Con Keating and Dennis Leech explain how the IMF's latest warnings on longevity are misleading.

The IMF recently sounded yet another clarion call for governments and others to act in response to increasing longevity (Global Financial Stability Review, Chapter 4: 'The financial impact of longevity risk'). This is the latest in a long line of similar calls - though most of those have clearly vested interests, and the underlying analysis reflects that.

This report states "if individuals live three years longer than expected - in line with underestimations in the past - the already large costs of ageing could increase by another 50%, representing an additional cost of 50% of 2010 GDP in advanced economies and 25% in emerging economies". Now that is not so much scary as scaremongering.

These estimates are misleading because they relate a change in a capital sum - the present value of future pensions liabilities, which is a stock - with GDP, which is a flow. This is the kind of elementary error that one would hope an undergraduate student would not make. Of course, GDP is the appropriate yardstick to use for the purpose of making comparisons over time and between countries, but it is misusing it to say the increase in the stock of liabilities is a "cost" against GDP.

It would be churlish to challenge the relevance of a three-year underestimate in the case of the UK, but the authors do display as their Figure 4.1 the UK Projected Life Expectancy at Birth for Males, 1966-2031. That clearly shows the 2010 revision of these projections is lower than the 2008 projection, something that has never occurred previously in that sample of projections. One is tempted to ask: could it be that we have not only caught up with the rate of increase of longevity but are now overestimating rather than underestimating it? The typical assumption now used in pension scheme evaluation is 21.2 years versus a UK Population Life Expectancy of 17.2 years - this difference is an allowance for future improvements in longevity and is slightly larger than that observed since 1990.

These though are minor quibbles relative to the issue of how important a three-year increase in longevity really is. Let's take their figures and do some simple arithmetic. Let us assume a pensioner receives an income in retirement of 80% of the average labour income - the higher of the IMF figures. Allowing this pension for the entire population means the total cost is 2.4% of national income, or 1.6% of GDP.

It is, of course, payable in each and every year going forward. But this does not mean we should now discount all of these payments to arrive at some horrendous proportion of current GDP. In fact, an increase of 1.6% of GDP in any one year covers it fully - so what is the true cost of a three-year increase in life expectancy in projections spanning as much as 40 years? Well, it represents about the same as next year's depressed projected growth in GDP. It does not require growth to recur every year in the future time-span of the projection - but saying we may have to allow the growth of one year in 40 does not sell newspapers.

We could go on and discuss the relative shares of labour and capital in national income and potential ageing effects upon these, but that would be calm rational analysis, and that doesn't sell newspapers.

Con Keating is head of research at BrightonRock Group, while Dennis Leech is a professor at the University of Warwick's Department of Economics