Jim Robinson wonders whether we'll ever get longevity right.

Whenever my four-year-old boy asks me how old I am and I tell him the truth, I am invariably met with a look of horror that anyone, anywhere, could ever be that old. His incredulous, somewhat disgusted expression is an apt reflection of how I sometimes feel when I read reports on ageing and it dawns on me how most of us are going to be doing a lot more of it in future.

The IMF's recent analysis of longevity risk is a case in point. In the decades ahead, as populations age, "the elderly will consume a growing share of resources", it warns. People will start living so long they run out of money, and governments and pension funds will find themselves owing far more in social security benefits and pensions than they ever expected.

One major source of this growing problem is that, so far, we have been utterly rubbish at the science of longevity. Our population forecasts have been consistent only insofar as they have been consistently wrong. According to the IMF, if people live just three years longer than they are supposed to - a figure that is fully in keeping with past underestimations - the "already large" cost of growing old could rise by as much as 50% of GDP in developed economies. Perhaps Leon Trotsky was right when he said old age was the most unexpected of all the things that could happen to man.

Of course, very few governments in the world have bothered to take a long, serious look at longevity risk (otherwise, there wouldn't be such a big problem), and those governments that have recognised the risk invariably discover it's a rather big one. One of the few countries that has grasped the nettle - and a country the IMF holds up as a shining example to the rest of the world - is the Netherlands.

According to the IMF, the beauty of the "exemplary" Dutch system is its flexibility. It singles out four "reform elements" in particular: contribution stabilisation (contribution adjustments can no longer be used to absorb changes in life expectancy); marked-to-market liabilities (liabilities had been discounted at the risk-free interest rate); the lack of unconditional nominal commitments (benefits are now "explicitly conditional" on a pension fund's investment performance); and adjustments for changes in longevity (the effective retirement age is linked to expected changes in longevity).

So, if the IMF is correct, as long as the rest of us do what the Dutch are now doing, we can carry on growing old, safe in the knowledge that, when we do eventually retire, we will be well taken care of. By my calculations, that's somewhere around the age of 104. I can't wait to see my son's expression when I give him the good news.