Can we be a little more responsible about the life expectancy figures we’re quoting? “The current life expectancy is 77 for men and 81 for women” it says on the BBC website. So this means that a retirement age of 66 would mean men working for over 40 years to spend only an average of 11 in retirement, right? Wrong.

First, these life expectancies are period life expectancies, based on the current probability of survival for one year at each age – so the probability that someone currently aged 30 will survive until 31, that someone currently aged 90 will survive until 91, and so on. In other words, they do not allow for any potential improvements in mortality rates. This is incredibly pessimistic. Even the worst of the Office of National Statistics’ projections gives a cohort life expectancy – which does allow for expected improvements – of over 81 years for men and 85 years for women. The central estimate is for life expectancies from birth of over age 88 for men and 92 for women, giving each more than a decade compared with the period expectation of life.

But even this isn’t quite right. These are life expectancies from birth, whereas it is more appropriate to consider how long someone will live once they have retired. This means allowing for the fact that such people will by definition have survived until retirement but also that since they are older future improvements in life expectancy will be less relevant to them. The result is that, according to the ONS, a man currently aged 65 could expect to live for more than 21 years more, whilst a woman could expect to survive for nearly 23 more years.

That’s easy enough, isn’t it?