The importance of getting demography right

The good news is that we are collectively living longer than forecasters have given us in the past. The bad news is that we’re not taking this information into account in our long term planning, which means that we may not be able to pay for the costs of this prolongation of life.

In April the International Monetary Fund (IMF) published a ‘The Quest for Lasting Stability‘, a report about threats to the long term future of financial systems. In the chapter ‘The Financial Impact of Longevity Risk’ the report highlights that one of the biggest long term threats to financial systems is the use of old demographic sources and models to calculate the life expectancy estimates which power long term planning in areas ranging from pensions and taxation to housing and health care.

Planning in all these areas requires a good estimate of how many inhabitants a country will have in each age bracket in decades to come. Of course, it has long been known that people are living longer now than they used to, and governments and pension funds have been preparing for the consequences of population aging. But these preparations have been based on population forecasts that have now been proven wrong. Data from several countries show that the estimates for people’s remaining lifetime made since the 1960s have consistently underestimated the real length of people’s life.

The difference between expected and actual life spans has been 3 years on average, regardless of the techniques used. For example, the graph shows that the 1971 forecast predicted that in 1991 life expectancy for men in the UK would be 71 years. In fact it was 74. The same pattern of underestimation errors is seen in forecasts from different decades and across countries: 20-year forecasts of longevity made in recent decades in Australia, Canada, Japan, New Zealand and the US have all been too low.

The main cause of these errors seems to be the assumption that longevity will not continue to improve at the same rate in the future, more or less following the law of diminishing returns. The underlying premise is that there is some absolute biological maximum to human life, and that if we make the ‘easy’ improvements to disease prevention and health care first, each subsequent improvement will be more difficult to achieve as we approach this absolute maximum. This line of thinking may make sense in theory; After all we don’t see people, even if they live in optimal circumstances, living beyond 120 years. But in practice, the point where improvements in life expectancy start slowing down has not yet been reached, mainly because advances in medical science continue to uncover new areas where improvements to health can be made. It is unclear what the future will bring in this area. It is impossible to predict whether medical advances will reach a point of exhaustion, or whether there is potential for even more revolutionary breakthroughs and extension of human life beyond the range we currently regard as normal.

At the same time, we are seeing a marked shift in where in the human lifespan gains are made. The biggest gains in life expectancy at birth in the 20th century have been due to marked improvement in infant mortality, while only a minor part was achieved through extension of life expectancy at older ages. Even so, life expectancy in older age has improved markedly in the past century. In advanced economies in Europe life expectancy at age 60 rose from 15 years in 1910 to 24 years in 2010, and it is expected to increase further.

At the younger end of the lifespan, infant mortality in developed countries is now so low, that further gains cannot be expected. Instead, future gains in life expectancy must come from gains in later life. It is important to realise that extension of life expectancy at young ages has no negative impacts on society, as the children who survive will grow into fully functional and contributing members of society. However, extension of life expectancy in the elderly directly extends the number of years people spend in retirement and possibly also in ill health. This has potentially severe consequences for the future planning of all sorts of things. Of course the IMF highlights the financial risks of an expanding group of elderly and dependent people.

The main consequence for pension models, public and private, is that people will need to receive payments for far longer than the current models assume. Unless far reaching changes in the amounts or duration of time people contribute to their pensions, it is likely that people will have a lower income in old age than they currently foresee. But this is only part of the problem. Governments are also highly exposed to unforeseen extensions of longevity through public pension plans, social security schemes, increased healthcare costs, and the risk that individuals will have insufficient resources to provide for themselves during the full length of unexpectedly long retirements. There are also threats to the solvency of private financial and corporate institutions exposed to longevity risk. If these institutions fail, the government may not have other options but to provide a last line of support analogously to the way governments were forced to step in to bail out banks deemed too big to fail during the first stages of the current financial crisis.

Mitigation measures include an increase of retirement age, and promoting higher levels of pension contributions during working life, but these measures will take years to have effect. Where does that leave us in the meantime? At the very least we should replace our current models with planning and prediction models which account in a more realistic way for expected increases in life expectancy. But more importantly, we should try to reimagine what it means to be 65 or 75 or 85. In the same way that 40 is now the new 30, 80 will be the new 65 and 100 may become the new 80.