Many people think there is a pensions crisis in Britain. There is a widespread perception that our average pensions are low, that poverty amongst the 11m pensioners is high, and that all this will get worse because the population is ageing and pensions are becoming ever less affordable. Fortunately, much of this is wrong. Some of it is so wrong as to be laughable-the whole notion of some sort of crisis stemming from an ageing population is spectacularly wide of the mark and some of the "solutions" to the problem positively dangerous. None the less, many of the concerns reflect real issues, but thinking they are pension issues is often a mistake. Poverty amongst pensioners is significant but it stems largely from poverty earlier in life. Why are a large number of people poor in old age? Largely because they have been unable to save enough during their working lives-which reflects the distribution of skill and productivity and the way in which it is remunerated.
THE FACTS
According to the latest official statistics, the proportion of pensioners living in low income households is just under 25 per cent. (The definition used here of a low income household is one with less than 60 per cent of median income after taking account of housing costs.) Just under 25 per cent is a depressingly high fraction. But it is not a number that is significantly different from the incidence of poverty amongst the general population. On the same definition, around 20 per cent of working age people live in low income households; and the proportion of children living in low income households is over 30 per cent. It is no longer the case that poverty amongst pensioners is disproportionately high. In the early 1960s, almost 50 per cent of the poorest 10 per cent of households were pensioners. Now under 20 per cent of the very poorest households are pensioners, according to the Institute for Fiscal Studies (IFS). Over the last 20 years the average real income amongst pensioner families has increased by about 3 per cent a year. This is much faster than the growth in real average earnings-thanks, in part, to the increase in relatively generous occupational pensions over the last 20 years. (About 6.5m pensioners, or 60 per cent of the total, are now receiving occupational pensions in addition to the basic state pension; a further 10m current employees are paying into schemes.)
Is the "pensions crisis" then that the levels of pensions we currently pay are not sustainable? If the crisis is supposed to reflect an inability of future governments to continue paying state pensions at around current levels then it is illusory. Overall resources spent on the provision of retirement income by the state is currently just over 5 per cent of GDP (excluding the cost of pensions to public sector workers). The basic state pension remains by some way the largest element of this spending. In 2000-2001, state spending on the basic state pension was around 3.6 per cent of GDP. The remaining 1.6 or so per cent of GDP that gets spent by the state on providing retirement income comes partly through spending on income support (now called the minimum income guarantee, Mig), partly spending on the (now being replaced) state earnings related pension, Serps, and partly on various allowances for pensioners (winter fuel, Christmas bonuses, widows' pensions and so on). So at the moment the government channels about 1/20th of GDP directly to the retired. Is this sustainable? Yes. Projections of likely future levels of total direct state spending on pensions and other support for the retired suggests that over the next 50 years there might be a slight increase in overall levels of spending, with the ratio of spending to GDP perhaps rising to 6 per cent of GDP by 2050. (These projections are based on the preservation of the basic state pension in real terms-it would not be indexed to earnings but it would not be eroded by inflation.)
The projections above already take account of the large increase in means tested benefits that are planned by the government. The plan is that while the basic state pension will be flat in real terms, pensioner incomes will be made up to a substantially higher level for those who have few other resources (see box). It is intended that the Mig will rise in line with average earnings. And because the Mig is set at around 60 per cent of median income then, assuming that one takes that as a definition of the poverty line, the prevalence of poverty amongst pensioners is likely to decline significantly. The IFS has estimated that the combination of the Mig and the new pension credit could reduce poverty amongst the over 65s by almost 50 per cent. So if one thinks that the way to define pension problems is in terms of how much poverty there is amongst pensioners, then the notion of there being a looming pensions crisis seems quite wrong.
MEANS TESTING AND TARGETING
But there are problems. The government hopes that, increasingly, pensioners will rely upon private pension income-occupational pensions, personal pensions, stakeholder pensions and accumulated savings. Currently around 60 per cent of the overall resources available to pensioners come from the state and 40 per cent are from private saving. The government's aim is that the state's share should fall to 40 per cent and that private assets should generate 60 per cent of retirement incomes. One of the big questions about pensions is whether this will happen.
Another key question is whether Labour's strategy-which should significantly reduce poverty amongst pensioners-will also undermine incentives to save for those on lower and middling incomes. Before considering this issue, let us consider where we start from in terms of private saving for retirement. In 2001, just under 50 per cent of working-age people made some sort of contribution to a non-state pension-in addition to the compulsory contribution that most employees (11 per cent of earnings) and employers (12.8 per cent) make through national insurance. That 50 per cent figure was slightly down on the average for the previous five years. Indeed, the distribution of private savings in Britain is stunningly lopsided. In 2000, 27 per cent of all households had no savings at all; 50 per cent of households had financial savings of less than ?1,500. Amongst households with gross weekly income of under ?300, around 40 per cent of households have no financial savings at all; and around 60 per cent have accumulated financial savings of under ?1,500. In 1997, one in ten households had no assets of any kind-that is no housing, no pensions, no life assurance, no financial assets. At the beginning of the 1980s, that figure was one in 20.
At the heart of Labour's strategy is the idea that state support will be channelled to those with limited resources-in other words means testing will play a much bigger role. The Mig will steadily increase relative to the basic state pension if, as is planned, the Mig is indexed to average earnings while the basic state pension is flat in real terms. Those who have small amounts of private savings will find that their receipt of the top up will be smaller the greater the past savings. Under the current system there is, in effect, a 100 per cent tax rate on the first part of private income since top up benefits are withdrawn pound for pound up to the level of the Mig. Thereafter private income is taxed in the normal way.
If this system were to persist, there would be a big disincentive to save for an increasing proportion of the population. The government is painfully aware of this. Next year the pension credit will be introduced. This is described by the government as a subsidy to saving. In fact, it is a 40 per cent tax rate on the income from private savings up to a given level-or rather a withdrawal of 40p of Mig for each pound of private saving. Beyond that level, the marginal tax rate then declines for those with higher saving. This sounds bad, but for some pensioners it will give an enhanced incentive to save since they currently face a 100 per cent Mig withdrawal. The bad news is that far more people in future will face at least a 40 per cent marginal tax rate on income from private saving-as well as the loss of 40p of Mig for each pound of income from savings, income tax also has to be paid on the money so the marginal tax rate for many of the retired will be nearer 60 per cent.
What the overall impact will be on aggregate savings for retirement is not clear. What we have is an enhanced incentive for those who in the past have done very little saving, but a reduced incentive for a large number of households who in the past have done some saving.
Is all this a big deal? Yes. Calculations by the IFS suggest that just over 50 per cent of pensioners would be eligible for the pension credit if it were to be introduced today; this figure could rise to just over 70 per cent by 2025 and to over 80 per cent by 2050. At the moment, a minority of pensioners face a 100 per cent tax rate on the first part of private saving because of means tested benefits. Government plans, if implemented, will mean that the vast majority of pensioners will face at least a 40 per cent marginal tax rate on private savings because of means tested benefits. To some extent the impact on saving will be offset by the compulsion to contribute to a second pension-either the second state pension or a private one. But by not working at all people can avoid paying national insurance and still get the Mig in retirement.
Is this a big problem? That depends on how one views the disincentives that are an almost inevitable side effect of any attempt to channel resources to the least well off. If one takes the extreme position that any disincentive to work or save is unacceptable-and that is a position which some otherwise sensible people, such as researchers at the Institute for Public Policy Research, seem to take-then, of course, the introduction of the pension credit and its interaction with the Mig is disastrous. But that cannot be the right way to think about the appropriate degree of redistribution within our society. Any attempt to redistribute resources to the least well off will create disincentives. Even if one aimed to generate above poverty levels of pension incomes for all by boosting the universal basic state pension, there would be disincentive effects created elsewhere through the higher taxes required.
Recent calculations by the IPPR suggest that setting the basic state pension in line with the proposed Mig would mean that government expenditure on benefits for pensioners would rise from just over 5 per cent of GDP to 8.6 per cent of GDP by 2050. This is relative to a base case (with the proposed pension credit and Mig) of expenditure rising to 6 per cent of GDP by 2050. Income tax now brings in around ?100 billion a year. This is just over 10 per cent of GDP. If income taxes were to yield an extra 2.6 per cent of GDP they might need to be about 25 per cent higher.
The bottom line is that governments that want to redistribute find it hard to do so without creating disincentives somewhere. The sensible question is what is the optimal trade-off between the benefits of redistributing income towards the least well off and the costs that the disincentives bring in terms of lower output, less work, fewer savings and so on.
Recent work I have undertaken with James Sefton, of Imperial College, attempts to calculate what the optimal amount of means testing might be when we take into account the negative effects on saving and work. Our results suggest that the optimal system might be one in which the pension you receive from the government declines with the amount of private income from savings you have. In other words there is means testing but the taper rate-the rate at which benefits decline with private resources, effectively the tax rate implicit in means testing-is substantially lower than the 40 per cent implied by the government's pension credit. We estimate that that the taper rate should be around 20 per cent rather than 40 per cent.
COMPANY PENSIONS
What about the incentives of employers to encourage pension saving through occupational pension systems? Things are changing fast in this area. A combination of rising life expectancy, falling returns on financial assets and the unexpected hit on pension schemes the government introduced in 1997 (reducing tax breaks on dividend income, worth an annual ?5 billion for pension funds), has meant that many companies are moving away from providing generous, defined benefit, occupational pensions. Is this the real crisis?
Not really. The first point here is that if all companies are doing is moving away from defined benefit, occupational pension schemes-that is schemes where what you get is linked to how many years you worked and what your final salary is-toward defined contribution systems, then it is not clear that it is something to worry about. Defined contribution schemes are ones where contributions earn a market rate of return and the pension you receive in retirement will depend on the value of the fund you have paid into, rather than on how high your salary is when you get close to retirement. From the point of view of risk diversification, there is good reason for thinking that the defined contribution scheme-where what you get depends on the rate of return rather than your final salary-is actually preferable. If you are thinking about your pension in your mid-thirties, you have no secure way of knowing what your final salary will look like; there is a risk that it will be less than you had hoped, which means that the salary-linked pension will be lower too. The defined contribution schemes at least spread this risk and such schemes are also more portable for people who move jobs a lot.
But this is rather too optimistic a conclusion, since companies in recent years have been moving toward defined contribution schemes while, at the same time, often taking the opportunity to reduce their own contribution rate. Is this the real pensions crisis? Again the answer is not really. If a company decides that it will reduce one aspect of the benefit package it offers its work force then, in a competitive labour market, it had better increase some other component of the overall remuneration package. So if a company switches the occupational pension scheme from a defined benefit one to a defined contribution one and, at the same time, cuts its own contributions by, say, 4 or 5 per cent of salary a year then, other things being equal, it will find some people leaving and recruiting more difficult unless it compensates by increasing wages and salaries by about 4 per cent or 5 per cent.
In fact, what has really prompted companies to switch towards defined contribution systems, and in some cases to pull out of offering company pensions at all-25,000, or one in five occupational schemes have closed in the last five years-is the realisation that the cost of offering pensions has turned out far higher than expected because of lower rates of return and higher life expectancy.
Employers are currently compelled to pay national insurance on their employees' earnings-the figure is rising next year to 12.8 of pre-tax earnings, although with an upper cap-but it is unlikely that any government would impose on employers another level of compulsory pension payment. That means that in the future a larger proportion of the non-state pension will be provided by private insurers rather than employers. This in turn raises questions about the selling, or mis-selling, of private pensions. The worst of the mis-selling is probably now behind us and the regulations have been tightened. But the cap on charges for government approved (but privately provided) "stakeholder" and other pensions means that the selling agents will have less incentive to spend time with the lower-paid people who will need the most advice.
CRISIS OF INEQUALITY NOT OF AGEING
A considerable proportion of the population who should be saving for old age and who will increasingly need to make those decisions themselves, as occupational pension schemes become less generous and less common, are either unable or unwilling to do so. The underlying problem here is the inequality in earnings power-something that has increased substantially over the last 30 years. In 1975, the real net disposable income of someone in Britain at the 10th percentile (someone who found that 90 per cent of the population earned more than them) was about one third of the income of someone at the 90th percentile. By the year 2000, the person at the 10th percentile earned one quarter the income of the person at the 90th percentile. In 1981, 13 per cent of people lived in households where disposable income (before housing costs) was below 60 per cent of median income; in 2000, that fraction was 18 per cent.
What has happened in Britain is what has happened in the US. Thirty years ago there were substantial numbers of relatively well-paid jobs for people with few qualifications. That is no longer true. The gap between the wages of those with and without qualifications has widened. This may partly reflect the decline of manufacturing which in the past had more middle-income jobs and more jobs for those with limited qualifications.
If there is a pensions crisis in Britain, it really stems from the effects of this widening distribution of earnings power. This problem ultimately is best addressed through improving the environment in which children are brought up-better education and parenting.
But another widely touted "solution" to the supposed problem of ageing is to encourage us to have more children or to allow much greater immigration. This seems to me utterly misguided as a solution to a non-problem. We have already seen that the government can pay for its pensions programme with only a small increase in the proportion of GDP directed through the state to the retired. So there is no crisis of affordability which requires that we sharply increase the number of earning tax payers. Those who focus exclusively on one particular measure of the scale of the "pensions problem"-the ratio of those of pensionable age to those of working age-seem to have convinced themselves that the real issue is that we will not have enough people in this country. It is worth considering where that line of reasoning goes.
Perpetual rapid growth in the population is almost certainly required to hold the demographic structure constant (that is the relative number of people in different age categories) in the face of rising life expectancy. The ratio of people over 65 to the total population now reflects the numbers of people born over the past 100 years. To hold that ratio constant requires that population keep on rising. In the 20th century, the population of Britain increased from around 38m to around 60m-an increase of about 60 per cent. A similar rise over the 21st century would take the population to about 100m. One of the things we are not short of in Britain is people. We are short of land, we have poor transport infrastructure, pressures on the environment are immense and open space around the cities in which most people live is under threat. The idea that we need to increase the population at faster rates than in recent decades-and keep the population growing at that faster rate in perpetuity-is simply mad.
A far more natural response to those who insist on seeing the "crisis" in pension provision as stemming from there being more old people is, of course, to redefine what we mean by old. Indexing the age at which people receive state pensions to life expectancy is sensible in the longer term. This would mean, say, raising the retirement age to 70, so that the fraction of people's lives that was spent in retirement would not keep on rising. That would also mean that current government proposals on pensions would look even more affordable. It is wildly implausible that a strategy of seeing Britain's population inexorably rise for ever more is a better strategy than gradually moving the retirement age in line with movements in life expectancy.
