life expectancies rates, so if there is an increase in mortality rates, then that really means average life expectancies have decreased. Again, if mortality rates have decreased, average life expectancies should have increased.
This is what Bernau Hadyn had to say:
Our published mortality rates have increased at older ages for both men and women since 2000-02, partly because of an error in the 2000-02 Tables, and partly because rates from age 94 onwards have genuinely increased.
We found some historical data which shows that, across most of the major developed countries, over the last 50 years or so, for very old ages, it has not been uncommon for mortality rates at very old ages (>90) to increase over time. We speculated in the document that the average health status of the very old group might be declining, as many lives are being prolonged through improved medicine.
One more issue which you may already understand is the issue of mortality improvement. The ‘life expectancy’ figures that we publish are based upon mortality rates relevant basically to the 2005-07 period; being the number of deaths in that period, divided by the population as at the 2006 census. So, we effectively have the rates of death which are occurring at this point in time.
Our ‘life expectancy’ estimates assume that these rates will continue into the future completely unchanged, that is, for a baby born in 2006, we assume that, at age zero, they have the probability of death within their first year as recorded in our Tables. Likewise, when they reach 20 in 2026, we assume their probability of death before they reach 21 will be the same as recorded in our Tables for people who were 20 in 2006. In reality, of course, mortality rates are likely to improve over the next 20 years, and so the 2006 newborn is likely to live longer than suggested by our standard ‘life expectancy’ figures.
Our cohort life expectancy estimates (p19) take mortality improvement into account, and are thus intended to genuinely estimate how long people will live, rather than our standard ‘life expectancy’ figures, which are more like a statistical summary of the mortality rates which occurred over the three-year period from 2005 to 2007.
For more information, you can refer to the Australian Life Tables 2005-07.
SIGN UP TO OUR EMAIL NEWSLETTER for all the latest information, Q&As, and tips about superannuation. View latest newsletter | View previous newsletters
Copyright Trish Power
Hi - I'm Trish Power and I am the author of 
Recent Comments