Adair Turner likes to scribble. He is trying to explain the future development of the British population, so he just reaches across, grabs this reporter's notebook and starts drawing. First a triangle. Then something that looks like a church candle that is taller than the triangle. The two drawings are intended to show that in past generations the number of people in work was roughly three times those in retirement, now it is a wobblier two. He then draws a candle tapered at bottom and top to show that in Italy, because the birth rate is below the two per family it takes to replace the population, the situation is less stable still. The point is that the candle is likely to grow taller, meaning not only that people will live longer, but also that they will not be balanced by young working types.

Enough scribbling for now. Turner, former director general of the Confederation of British Industry, now a Merrill Lynch bigwig, is working one day a week for the Government as head of a pensions commission - or the 'yet another pensions commission following the inconclusive green paper', as critics call it. This time next year he hopes to diagnose the problems of a long-term Cinderella of social policy whose profile is now rising as workers find their company schemes are slashed back.

His thoughts so far are not comforting. He says: 'As the ratio of those in work to those out falls, one [or more] of four things must change. One: workers defer more income through contributions or savings, two retirees are less well off, three the retirement age is going to go up, four some way of increasing national income.'

Number four is pension Nirvana, negating the need to come up with sticky answers to the other three. But Turner is sceptical. His formula is 'If not four, then which of one, two or three?'

Turner concedes that if the gloomy calculations of the UK pension gap - the deficit between the amount defined-benefit schemes promise to pay out in future and the amount they are calculated to need - of up to £300 billion are true then 'almost everyone is not saving enough'. But he is not convinced - calculating a number he is happy with is part of the commission's work.

But even if the aggregate gap is much lower, there will be groups not saving enough. Making sure they do seems to be what Turner is focusing on.

He says: 'At the low end of the income distribution - maybe deciles 1 and 2 - people in almost any circumstances will get their retirement income from the state. For that group arguments about tax relief to encourage saving and voluntary or compulsory saving are not relevant. At higher income levels you have to take the attitude that if they did not make adequate provision that was their fault.'

It is the in betweens he is worried about - a group that should be saving privately now, but which, as a Financial Services Authority report shows, is not. And within this, there are exposed groups, women and the self employed, who traditionally have fallen outside the pensions net.

So much for the problem. What does Turner think is the solution? He is surprisingly forthright. Option three will happen.'The one thing that is clear in this story is the increase in average retirement age. As far as we know at the moment, the ageing process has no necessary limit - I believe that the estimates of life expectancy that the Government actuaries have been working on are likely to be revised up.' The candle, in other words, gets longer.

So the retirement age will have to rise - but Turner says this does not point to a future of supermarkets and call centres full of centenerians. 'The good news is that a fairly small increase in retirement age produces dramatic changes in the solvency of pensions.'

For example if, someone started working at 25, retired at 65 and died at 85 the ratio of work to retirement is two to one. If you raise the retirement age to 70 it becomes three to one, which means you can save less.'

So Turner thinks retirement age - not the age for state pension qualification - will rise, perhaps not to 70 to start with, but above 65. That will begin to happen automatically after 2005 when companies will be unable to set retirement ages.

Option two - people receiving less in retirement - is what the commission is trying to avoid, so it is not 'desirable'. So that leaves option one - encouraging people to save more. Here Turner's thinking crosses with other controversial areas of policy. As he puts it: 'How should we make them? Should they be compelled?'

Good questions. The compulsion one, Turner won't answer. 'I am genuinely open minded about it,' he says.

But there are other issues he will discuss, such as the highly controversial one of the demise of defined benefit schemes for new entrants in the private sector. Turner is crystal-clear here - the decline in these schemes should go no further, and employers are wrongheaded about shutting schemes off.

Turner believes that some half of schemes have been closed. 'If the new inflow has fallen by over 50 per cent, over time the stock of employees and pensioners of these schemes will fall by 50 per cent. That would be very unfortunate. These schemes are something that do not leave them subject to the vagaries of the market. The members of these schemes get a return that is double secured - by their assets and by the guarantee from their employers.'

Employers themselves, he says are being too 'binary' - looking at the fall in the stockmarket that ended the 'fools paradise' of the 1990s pension fund surplus, deciding that the final salary scheme is unsustainable, and closing it down. Turner says he hopes they become more flexible. 'Instead of saying "I have a final salary scheme with a 65 retirement age, lets close it", why not look at, say using average salary, or raising the retirement age, or doing a hybrid scheme, where 60 per cent is final salary and 40 per cent is defined contribution [where employees build up their own fund and buy an annuity at the end].'

These problems are all in the context of 'If not [option] four.' What about four? Turner says the question of whether investment activity can raise GNP, by taking funds and investing them in areas of the world where the demographics are different, are controversial questions which the commission will have to think hard about.

It would be nice, he says, to think there would be a silver bullet, but he is not convinced.

Unless you have a valuable work of art hidden away, you need to start thinking about it now. Talking of which, perhaps its time to get that Turner drawing framed. Just a shame it's by Adair, not JMW.