Don't blame the actuaries for longer life

There is an old adage that an actuary is someone who found life as an accountant too exciting. All that has been changing. The difficulties which pension funds have experienced in the market turbulence of recent years, along with the effects of an ageing population on the finances of the state provision for retirement, have thrust actuaries into the limelight. Indeed, it is the growth in life expectancy – the other side of the coin of the rising average age of the population – that has done far more than other factors, such as the collapse of equities between 2000 and 2003, or Gordon Brown’s ‘raid’ on the dividend income of pension funds, to produce the present crisis. And actuaries have been saddled with the blame for not keeping up with events.
Calculating future liabilities of a pension fund and setting the contributions of employers and employees to meet them, is a constantly moving target. A given sum of money has to cover a pension to be paid many years ahead and must allow for a wide range of fluctuation in uncertain variables. To some extent, these can be taken into account: inflation rates can be partly hedged by the use of long-term indexed bonds, and a broad assumption made of some long-term growth in the economy. But life expectancy is far more difficult to hedge.
Actuaries have been under attack for not predicting with any accuracy the pace of declining mortality rates. This is rather like blaming the weather forecasters for not getting next June’s rain and sunshine figures right a year in advance. There is more to predicting the future than continuing yesterday’s trends through today to tomorrow.
Even looking backwards is not so easy. It may have seemed a simple matter, at the end of 2006, to find out just how many people died at the age of 65 in that year. But the Continuous Mortality Investigation, which does just what its name suggests, was only able to base its most recent analysis (for 2006) on the pattern of deaths reported to insurance companies between 1999 and 2002. So, however accurate a snapshot is provided, time had moved on. While there was anecdotal evidence of lifespans continuing to lengthen, there was not, and could not have been, any firm basis on which to project the pace of continuing change. It is therefore unhelpful for the National Association of Pension Funds to simply complain that fund trustees are ‘becoming increasingly frustrated about the inability of their actuarial advisers to offer clear and accurate guidance on members’ life expectations’.
In fact the actuaries have (wisely) abandoned any pretence to be able to provide a unique insight into the future. Their 2006 series of mortality tables, known as the ‘00’ series after its base year, makes no predictions of trends beyond the year 2000 – unlike all its predecessors. Rather like the Bank of England trying to foresee the future trend of inflation, it has adopted a diagrammatic fan shaded to indicate the degrees of probability by darker and lighter hues around a central trend. One such projection suggests that a man born in 1950 and retiring in 2015 is likely to survive for between 21 and 26 years.
That variation obviously has significant implications for the appropriate level of contributions required to cover his pension commitment. But it may be preferable to picking out a single figure (or a much narrower range) which is correspondingly more likely to prove wrong.
It is better to be roughly right than precisely wrong. Actuaries are in fact in a good position to provide a combination of precision and reasonably well-founded prediction. While it is possible to identify factors which have been increasing longevity (such as the decline in smoking; more successful treatment of a range of diseases – particularly a drop in deaths from stroke and heart attacks) – it is impossible to say how much further this progress will go. Nor is it possible to say anything meaningful about the chances of an increase in death rates from, for example, the potential effects of more widespread obesity among the younger age groups in society.
The Scottish actuary who was asked by a farmer friend to estimate the number of sheep he had in two fields they were looking at, and who replied, instantaneously, ‘317’ was probably wiser than he may have appeared. Asked how he arrived so quickly at a precise answer, he said: ‘I can see 17 in that field, and there are about 300 in the other’.
Attempts to forecast a pattern of life and death fifty years ahead, and making this the centrepiece of pension-fund management is to look for a chimera in a mirage. However, one clear implication is that this must spell the final, lingering death of the already evaporating defined benefit pension fund, which purports to be able to provide not merely a pension linked with a given proportion of final or average pay, but one which is also inflation proofed.
It is not just actuarial uncertainty that leads to this conclusion. In a fluctuating economy – and it would be wise to be prepared for ups and downs to continue on a larger scale than we have been used to over the past twenty years or so – the whole process of valuing a fund’s assets and liabilities becomes unstable. A rise in interest rates has the effect of making a fund apparently more viable by lowering the discounted value of its future liabilities perhaps thirty years ahead. A decline in share prices cuts its current assets – and if interest rates are reduced to try and stimulate the economy, there is a double whammy.
One practical alternative to pension funds looking to actuaries for detailed advice which nobody can produce, is to fund their schemes deliberately at the high end of life expectations. Such over-funding would avoid most of the risk to viability, but would mean modifying the principle of equating a pension with an annuity. Instead of any pension payments not drawn because of the premature death of a beneficiary going back into the general pot, a proportion represented by the over-funding at the time of death could be returned, untaxed, to the deceased’s estate.
This would introduce an element of individual benefit to pension schemes, but only to the extent that an individual’s own contributions were affected by the date of his or her own death. This is far preferable to the emerging indications of possible attempts to tailor pension contributions and benefits more closely to the assumed characteristics of groups and sub-groups of fund members – with the clear potential eventually to tie them to the type or work, place of residence, health record and other characteristics of the lifestyle of each individual.
Box 1 ('Living longer’) sets out the findings of some recent studies into such variables. So far insurance companies have denied any intention of using such information in setting individual (or even group) contribution rates, but it may not just be simple cynicism that suggests that this may be because the information is not yet sufficiently robust.
For example, the variation in mortality rates within local government areas (shortly to be analysed at postcode level) does not allow for people spending different periods of their lives in different places. Similarly, the type of job, and the kind of employer, as well as the frequency of moving from job to job, must affect the figures that appear in the box.
Once these problems have been tackled, the threat of the individually assessed pension contribution will loom, and the whole basis of the insurance approach will be eroded. A significant hint as to the way this is likely to be presented comes from a spokesman for Legal & General who said that:
“Our test run shows that around one in five could benefit from postcode pricing. But those who don’t qualify will still get the standard rates – there won’t be reductions for people living in the leafiest areas.”
For that to be true, the clear implication is that, although payments of pensions will not be cut for the ‘losers’, contributions will have to rise if profit margins are to be maintained. Given the likelihood that, at least for the next decade or two, lifespans will continue to grow, firms will need to adapt.
Already, without such growth, there is an even chance that one of a couple now aged 65 will reach 90 – and one chance in six of them reaching 100. Some estimates suggest that up to one in three of today’s 30-year olds will become centenarians.
For every year over 65 you live today, you can expect to survive for an additional month. (That sounds even more impressive when expressed as 5 minutes of extra life for every hour survived).
Even more modest increases will be costly. A 10 per cent improvement in life expectancy would reduce an annuity on £100,000 from £6,700 to £6,250 (assuming unchanged interest rates). The problem of ensuring continued provision of pensions is also highlighted by recent wrangles over the viability of pension funds run by companies which become takeover targets: since 1984 as many as 68 of the companies then in the FTSE 100 have either failed and gone out of business or been broken up, merged, or taken over.
The most striking example of how pension payments can cripple a company is perhaps General Motors in the US. Until a recent complete financial reconstruction, it was best described as a generous pension fund operating a small loss-making car manufacturing business. Liabilities to retired workers, and contributions to current employees’ pensions were taking over 30 per cent of the sale price of every car produced and accounting for losses of around $250 per vehicle.
There is one final scenario worth a brief look. There is serious talk in some scientific circles of being able – within the next century – of extending human life almost indefinitely, by discovering how to repair and sustain organs and tissues. Whether that would be nirvana or nightmare is debatable. Certainly, it would solve the pension problem (a healthy and productive life of 900 years followed by 100 years in retirement should not be difficult to finance). However advocates of such an achievement, such as Aubrey de Grey and Bryan Appleyard, tend to overlook the implications for population growth – and the political totalitarian system that would inevitably become necessary. Box 2 ('Immortality would be deadly’) sets out a simple calculation.
In all this, actuaries have a role to play, but it is one of analysis and advisers, rather than to be seen as mysteriously omniscient witch-doctors with privileged insight into the future.
Two actuaries who have not seen each other for a long time meet in the street. ‘Good Heavens, John’, says the first. ‘I didn’t know you were still around. How old are you now?’ The other gets out of his pocket book of mortality rates. ‘Ninety-two,’ he says. ‘There seems to be an 80 per cent probability that I’m dead. How are you?’

Harvey Cole is ECA’s economic consultant. He can be contacted on 01962 865930.

Living Longer

Who?
It has long been known that employment can affect longevity. A new study by the Office of National Statistics analyses more closely the current pattern. It divides the population into seven groups rather than the familiar ABC1, C2, D and E. The top of the hierarchy shows only one third of the chance of dying before the age of 65 than those on the bottom rung of the ladder. The rankings are:

Sub-group 1.1: Out of every 100,000 in this group, 182 die before reaching 65. They include directors and chief executives of major companies.
In sub-group 1.2 come professionals, like doctors, lawyers, accountants, architects – and actuaries, for whom the figure is just over 200.
Rates then rise steadily through: Group 2 (including social workers, teachers and nurses); Group 3 (counter clerks, secretaries and designers); Group 4 (opticians, farmers and shopkeepers); Group 5 (bakers, plumbers and electricians); Group 6 (farm workers, shelf-fillers and traffic wardens); and, finally, Group 7 (bus drivers, bar staff, car park attendants, labourers and cleaners), whose mortality rates reach over 500 per 100,000.

What?
Another analysis looks at the figures for 76 of the largest UK companies quoted on the London Stock Exchange who publish details of their pension funds. This indicates that your chances of a longer (not necessarily happier) life are best if you worked for a large property company: British Land and Land Securities topped the list with 24.6 and 24.0 years of pensionable retirement respectively.
Variations between companies in the same line of business are considerable. Other companies to figure in the top ten include: Intercontinental Hotels; 3i; Cable & Wireless; BAE; BG, and Enterprise Inns.

Where?
Regional variations in longevity have also been known for a long time. Now a leading insurance company, Legal & General, has put some new figures on them.
The ten local authority districts with greatest life expectancy and the ten with the lowest are set out below. All the 400-odd others come between 73.2 and 79.8 years

Immortality would be deadly

SUPPOSE THAT it became possible for everyone to live (apart from the odd accident or suicide at the prospect of such longevity) not to the age of 1,000 – as envisaged by Aubrey de Grey and Bryan Appleyard – but to a mere 500 years.
For simplicity, the calculations are based on just the UK population, and assume that a gradual rise to the average age of 500 has been achieved at a future base date — and that there is no continuing immigration.
Taking an initial population of 60 million, of whom only 90 per cent will choose to have children, and that there is official restriction on the right to do so (limited to one every 15 years between the ages of 30 and 45), the initial population of 27 million prospective parents will produce over 750 million offspring.
With virtually nobody dying for 500 years, the population would rise to over 800 million.
But that is just the beginning of the story. The first-born to the long-livers — the 27 million born in year 30 of the new era — will themselves produce 26 children each by year 450, and each of their siblings will give birth to offspring reducing from that figure by 2 million each time a step is taken down the birth order.
After 450 years, the prospective population of this country would be several times larger than the present numbers inhabiting the globe.
Even a much more radical restriction on parenthood would lead to impossible conditions. Certainly, any substantial increase in longevity, accompanied by a corresponding extension of healthy and vigorous life, could only be achieved by transforming society beyond all recognition, and at the cost of a ruthlessly dominant central authority to control every aspect of decreasingly human life.
If something can’t happen, it won’t.