The local government pension scheme, like all such schemes, depends on accurate data and assumptions so that decisions about its long-term viability can be made. Of all the assumptions, one of the most important is the longevity of scheme members.
In the past, the life expectancy of people paying into the LGPS was badly underestimated, leading to underfunding.
This happened in the 1980s and 1990s, when some councils paid lower contributions than their scheme members - who continued paying their contributions in full - for years.
Now we are seeing the pendulum swing in the other direction. Actuaries are making more cautious assumptions, and predicting large rises in the life expectancy of scheme members.
While improvements in longevity are predicted for the population as a whole, we need to look more closely at the people paying into the LGPS.
The majority are low paid, and will receive low pensions. For women, the average is just £2,000 a year, rising to £4,000 for men. It is a sad fact that this will have an effect on their life expectancy.
The Continuous Mortality Investigation Bureau says that men with a pension of less than £13,000 pa, and women with less than £5,000 pa can expect a below-average rise in longevity.
Clearly, this should be taken into account. But the LGPS is heading for a valuation, and Unison is worried that the reduced life expectancy of low-paid scheme members will not be accounted for. If life expectancy is overestimated, scheme members will pay the price, either in higher contributions or poorer pensions.
Unison is working to make sure that all assumptions made about the scheme are based on experience, and not on dogma.
Glyn Jenkins, head of pensions, Unison