Thursday, March 23, 2006

A note on life expectancy

I just saw a graph of how female life expectancy has increased steadily over time, passing points at which it was predicted to stop and level out without wavering. However, it seems the truth may be more complicated. Life expectancy is generated by a mean of years lived. Part of the reason that such large increases have happened is not because we have vastly increased the maximum age of death. After all, the bible, two thousand years ago, sets our span at threee score years and ten, yet we regularly speak of mediaeval life expectancies in the forties. What we have really changed is not death in old age, but deaths associated with birth. Child mortality massively skews any mean -- obviously, if lots of children die at birth, that will pull down enormously any average of lifespan across the population. Equally, fewer women now die in childbirth, which, as an event occuring relatively early in life, especially in historical times and especially in cases where the risk factors are higher, also pulled down the statistics. So we must be very careful in seeing a rise in life expectancy as proof that we are doing a lot to extend the lifespan overall. What we have done is let more people live out their biblical span. Of course, if we wanted to be gloomy, we could also argue that given how statistics have in the past been skewed by child mortality, they ought now to include abortions and infanticide. It would be interesting to see how that would affect mean life expectancy, especially given that abortions in Russia exceed live births and female infanticide in Asia is reported to have been widespread in recent decades. The claim might have to shift from, 'we have done wonders for the lifespan,' to 'the individuals who die now do so without suffering'. Which is an achievement of sorts, but not at all the one we currently celebrate.

No comments: