If you wanted to design a state pension for the UK from scratch, you would not do it this way. Every commentator has his own preferred wheeze, with its unique bells and whistles, but across the literature there is a degree of consensus. We want to encourage people to save to provide their own funded pot, so there should be a tax-shelter in which they can accumulate reserves, with a generous annual allowance—something like a super-ISA. In fact, it would be sensible to amalgamate the various forms of tax-favoured saving (ISAs, pensions, etc) which would reduce management costs, and so improve returns and thus the eventual pension paid. Legitimate concerns about pensioner poverty could be addressed by contributions from the Exchequer. My favourite mechanism would be a seed capital contribution to each pension plan, paid at an early age and set at an amount that on pessimistic assumptions about investment returns and optimistic ones about life expectancy, would accumulate over a working life to provide an income at least as high as the state dole-pension would have been. As a nice touch, you might want to exempt the new arrangements from inheritance tax, so that when people die before they draw their pension, or before exhausting their fund, they can pass on their savings to their children's pension plans. The quid pro quo would be an increase in the retirement age.

It is not very difficult to devise a new form of saving vehicle along these lines. Nor is it too difficult to see that such an arrangement would cost far less than the current system. In hard economic terms there is no difference between paying a dole-pension to a retiree every week for the rest of his life and giving him a one-off lump sum equivalent (currently, at least £60,000). That is the principle behind annuities. However, there is a significant difference if instead you pay a one-off lump sum to a young person so that it can accumulate over his working life to a fund equivalent to a dole-pension by the time he has reached 70. One-off lump sums to a few hundred thousand people born in the same year would cost far less than weekly dole to several million pensioners for as long as they live. Increased tax relief and exemptions would be only a fraction of a fraction, and would be offset by reduced administration costs and the wider benefit of higher levels of saving and investment.

State pensions cost around £100 billion a year, but within a few decades this can be expected to more than double in real terms and to keep rising. The cost of "super-ISA" pensions would depend on the age of the recipients but would probably be about the same as the current cost and, crucially, it would remain stable. (Older people, having a shorter time to accumulate savings, would require a larger seed contribution than younger people, but there would be fewer of them as the grim reaper thins their ranks, so it would most likely all balance up.) Hence, if we had this sort of state pension system, by about 2040 (when I will be collecting my pension) it would save the state at least £100 billion. Which is quite a nice little windfall for our children, don't you think?

The Coalition, of course, would claim that they are doing something. They have "faced reality" and "taken the hard decisions". And in fairness, the Coalition has accelerated the increase in retirement age to 70 and implemented a footling scheme of auto-enrolment for personal pensions. These measures are little different from ideas proposed by Gordon Brown. They still assume that people will save themselves by saving for themselves—even at a time of austerity when it is harder for them to do so. Raising the retirement age by itself will not solve the inherent flaws of the dole-pension, which are a matter of demography. (Measures to reform welfare by tackling its alleged abuse by feckless unmarried mothers seem particularly unfair: the feckless mothers are, at least, doing something to relieve the coming demographic crisis.)