The pensions crisis and flexibility
Opinions17 April 2008
In many countries and for many people, pension provision has been something that is done for them and not by them. They are automatically enrolled in state or occupational schemes that determine how much they contribute and what they receive. They have little individual control over contribution levels, the way their savings are invested, and how they are drawn down in retirement. This model of collective pension provision has come under severe challenge, with the closing of occupational defined benefit schemes and proposals in many countries for some form of individual accounts within the state system.
To see why traditional pension systems are now coming under attack, and to evaluate proposals for a more flexible and individualised approach, it is worth understanding why they were set up in the first place. A useful place to start is the state pension system, which tends to be highly standardised, and which accounts for much of pensioner income. Even in the UK, which has a particularly well-developed private pension system, 60% of pensioner income comes from the state.
A compulsory state pension is very much cheaper to provide than an individual voluntary pension. If people are forced to participate, there is no need to spend money on advertising. If little choice is provided, it is cheap and easy to inform participants of their rights, and little need to spend money providing tailored advice on choices. With no individual accounts but simple entitlement rules, record-keeping is reasonably straightforward. If provision is universal, the insurer's concern about adverse selection - attracting only the people who are most likely to claim - is banished. If the state is backing the scheme, the chance of people being left destitute by fraud or mismanagement of their savings is small.
The state schemes typically provide little choice. But the costs of choice are high. The costs (administration, marketing, customer support) associated with an individual pension product are around 1.5%/year. Taking account of the long period over which a pension is built up and spent, this means that for every £10 paid out in pension, around £3 is spent in administration and other costs. So the savings from a compulsory universal scheme are not insignificant.
At the same time, a public pension system allows many other social and political objectives to be fulfilled. It can achieve near universal coverage. It ensures that people, however myopic and imprudent, will have a decent standard of living when they age, without recourse to charity or poverty relief. The system can be designed to be mildly or strongly redistributive. The UK basic state pension for example is unusually redistributive, with contributions being proportional to earnings while the pension itself is flat rate; the state second pension is going the same way. In almost all other countries, the level of the state pension is more strongly related to the level of contributions.
There are many other subsidies built in to the state system. With an automatic spouse's pension, there is a transfer from singles to those who are married. Smokers subsidise non-smokers. Traditionally, early retirees are subsidised by those who go on working. It is possible to have subsidies between one group and another because participation is compulsory - those who are doing the subsidising cannot opt out. While some of these cross-subsidies are accidental, many reflect collective social priorities that would be more difficult to advance through other mechanisms.
The advantages of a standardised system also apply to defined benefit occupational pension schemes. There is little choice and administrative costs are low. While membership is not generally compulsory, most of the cost is normally borne by the employer, and by opting out the employee loses the employer contribution. Within the constraints of the law (and they have got much tighter over the years) employers can build in cross-subsidies that advance their commercial objectives - rewarding long-serving employees at the expense of early leavers, and those who experience substantial wage progression as opposed to those who never get promoted. They can induce early retirement or encourage people to stay on, depending on the industry and the labour situation.
With all these strengths, why is collective provision under strain everywhere? It is too glib just to point to demographics, and say that they are under strain because the cost is increasing. It is true that people are now much more likely to live to pensionable age, and when they get there, they draw their pension for much longer. 50 years ago, 30% of 20 year old males in England and Wales could expect to die before reaching the age of 65; now the proportion is less than half that. If they did get to 65, they could expect to live less than 12 years; now they can expect to live at least for a further16 years, and this figure is rising fast. The cost of providing a pension has risen. But that does not explain why there should be a shift from collective schemes to individual accounts. Demographic trends mean that more money has to go to pensions if pensioner living standards are not to fall, but they do not obviously favour one route over another.
The reason is subtly different. Offsetting the great cost efficiencies of collective provision, there is the major disadvantage of inflexibility. The pension scheme member has no control over contributions; for those who have debts to pay off or who have children to raise, there may be much better uses for the money than putting it into a pension. The member has no control over investment strategy. In a defined benefit scheme, the member is getting a fixed return. This may suit the risk averse, but elementary economics suggest that many people would prefer to take more risk in exchange for higher returns on average. Collective schemes often impose severe constraints on when the pension is taken and in what form.
This inflexibility is becoming an increasing handicap for a number of reasons. One is the rise of consumerism and the decline of deference. If a pension is seen as deferred pay rather a recognition of a continuing obligation by an employer to look after his retired workers, or the state to look after its senior citizens, then the member is bound to demand control over it and access to it. Demographics have also been important in the following sense. When pensions were first introduced, those people who did reach the age of 65
were often exhausted, and in ill-health. The assumption that they could not earn a living was generally valid. Now, many people at 65 are fit and healthy and capable of working. Designing a pension system on the basis that people will not work beyond some specific retirement age has become extremely wasteful.
But perhaps the most important reason for the pressure on collective schemes is that with increasing longevity, pension contributions have risen as a proportion of wages and pension wealth has also grown. Inflexibility that might be tolerable if it extends to say 5% of salary may become intolerable when it extends to 15%. A pension pot that allows a couple to live on two thirds of final salary, and protects them against inflation will have a value, at retirement, of over 10 times final salary. It is not surprising that with such large amounts of money at stake people want more access to it and are less happy about "their" wealth being used to subsidise someone else.
Steps have been taken, with both occupational and state pensions, to give members more flexibility and choice. But it is difficult to go far without losing the advantages of a standardised system. Costs rise. Members need advice on how to exercise their choices. Choices have to be constrained or else members must be allowed to make bad choices. Choice with constraints makes the system complex, hard to understand and expensive to administer. The ability to cross-subsidise also falls as people are given more options.
This creates a dilemma in designing a pension scheme. One possible direction is the one we appear to be going towards in the UK. That is a two tier system: a compulsory, collective state system with little choice, that aims to provide no more than a tolerable standard of living for all in retirement - say at 40% of average earnings from 65 or 70 - with a voluntary system that allows and even encourages people to save so that their standard of living in retirement is not much lower than it was while they were working. Under the voluntary system, people would be free to choose how much and when to contribute, how to invest the money and how to draw it down in retirement.
The fear is that many people will chose to do minimal saving, and will then suffer a very substantial fall in living standards when they retire. This is the so-called savings gap. The big question is whether this is an important public policy issue. If people who are earning above average earnings chose not to save, but rather consume their earnings during their working life, and live on the same income in retirement as people who earned much less than they did, does that really constitute a crisis?
To see why traditional pension systems are now coming under attack, and to evaluate proposals for a more flexible and individualised approach, it is worth understanding why they were set up in the first place. A useful place to start is the state pension system, which tends to be highly standardised, and which accounts for much of pensioner income. Even in the UK, which has a particularly well-developed private pension system, 60% of pensioner income comes from the state.
A compulsory state pension is very much cheaper to provide than an individual voluntary pension. If people are forced to participate, there is no need to spend money on advertising. If little choice is provided, it is cheap and easy to inform participants of their rights, and little need to spend money providing tailored advice on choices. With no individual accounts but simple entitlement rules, record-keeping is reasonably straightforward. If provision is universal, the insurer's concern about adverse selection - attracting only the people who are most likely to claim - is banished. If the state is backing the scheme, the chance of people being left destitute by fraud or mismanagement of their savings is small.
The state schemes typically provide little choice. But the costs of choice are high. The costs (administration, marketing, customer support) associated with an individual pension product are around 1.5%/year. Taking account of the long period over which a pension is built up and spent, this means that for every £10 paid out in pension, around £3 is spent in administration and other costs. So the savings from a compulsory universal scheme are not insignificant.
At the same time, a public pension system allows many other social and political objectives to be fulfilled. It can achieve near universal coverage. It ensures that people, however myopic and imprudent, will have a decent standard of living when they age, without recourse to charity or poverty relief. The system can be designed to be mildly or strongly redistributive. The UK basic state pension for example is unusually redistributive, with contributions being proportional to earnings while the pension itself is flat rate; the state second pension is going the same way. In almost all other countries, the level of the state pension is more strongly related to the level of contributions.
There are many other subsidies built in to the state system. With an automatic spouse's pension, there is a transfer from singles to those who are married. Smokers subsidise non-smokers. Traditionally, early retirees are subsidised by those who go on working. It is possible to have subsidies between one group and another because participation is compulsory - those who are doing the subsidising cannot opt out. While some of these cross-subsidies are accidental, many reflect collective social priorities that would be more difficult to advance through other mechanisms.
The advantages of a standardised system also apply to defined benefit occupational pension schemes. There is little choice and administrative costs are low. While membership is not generally compulsory, most of the cost is normally borne by the employer, and by opting out the employee loses the employer contribution. Within the constraints of the law (and they have got much tighter over the years) employers can build in cross-subsidies that advance their commercial objectives - rewarding long-serving employees at the expense of early leavers, and those who experience substantial wage progression as opposed to those who never get promoted. They can induce early retirement or encourage people to stay on, depending on the industry and the labour situation.
With all these strengths, why is collective provision under strain everywhere? It is too glib just to point to demographics, and say that they are under strain because the cost is increasing. It is true that people are now much more likely to live to pensionable age, and when they get there, they draw their pension for much longer. 50 years ago, 30% of 20 year old males in England and Wales could expect to die before reaching the age of 65; now the proportion is less than half that. If they did get to 65, they could expect to live less than 12 years; now they can expect to live at least for a further16 years, and this figure is rising fast. The cost of providing a pension has risen. But that does not explain why there should be a shift from collective schemes to individual accounts. Demographic trends mean that more money has to go to pensions if pensioner living standards are not to fall, but they do not obviously favour one route over another.
The reason is subtly different. Offsetting the great cost efficiencies of collective provision, there is the major disadvantage of inflexibility. The pension scheme member has no control over contributions; for those who have debts to pay off or who have children to raise, there may be much better uses for the money than putting it into a pension. The member has no control over investment strategy. In a defined benefit scheme, the member is getting a fixed return. This may suit the risk averse, but elementary economics suggest that many people would prefer to take more risk in exchange for higher returns on average. Collective schemes often impose severe constraints on when the pension is taken and in what form.
This inflexibility is becoming an increasing handicap for a number of reasons. One is the rise of consumerism and the decline of deference. If a pension is seen as deferred pay rather a recognition of a continuing obligation by an employer to look after his retired workers, or the state to look after its senior citizens, then the member is bound to demand control over it and access to it. Demographics have also been important in the following sense. When pensions were first introduced, those people who did reach the age of 65
were often exhausted, and in ill-health. The assumption that they could not earn a living was generally valid. Now, many people at 65 are fit and healthy and capable of working. Designing a pension system on the basis that people will not work beyond some specific retirement age has become extremely wasteful.
But perhaps the most important reason for the pressure on collective schemes is that with increasing longevity, pension contributions have risen as a proportion of wages and pension wealth has also grown. Inflexibility that might be tolerable if it extends to say 5% of salary may become intolerable when it extends to 15%. A pension pot that allows a couple to live on two thirds of final salary, and protects them against inflation will have a value, at retirement, of over 10 times final salary. It is not surprising that with such large amounts of money at stake people want more access to it and are less happy about "their" wealth being used to subsidise someone else.
Steps have been taken, with both occupational and state pensions, to give members more flexibility and choice. But it is difficult to go far without losing the advantages of a standardised system. Costs rise. Members need advice on how to exercise their choices. Choices have to be constrained or else members must be allowed to make bad choices. Choice with constraints makes the system complex, hard to understand and expensive to administer. The ability to cross-subsidise also falls as people are given more options.
This creates a dilemma in designing a pension scheme. One possible direction is the one we appear to be going towards in the UK. That is a two tier system: a compulsory, collective state system with little choice, that aims to provide no more than a tolerable standard of living for all in retirement - say at 40% of average earnings from 65 or 70 - with a voluntary system that allows and even encourages people to save so that their standard of living in retirement is not much lower than it was while they were working. Under the voluntary system, people would be free to choose how much and when to contribute, how to invest the money and how to draw it down in retirement.
The fear is that many people will chose to do minimal saving, and will then suffer a very substantial fall in living standards when they retire. This is the so-called savings gap. The big question is whether this is an important public policy issue. If people who are earning above average earnings chose not to save, but rather consume their earnings during their working life, and live on the same income in retirement as people who earned much less than they did, does that really constitute a crisis?
