Everything you always wanted to know about death projections, but were afraid to ask

November 2015 saw the ONS release their latest death projections, with some significant revisions to what was predicted back in 2012. But how important are these changes and what does it mean for legacies?

At first glance the changes in death predictions do indeed look significant, and should mean an increase in legacy incomes over the next few decades. The latest forecasts predict 17,000 more deaths per year in the rest of this decade, and 10-15,000 more deaths per year through the 2020s and 2030s. Cumulatively that implies over 250,000 more deaths over the next twenty years!

In legacy terms, this suggests that between now and 2020 total charitable legacies are likely to be 3% higher in each year than previously expected – that’s 18,000 more legacies over the five-year period. And from 2021 to 2030, legacy numbers will be 1.5% higher than previously expected – meaning 27,000 more gifts to charity over those ten years.

However, we wanted to delve a little deeper into these projections and understand: 1) Why have these numbers changed so much? and 2) Are these changes really significant or are they just part of the natural ‘noise’ of ONS’ population projections?

In both 2013 and 2014 the number of deaths was higher than the ONS had predicted. This in turn implies that life expectancies are not improving as fast as they previously thought. Going forward, the ONS has reduced projected life expectancies for both men (down from 83.0 to 82.8 years by 2030) and women (down from 86.3 to 85.8 years in 2030). Although this may not sound much, it makes a big difference to the number of deaths over time.

This is not, however, the full story. The rates of improvement discussed above represent just one element of the projections; equally critical in obtaining projected figures for the number of deaths is the base year from which the rates of improvement are applied. Although the deaths figures for ‘2014/15’ (actually July 2014 – June 2015) are not used as an input into the calculations for projecting life expectancy, they are used as the ‘base’ figure from which those projections are applied. The newly published Winter 2014/15 death rates reveal a big spike in deaths in January 2015, with the highest Winter deaths since 1999/00; deaths between July 2014 and June 2015 exceeded 600,000 as a result (over 50,000 more deaths than projected two years earlier). This spike is only partly explained by the usual suspects for driving excess winter mortality: overall, the Winter was relatively mild (albeit with a short cold snap) and there wasn’t a major flu epidemic (although the flu vaccine was more ineffective than usual so there were some minor outbreaks, particularly among the elderly).

So part of the increase to projected deaths is due to life expectancies [that] are slightly lower than were projected last time, but more important is that these projections are applied to a high ‘base-year’. The projections are fairly sensitive to changes in the value of the ‘base-year’: 2014/15 was a high base-year and the projections of deaths are fairly high as a result. For this reason we suggest that the best way to interpret these projections is to focus on the shape and direction of change, and not to worry too much about the precise number of deaths projected. Viewing the figures in this way helps us to answer the second question above: Are these changes really significant, or are they just part of the natural ‘noise’ of ONS’ population projections?

The most significant change implicit in these figures is that the long-predicted upturn in numbers of deaths has happened earlier than expected. ONS have long projected that the steady decline in the number of deaths through the 1990s and 2000s would turn around during the latter part of the current decade; the 2012-based projections suggested that 2017/18 would be the first year of increasing deaths. These latest projections suggest that the turning point has come 2-3 years earlier than previously predicted. The number of deaths will continue to be volatile from year to year but from 2015-onwards the trajectory will be up not down.

However, beyond the turning-point, the upward trajectory for UK deaths implied by the latest projections is not dramatically different to previous projections. Over the next ten years the underlying annual increase in deaths will be small, despite the continued possibility of quite dramatic year-to-year fluctuations. These fluctuations – and the sensitivity of the projections to the level of a volatile ‘base-year’ – suggest that projected deaths figures are almost certain to change again when ONS produce their next revision in two years’ time. But the underlying shape of the curve should remain relatively stable: the outlook for UK deaths is for a gradual increase in their number over the next ten years; this outlook is a common feature of the 2014-based projections and previous ONS projections alike.

Legacy giving during the next few years is – all other factors remaining equal – likely to be a little higher than previously expected, as a result of a slightly higher number of deaths. But the expected rate of growth in legacy giving implied by the 2014-projections of deaths is broadly unchanged against the 2012-projections.