Long live a proper pensions policy: Andrew Dilnot and Paul Johnson welcome the new thinking in Labour's social justice report

Nearly all of us will have to rely on pensions of some sort to provide the greater part of our income in later life. The way in which these are provided, their security and fairness, should concern us all.

Two reports published this week have put pension policy back at the heart of the political agenda. The first, from Labour's Commission on Social Justice, talks of a new 'minimum pension guarantee' and of ensuring that everybody has a good second pension. The second was published on Tuesday by the Securities and Investments Board, the watchdog of the financial services industry. It tackles one of the most worrying failings of the current system: the poor advice which has led many people to choose the wrong pension plan, particularly through inappropriate selling of personal pensions.

Both raise fundamental questions about how we should develop pension provision. We should be thankful that these issues are being considered now and not in the next century, when current problems could prove disastrous for future pensioners.

The main change proposed by the social justice commission is the introduction of a state pension guarantee. By integrating the tax and benefit system for pensioners, the commission intends that all pensioners should be guaranteed a level of income that is (an unspecified amount) above current basic pension and means-tested income support levels. This would be achieved by providing a supplement to the basic pension for those who needed it, to bring their total income to the minimum level. In other words, receipt of extra money on top of the basic retirement pension would depend on income.

If this sounds familiar then that is not surprising. Although the commission claims, rather disingenuously, that this 'would mean an end to means testing for pensioners', it would in fact be providing a means-tested supplement to the basic pension. And that is no bad thing. As the number of pensioners grows, and the gap between rich and poor pensioners widens, we can no longer afford ever-bigger basic pensions for all. We need to target resources where they are most needed.

The commission's proposal would place on a more acceptable basis what the government is doing anyway. Means-tested income support already provides a top-up to the basic pension for the 1.5m poorest pensioners. Reforming the system as the commission proposes would integrate that top-up into the whole tax and benefit system, and probably make it more acceptable to recipients.

The commission proposes that the level of the pension guarantee be increased in line with net earnings growth in the economy, typically around 2 per cent per year faster than prices. No recommendation is made on the flat-rate retirement pension, which is to be left to the 'discretion' of the government of the day. This inevitably means that the policy of price-indexing (rather than earnings-indexing) the flat-rate pension, in place since 1980, will continue. Over the past 15 years the single pension has fallen from 20 per cent of average earnings to 13 per cent, and it will be below 10 per cent by 2020. The gap between the flat- rate pension and the more generous pension guarantee will grow steadily, so in tackling poverty among the elderly we will become more reliant on the means-tested pension guarantee, and less on the flat-rate pension.

The acceptance of a continued decline in the relative value of the flat-rate pension, and the need for greater reliance on improved means-tested benefits, is a huge leap for the Labour Party to contemplate. The alternative, of proposing again, as in the 1992 election manifesto, immediate increases in the flat- rate pension and subsequent uprating in line with earnings growth, presupposes substantial tax increases or reductions in other spending areas. The commission has, rightly in our view, recognised the massive change in the circumstances of the pensioner population in the past half-century, which makes a universal flat-rate benefit a far less appropriate primary means of tackling pensioner poverty than it was in 1942, when Beveridge reported.

One odd aspect of the social justice report is the suggestion that while the pension top- up would be reduced to take account of income from private pensions and earnings, it might not take account of interest income on investments. This is foolhardy. It would encourage people to save in forms other than pensions. It could encourage a generation of pensioners to rely on savings that run out when they reach extreme old age, rather than on pensions that insure against such longevity.

But even a more generous guaranteed pension paid by the state is unlikely to be enough to make people comfortable in retirement. Additional pensions - such as occupational pensions, personal pensions and Serps - will always be necessary to achieve this. The commission recognises this and so makes it clear that all employees and the self-employed should have such additional pension arrangements - properly regulated.

Given that only about half of employees have access to company-run occupational schemes, and that Serps is unlikely to be very generous because of the burden it would then place on future taxpayers, personal pensions will be one important part of this supplementary provision. That is why the finding of SIB that companies have been mis-selling these products is so disturbing. Without adequate compensation, hundreds of thousands of people will be left with a lower income in retirement than they should have had.

Why is that? The worst cases occurred when people were advised to leave their occupational pension schemes. This proved a bad idea in many cases, not because personal pensions are inferior, but because employers typically do not pay money into them. So leaving an employer's scheme usually means giving up the employer's contributions.

There are few circumstances in which shifting from an occupational scheme with significant employer contributions to a personal pension scheme without them would make sense. But for mobile or young workers, a personal pension with contributions from an employer will often be better than a traditional final-salary occupational scheme, which is best for long-staying employees with high final salaries. And for many employees an employer-provided occupational pension is not available at all: for this group personal pensions provide one of the few routes to adequate pension coverage.

Different forms of private pension have differing characteristics, and each will suit different types of individual. The appalling mis-selling of personal pensions to one group for whom they are inappropriate should not be taken to demonstrate that they do not work and cannot be helpful.

And if the Borrie commission recommendations are implemented, or if current government policy continues, all except the least well-off will need to rely on some form of private provision in retirement. This must be properly understood and regulated. Maybe we are now moving in that direction.

The authors are at the Institute for Fiscal Studies.