Does life expectancy tell us how long we will live?

I’ve written a PhD on this, but assuming you don’t want to read the full 100,000 words, here’s a quick summary: No.  It’s all down to probabilities.  But we tend to underestimate how long we are going to live, partly because of poor statistics and partly because we just can’t believe longevity is increasing.

Poor statistics:

  • “Life expectancy” is the average of a probability distribution. This distribution – of ages at death for a given group of people – is not symmetric.  More people die older than the average than die younger.
  • The figures often quoted for “life expectancy” don’t allow for future trends and so don’t give a real measure of potential lifespan for an individual still with many years to live**.  Even otherwise rigorous academics often use a statistic called “life expectancy” wrongly thinking it is “the lifespan we should expect”.

And just about everybody finds it hard to believe that we have been and still are living longer:

  • When demographers or actuaries have made assumptions about future longevity trends their techniques have generally led them to underestimate how fast lifespans are lengthening.
  • If we don’t use any statistics but just think about what feels right, our first thought is often about our parents and grandparents.  On average we will live longer than each preceding generation, but we tend not to factor that in.
  • Lifespans are getting longer, and most experts think that will continue, despite obesity, diabetes and other ills.  Medical technology helps to keep us living as the nature of ill-health is changing.

My own survey of adult New Zealanders suggested that on average men underestimate their likely age at death by 5 years and women by 7 years.  Other studies around the world show similar answers.

** I’ll take a few more words to explain this one.  Often a life expectancy is used from a “period” life table.  For example, it might be said that the life expectancy at age 65 for Kiwis is currently 21.4 years for females and 19.1 years for males.  This means that if patterns of mortality stay the same as they are in the current period, then on average current 65-year-old Kiwis would be expected to live until age 86 years (women) or 84 years (men).

The one thing we know about mortality patterns is that they change over time, so this is not a realistic assumption if we are interested in what a real lifespan might be.  Instead, making some assumptions for future mortality trends consistent with their population projections, Statistics NZ make a “cohort” life table. From this we can calculate cohort life expectancies, which do take future trends into account and so are more realistic indicators for average lifespans.  At the time of writing this, 2015, Kiwi cohort life expectancies at age 65 take average expected lifespans to around 89 years (women) or 86 years (men).

Period life expectancy is a useful measure, but just not for estimating realistic lifespans. It incorporates mortality at all ages at a point in time, so is an overall summary of the health of that population at that time.  Period life expectancy is often used to compare the general health of different populations, or to track the health of a single population over time.

Leave a Reply