Pensions are finally capturing the public imagination. But, unfortunately, for all the wrong reasons.

This week, up to 1m public sector workers prepared to strike over government plans to change their pension schemes. With an election looming, the government eventually climbed down and agreed to new talks on the key issues, which include plans to raise the age of retirement for most public sector workers from 60 to 65.

The workers' victory is almost certain to be short-lived. Once the election is over, we are likely to hear a lot more about "longevity risk" - a new buzzword that measures the financial costs of longer lives.

The startling burden that will fall on all of us is only now starting to be fully understood by experts. With overall life expectancy at birth rising at an average of two and a half years each decade, assumptions about the total costs of retirement are being shattered.

In crude terms, men now have to fund a retirement at 65 that's likely to last 19 years on average - up from 13 years in the early 1970s. And most of them also have to fund retirement income for wives who will probably outlive them. A 65-year-old woman in 2005 can expect to live another 22 years.

Adrian Gallop at the Government Actuary's Department (Gad), says the rise in life expectancy is likely to continue well into this century, although looking a long way ahead is an inexact art. "We produce mass population projections 40 years and 70 years ahead - some bodies such as National Insurance need them - but effectively you can't predict this far ahead, it is just a projection. There may be a miracle cure for cancer, cures for genetic diseases - or possibly the rise in obesity among young people will affect mortality," says Gallop.

A recent paper on longevity, co-written by Gallop, points out that very elderly people are vulnerable to a range of medical conditions - so even if some causes of death among the very old are eradicated, other life-threatening diseases are just as likely to take their place.

Nevertheless the authors suggest there is room for substantial improvement in life expectancy at retirement age through medical advances.

Heart disease is considered a candidate for eradication - modern drugs have already revolutionised the treatment of heart disease since the late 1980s.

Stephen Richards, head of mortality risk at Prudential, says: "These drugs have halved heart disease mortality and I imagine heart disease mortality will go on falling steeply in the next few years. These are real modern day wonder drugs happening now."

Individual retirees are already feeling the impact of these trends. Final salary pension schemes which pay a set level of income based on salaries at retirement are being closed down, mainly because of rising longevity.

Similarly, annuity rates have plummeted over the past decade, partly because of falling long-term interest rates, but also because we are living much longer.

Yet the biggest burden of financing longer retirements currently falls on the government - through its myriad roles as financier for the NHS, state pension provider and public sector employer - and on companies running defined benefit (final salary) schemes for employees.

Life assurers will be less dramatically affected, partly because they have already been dealing with the rising costs of longevity. For example, Legal & General set aside up to £100m in reserves last July to allow for revised longevity expectations in its annuity business.

A recent report on longevity risk by Richards and his Prudential colleague Gavin Jones suggested that total shareholder exposure to the costs of longevity is some £30bn across all UK-listed life companies. Shareholders in other publicly-quoted companies, by contrast, face exposure to costs of some £762bn to finance retirements.

And there's very little publicly available information about the assumptions made by public companies when calculating the expected retirement cost of their employees - some companies are using longevity data that is 10 years out of date. BAE Systems, for example, increased its pension scheme liability from £2bn to £14bn after its once-a-decade mortality review.

It seems there's no escape from the cost of long lives: most people's pension funds are invested at least partly in shares in UK companies, and the long-term returns on those shares may be lower than expected if they are in companies which are underestimating the costs of longevity within their own employees' pension scheme.

Other financial schemes potentially affected by longer lives include lifetime (or equity release) mortgages sold to older people. These deals work by remortgaging the home to release a lump sum or income, with nothing to repay until the homeowner dies or moves into residential care.

Reputable lifetime mortgages have a "no negative equity" guarantee on these deals, meaning that no-one can owe more than the value of the home to the lender. But if house price inflation in future slows down to less than the interest accumulating on the outstanding mortgage balance, then a long life could mean there is nothing left to inherit.

There is much debate among longevity experts about the possible limits of the human lifespan. Medical advances can nudge life expectancy upwards, but unlocking the key to a lifespan of 130 years or even longer would take massive breakthroughs in our understanding of the body's ageing process. The social and medical costs would be huge, but the financial burden for the individual might be unbearable.

It is individuals who are likely to have to take on much more of the cost of longer lives, as government and companies seem certain to scale back pension benefits.

The limit of our financial tolerance for a long lifespan might come much earlier than we might think.

Assuming a future average lifespan of even 95 years, which might well be achievable for some of the professional classes now in their early working lives, a pension pot of £100,000 at age 65 would buy, at current rates, an income of £5,900 a year. That's a drop of 19 per cent compared with £7,300, a level annuity income sold for £100,000 on current longevity expectations.

This is a crude measure, warns Peter Quinton, of independent advisers the Annuity Bureau, who drew up the calculation. But he already sees warning signs that individuals are disastrously underestimating the length and cost of current retirements.

"Just about everyone now demands a level annuity, rather than an indexed one, because you get more income for your money. But £1,000 of buying power now over a 20-year period will drop to £603 even at an average inflation rate of 2.5 per cent," he says.

"I really think people are very seriously underestimating how long they are likely to live for."

Words that are likely to come back to haunt many of us in later life.