Summary
Life expectancies show a surprising degree of variation amongst developed nations. Circumstantial evidence suggests that the U.K., U.S.A. and Germany have plenty of scope for further increases, compared to similar wealthy nations. The implication of such increases would be increased cost of defined-benefit pensions promised by corporate pension plans.
Analysis
On 8th March, the actuarial profession in the U.K. posted a discussion
paper on its website. Entitled "Two-dimensional mortality data: patterns and projections", the paper compares and contrasts the
mortality patterns of seven industrialised nations: England and Wales,
the U.S.A., Canada, Japan, France, Germany and Sweden. The mortality patterns are particularly focused on the generations who are currently in, or about to enter, retirement. The results in the paper are therefore of considerable interest to those studying the affordability of pensions promises, either from corporate pension schemes or from government social-security plans.
The first sections of the paper contain much circumstantial evidence
that countries such as the U.K., U.S.A. and Germany still have plenty
of scope for further increases in life expectancy, especially when
compared to other industrialised nations. The paper also finds
several countries where the so-called cohort effect dominates
mortality patterns in the recent past, i.e. where the year of birth
proves to be more important than period effects. The U.K. is a
country with a well documented cohort effect, but this paper shows
that other countries like the U.S.A. and Germany also have one. The
cohort effect is important because it usually means a strong increase
in life expectancy in retirement compared to previous generations. If
not already adequately allowed for in funding calculations, it means
that retirement benefits will likely cost more than previously assumed.
The paper also uses different techniques to show that, in addition to
a cohort effect, researchers can also detect a separate time-based
trend. This trend reflects the fact that many sources of increased
life expectancy are time-based, i.e. improvements in diagnosis and
medical treatments. The implication is that projections of retirement
benefits must allow for the fact that there will be a certain minimum
amount of improvement due to technological advances. Again, to the
extent that such a minimum level of improvement is not allowed for in
funding calculations, retirement benefits may cost more than previously assumed.
The paper itself can be downloaded at
http://www.actuaries.org.uk
