Is it realistic to expect people to take personal responsibility for self-support during old age? Of course, there should be an infallible safety net. But it is widely accepted that the presence of a safety net may encourage some individuals to rely on the state rather than their own endeavours. This difficulty is at the heart of the current debate about pension reform. To what extent should provision of a pension be compulsory?

Advocates of compulsory provision by the state usually recommend one of two main alternatives: either a funded state pension, such as Singapore's; or a compulsory funded private pension of which the Chilean scheme is the prime example. The strongest argument for compulsion is that, if the state maintains a safety net, then some individuals will be more likely to rely on it rather than to provide for themselves. Therefore, so the argument goes, people who are capable of saving but do not or will not, should be required to do so. But, should the desire to avoid free riding outweigh all other considerations? What are the potential disadvantages of compulsion?

1. A compulsory pension is defended as a means of eliminating free riders, but it does not completely achieve that objective. Some people never earn enough to accumulate a pension fund. Some refrain from work. Some work in the black economy. Moreover, compulsion increases the incentives to work "off the books". Countries with compulsory schemes, such as Chile, still have to maintain a state guarantee funded from taxes.

2. Compulsion is applied to people who would have saved anyway. Many are compelled to save in a manner which may not be the best for them. A self-employed person, for example, may prefer to invest in his own business and such an investment may well prove to be a more sound method of provision than a compulsory contribution to an annuity.

3. Compelling individuals to contribute a fixed percentage of their income to a private pension throughout life assumes a stable pattern of employment. But the pattern of work has changed. Earlier in the century it was possible to assume that most people were employees, but this is no longer true. The labour market continues to change, with far fewer workers now expecting to spend their career with one employer and many more preferring self-employment. A compulsory scheme crowds out flexible alternatives that would allow adjustment to changing circumstances.

4. A compulsory scheme, based on a percentage of income, also fails to acknowledge the legitimacy of changes in saving and spending priorities over the lifecycle. The importance of the risks we face changes over time. It is legitimate to have different priorities when younger. For example, when a married couple have children to support then the death of the breadwinner would be very serious, as would incapacity for work. But when the children are grown up, early death is less important.

5. Inevitably a compulsory scheme must be regulated, and the regulations limit the access of new companies to the market and diminish the scope for innovative competition. In Chile, for example, a relatively small number of regulated companies have a guaranteed market. The result is that they do not need to try so hard.

6. To contribute to a pension by locking away savings for 20-30 years is not the most prudent savings strategy. To hand over money today to a private company knowing that you will be forced to buy an annuity in 2034 is inherently risky. The company might be a mutual at the moment, but what will it be in the future? Companies can and do increase their charges (which might be from 13 per cent to 39 per cent of contributions); investment performance varies; annuity rates fluctuate; policy towards bonuses may change; and the government might change the tax regime again.It is not short-sighted to want to avoid relinquishing control of your retirement income. It is prudent.

7. The idea of a pension assumes retirement from work, but not everyone wishes to retire, preferring to lead a "useful life" until they die or become incapable of work. Most people retiring at 65 are capable of working for many more years.

8. If the government sets a percentage of income that must be paid anticipating that it will produce a fund at retirement sufficient to yield an annuity of a certain amount, what happens when investment performance is poor or annuity rates are bad? In practice, the government may find that it comes under pressure to guarantee investment performance.

Thus, compulsion does not achieve its primary objective, to avoid free riding,but it has a harmful effect on other people, many of whom would have made sensible provision. These are arguments against compulsion but they are also strong arguments against tax concessions. If compulsion is a requirement to put money into a bad investment, tax breaks are an inducement to put money into a bad bet. The objective of policy should, therefore, be to create a regulatory framework which allows the maximum flexibility whilst guarding against fraud and financial abuse. Before suggesting how such a framework could be established, it will be useful to examine more fully whether it is feasible to rely on personal responsibility. The assumptions still being made by many politicians belong in the age of social determinism, whereas we are now in a new era of personal responsibility.

Are there other strategies that a prudent and unselfish person could pursue? And if so, why should the Government make it more difficult for them? I will suggest two such strategies. First, there is 'work until you drop'. A prudent person could plan to go on working until he was no longer able to do so. In some cases such a person would die while still working, and in others he would work until illness or frailty intervened. The main risk is that a relatively young person may become too ill or frail to work, and so a prudent person would ensure that he had a disability insurance policy, payable at any age, and which preferably paid a high income replacement benefit linked to salary. But, as we have seen, if an individual were compelled to pay a percentage of his income into a pension, his ability to provide for possible misfortunes would be diminished. A person pursuing a 'work until you drop' strategy would be self-sufficient and at no stage of his life a burden on other people. The former is a one-generation strategy, but it would be perfectly prudent for a family to follow a two- or three-generation plan. A family could build up assets - property, durable goods, shares, cash - with the intention of handing them on from generation to generation. Purchasing an annuity involves giving a capital sum to an insurance company which undertakes to pay an agreed sum per year until death. The insurer takes the risk that the capital will run out, but members of a family might prefer to take that risk themselves. If the oldest surviving generation, for example, opts to live on its capital rather than to buy an annuity, the risk of their capital running out before they die gives their children an incentive to diminish expenditure, perhaps by caring for their parents themselves. They might use the capital to build a granny annex, so that their parents were close at hand. Or they might purchase long-term care insurance to guard against the high cost of residential care.

Such a property-based strategy offers considerable flexibility for families of quite modest means. For most people, investing in the trading up or down, or using the property as collateral for a loan, cash can be made available at different stages of the lifecycle. It has also been the strategy successfully pursued by some of the poorest overseas immigrants to this country. Of course, parents could find not only that their capital had run out, but also that their children proved unreliable. However, the danger of falling back on the state safety net could be avoided by simultaneously pursuing the 'work until you drop' strategy.

Thus, there are alternative strategies which could be pursued without free riding, but is it possible for the government to create a new framework which maintains a safety net, minimises free riding, and encourages personal responsibility? The temptation to free ride is the result of offering unconditional benefit payments on reaching a certain age. The Government has accepted that this system creates a moral hazard, but it has chosen to deal with it by coercing people who do not need to be coerced, many of whom have made arrangements more suitable for their own circumstances than any pension.

A more effective method of removing the moral hazard would be to abolish the state retirement age. Thus, individuals of any age who relied on state benefits would be obliged to find a job or take part in a government work scheme. If a work obligation is attached to benefits at all ages, there is little temptation to take advantage. The obligation to work, obviously, should only apply to people who are capable of it and individuals who are too ill, injured or frail should be able to fall back on the safety net.

If the safety net were a system of transitional relief, as I have argued, this would mean that elderly persons claiming benefit would have to undergo a medical examination to show that they were unable to be fully self-sufficient. It can be argued that tests of fitness for work should not be necessary after a certain age. Perhaps so, but age 65 is too young. We now live much longer and enjoy good health well beyond 65 and an alternative approach would be to raise the pension age in stages, perhaps by six months per year so that in ten years it will be 70 and in 20 years, 75. Having achieved the relevant age, no test of fitness for work would be necessary and the basic state pension would be payable. Eventually, the pension age could be phased out altogether so that we would no longer become members of the demeaning category "old-age pensioner" towards the end of our lives. Instead of having a group identity imposed on us by the system, we would continue to play our part as individuals, each with a unique contribution to make.

Finally, the current tax regime for pensions should be revised. If compelling people to contribute to a pension is wrong, so too is offering them a tax inducement. A better system would be to replace it with the tax regime for ISAs, but with much higher investment limits. Schemes could then evolve with the minimum of regulation offering any combination of saving, protective insurance, investment for income or capital growth, longterm lock-in, short-term access, linkage or non-linkage to annuities. The regulatory framework would be lighter, a trend already under way now that benchmarks have been preferred over top-heavy regulation for ISAs. Vast opportunities would be opened up for innovative and enterprising companies. And it would maximise self-sufficiency and independence for the people whilst ensuring that no one could fall below a clear national minimum income.

From An End to Welfare Rights by David G. Green