Commentary: retirement and economic recovery.
One of the most striking ways in which the policy response to the recent recession differs from earlier practice is in policy towards early retirement. In the 1980s early retirement was seen as a 'solution' to high unemployment. In the current circumstances in the UK not only is the policy of raising women's state pension age continuing, but the new government seems likely to implement proposals it made in opposition for an early increase in the age at which the state pension is paid to men. In addition a number of other countries in Europe have announced plans to increase the state pension age in response to pressures on the public finances induced by the recession. Increasing the age at which people retire can achieve two objectives; it is likely to improve the public finances and it is also likely to increase potential growth for a period and the sustainable level of output permanently. After the shock to output and government budgets induced by the recent financial crisis, both of these can be seen as beneficial. In addition, raising retirement ages appears to be a believable commitment that financial markets will take more seriously than commitments on spending that can be reneged on after the election of a new government.
The policy of the early 1980s could be justified only by the assumption that there was a number of jobs to be shared around which, if not fixed, was independent of labour supply. This assumption, in combination with a belief that it was more socially useful to allocate these jobs to younger workers, drove a wave of early retirement schemes. Since high unemployment was a political problem, for unemployed workers to convert into workers who had retired early yielded a political bonus in terms of lower unemployment figures. The view that the stock of jobs is essentially fixed is sometimes called the 'lump of labour' fallacy by those who believe that markets work. The alternative view is that if labour supply increases, then for a period wages are likely to be lower than they otherwise would have been, and new jobs will be created, and a new higher equilibrium level of employment will be reached. This may take time, but experience suggests that markets normally do work in this way.
This policy of promoting early retirement was widespread in Europe. In the United Kingdom it received a second impetus from the perceived state of private pension funds. In part as a result of the sharp stock market increases between 1982 and 1999, these funds seemed to show large surpluses and trustees did not generally object to the surpluses being used to provide generous early retirement arrangements for scheme members. Retirement ages in the UK are not particularly low, as we can see from table l, and they are rising. In 2009 men in the UK retired on average at 64.6 whilst women retired at 62.4, both of which represent a noticeable increase over the average between 2002 and 2007 in the table. (1) The Scandinavians generally have high retirement ages, and Sweden is a good example of this, with men working on average beyond the state pension age (which we denote 'official' although this is a misnomer in the UK). Retirement ages in Spain and the Netherlands have been a little lower than in the UK. France has noticeably lower retirement ages for men than does the UK. The Spanish and the French have announced their intention to raise their official retirement ages to 67 and 62 respectively, but the French clearly enjoy leisure. Recent discussions in the UK of retirement ages in Greece (and Italy) have focussed on the low retirement ages for public servants. The associated pension arrangements are costly, but these ages do not reflect the average for society, and clearly many retired public servants in these countries go on to further work whilst receiving a pension. The Greeks plan to change this arrangement, as do the Italians, mainly in order to improve the public finances. There may, however, be initial adjustment problems for these economies because both plan declining employment in the public sector and, if job tenure is lengthened there, then hiring will be limited.
Over the past twenty years mortality rates of people aged sixty and over have fallen in a way which has no historical precedent, with the obvious consequence that the cost of supporting an ageing population is increasing. Life expectancy of men reaching the age of sixty has risen by three years in each decade. Although the average age of retirement has risen marginally since the early 1990s, expected retirement has lengthened by at least five years. The situation is compounded by the fact that a baby boom took place in the UK between 1946 and 1970, leading to an increased proportion of old people even before taking account of rising longevity. Funded private sector pension schemes should be run so that they are not affected by fluctuations in birth rates. But the UK's schemes have been affected by declining mortality rates, which could not have been predicted twenty years ago. Poor stock market returns since 1999 which, if not forecastable, were at least a risk against which prudent pension schemes should have made provision. Thus, even if it might be seen as desirable, the UK is not in a position to address the problem of rising unemployment by inducing early retirement, as was the approach in the 1980s.
In this Commentary we summarise results which suggest that policy should aim to achieve the exact opposite of what was done in the 1980s, and that such an approach can play a significant role in reducing the fiscal burdens faced by many advanced countries. We begin by discussing motives for retirement. We then proceed to discuss the implications of later retirement.
Reasons for retirement
A basic economic assumption is that people are likely to want to smooth their consumption over time. If the marginal utility of consumption is diminishing, then people can raise their welfare by keeping consumption roughly constant instead of having high consumption in one year and low consumption in another year. Much the same argument might be thought to apply to leisure time. Thus, the observation that people choose to work on a concentrated basis for only a part of their lives rather than more evenly throughout their lives needs some explanation. At the start of people's working lives they may take gap years or spend some time settling into a job, while retirement may well be a gradual process.
The concentration of work can be partly explained by three economic phenomena. The first factor is that for some people earning capacity may decline with age. This means that, as people age, the cost of work, in terms of leisure foregone, becomes increasingly expensive relative to the benefit they gain from working. Working becomes less attractive and, when people feel they can afford an acceptable living standard without working, they retire. The second reason for people to choose the bunching of leisure in retirement is the potential for receipt of a pension, which means that their possible consumption is increased and therefore its marginal utility is lower at any given level of earnings. The marginal utility of leisure is much less affected, making it more likely that people will choose not to work. The third factor is that concentrated working increases people's earning capacity as skills and knowledge are temporarily enhanced. These factors provide reasons for people to take relatively more leisure late in life and thus to retire.
A number of implications follow this. First of all, people will tend to retire when they feel they can afford to. People with low earning capacity will retire if they feel they do not lose very much by living on benefits even if this happens before the state pension age. And people who have accrued substantial wealth, whether as pension rights or in other forms, may also choose to retire ahead of the state pension age. Sefton, van de Ven and Weale (2008) show that a theoretical model of rational decision-making leads to this pattern and that it is also observed in the data. Broadly speaking, low earners and high earners tend to retire earlier than the middle earners. For everybody the provision of state benefits associated with old age makes it less likely that people will choose to work, with the influence the strongest on those with least in the way of other resources, i.e. the poorest.
These observations immediately allow us to identify factors which are likely to encourage people to retire later than they currently do. First of all, a delay in the date at which normal state benefits are available has the effect of reducing the present discounted value of everyone's prospective future income and is likely to promote later retirement. But the effect is bound to be most marked on the poorest people, because, in proportionate terms, it is a much larger reduction in their wealth than it is for rich people. Secondly, poor investment returns, perhaps a consequence of a more general reduction in rates of return or as a result of taxes on income from capital, will mean that people need to save higher proportions of their labour incomes to achieve any given level of retirement consumption. A probable response to lower returns is both that people Will reduce their consumption in retirement relative to that in working life and also that they will delay retirement. This effect is bound to be more marked on rich people than on poor people since the latter do not save very much anyway. The age of retirement that people choose will also depend on the expected generosity of the state benefits in their retirement, with a more generous path for future benefits inducing people to retire earlier. An announcement in the Emergency Budget in June that pensions would be uprated in line with earnings, inflation or 2 1/2 per cent, whichever were to be the highest, was perhaps more generous than wise, but as with all political commitments, it can be revisited.
Implications of later retirement
Over the past three years the government budget deficit has risen from 2 3/4 per cent of GDP (on the Maastricht definition) to probably over 11 per cent this year. Even on current fiscal plans the debt stock on the same basis will almost double as a per cent of GDP. As Barrell and Weale (2010) stress, we are borrowing from our children, reducing what they can consume. There are a number of ways of reducing this borrowing, some of them involving consuming less now, and some involving working more now in order to be able to consume more now. Our children will also be better off than they would have been if we work more. At a time of fiscal stringency it is clear that policies which reduce budget deficits are attractive. Raising the state pension age has a direct and favourable fiscal effect, in that outgoings on state retirement pensions will be reduced. Barrell, Hurst and Kirby (2009) suggested that these benefits would represent only about a quarter of the gains (excluding the impacts on debt interest payments which cumulate over time) to the public purse. People's earnings are typically higher than the state pension, partly because choice about working hours is relatively limited. In addition, if work has any disutility, to achieve the same level of welfare as they would have had on the state pension then individuals will have to choose to earn more than the pension they would have received. (2) Thus, without a state pension, people who choose to work rather than not working are likely to work for longer than is necessary just to make up for lost income. As the level of income in the economy would be higher, income tax receipts would rise. The scale of the increase would depend on the impacts on output.
A rise in the age at which people retire, induced by a rise in the state pension age, an increase in age at which other benefits become more generous and the abolition of the Default Retirement Age, as proposed in the budget, will induce people to work longer. It will also allow those who are currently constrained to retire earlier than they desire to work longer in future. If those who are enabled or induced to remain in work have similar habits over hours and participation then, as Barrell, Hurst and Kirby (2009) suggest, labour input into production will rise by around 1 2/3 per cent for each additional year on the state pension age. (3) Barrell, Hurst and Kirby (2009) analyse a situation where the announcement of the changes needed precedes their implementation by three years, and hence the increase in labour supply takes place at a time when the labour market is less stressed than it now is. The empirical studies of the UK labour market on which their work is based suggest that the increase in labour supply will be fully absorbed into work within four years (Barrell and Duty, 2003), and that any increase in unemployment would be small and temporary. As a result of an increase in retirement ages of one year, output would be at least 1 per cent higher than it would otherwise have been in seven years time. This is a major gain in a situation where the financial crisis has probably made us all around 4 per cent poorer than we had expected. (4) If incomes are around 1 per cent higher than otherwise, then income tax receipts will also be higher, and Barrell, Hurst and Kirby (2009) suggest that these increases will have about 50 per cent more impact on the government finances than the direct savings on pension payments. Of course the government might decide in 2016 that it wishes to use the extra revenues to reduce income tax rates rather than pay off the debt stock more rapidly. We would consider that decision to be unwise as we have to move as rapidly as is reasonable to reduce debts in order to leave us more room to deal with the next crisis, whenever that might occur. (5)
Increased incomes and later retirement ages will mean that the consumption of goods and services is likely to increase. This will mainly follow from the higher levels of incomes, but will also be a result of a realisation that less saving is needed for a shorter retirement. Since consumption of goods and services is taxed, this has a further favourable budgetary effect, and Barrell, Hurst and Kirby (2009) suggest that the effect on the budget of increases in indirect taxes will exceed that of the fall in pension payments. This effect is likely to be larger in future with a higher VAT rate as announced in the recent Budget. Anticipation of longer working lives and hence of higher lifetime incomes is also likely to raise the consumption of younger people, as well as those near retirement or who are working longer. This effect is unlikely to be large, and Barrell, Hurst and Kirby (2009) suggest that if we announce a one year delay in retirement to be implemented in three years then consumption would immediately increase by 0.1 per cent a year more than it otherwise would have done for the three years before implementation. Even though this would be small, it would give a small boost to demand now. This demand effect would be augmented if employers realised the implications of longer working lives and started to invest in order to ensure the necessary capital is in place to use with the additional workers. Such an effect is possible, but we should not rely on it in the current conjuncture.
The overall impacts of raising the retirement age on the public finances depend upon how the government chooses to use pension savings and higher tax revenue. It would be wise to allow government investment to rise in line with the increase in output that longer working lives might bring, but decisions have to be made over paying down the debt or reducing the cuts to services that the current expenditure plans are projecting. Barrell, Hurst and Kirby (2009) calculate that if working lives are increased by one year then before the end of the decade the public finances would improve by two thirds of a per cent of GDP, allowing debt to be paid down more rapidly. This would require government consumption plans to be unchanged. The rate at which the improvement takes place depends upon the assumption that an announcement is made now that retirement ages for both men and women would slowly rise by one year from three years ahead. (6) Similar calculations are undertaken in Barrell, Hurst and Kirby (2010) for the Euro Area countries, where they show that three years on retirement ages (about two years on effective working lives) would reduce deficits by 2 per cent of GDP after a decade, and allow government debt to be reduced by 40 per cent of GDP in thirty years or so.
Retirement, working and trend output
The sustainable level of output depends upon the supply of labour in the economy, the skills of individuals, the capital they have available to work with and the technology they have at hand. It is hard to change skills and technology quickly, and the capital stock adjusts only slowly. The recent crisis has widely been seen as producing a repricing of risk and hence an increase in the rate of discount used when evaluating future income flows from investment. This will mean that less capital will be used in the production process, all else being equal. As a result, demand is weaker in the short run and this is reflected in lower sustainable output in the long run. The supply of labour can be augmented in a number of ways. Between 1995 and 2008 the population of working age rose by three million in the UK, with 80 per cent of this being driven by net migration from Europe and elsewhere. With constant participation rates, this will have raised the labour force by around 10 per cent and will have added almost l per cent a year to trend growth. Going forward, raising the pension age for women from 60 to 65 (and hence the effective age from 62.4 to 65) will add about 0.1 percentage points to potential growth each year between now and 2020. Trend growth would rise by the same amount over the next decade for each additional year we manage to increase the age at which people retire. Shortening retirement to the same proportion of expected adult life (7) in 1990 would raise the sustainable level of output by around 5 per cent, raising consumption levels in both working lifetimes and in retirement by similar amounts. It would also give us the possibility of a much less painful debt reduction process than we now contemplate.
Making choices over consumption and investment
The Labour government put in place a programme of spending reductions that was the most stringent since the end of the Korean War, and the current government has strengthened the consolidation programme. It is necessary to pay down the debt stock, and the objective set out in the Budget to achieve a net debt to GDP ratio of 40 per cent is achievable, as was discussed in Barrell and Kirby (2010). It is clear that plans are still intentionally vague as the early part of a new regime can be seen as a period of negotiation with the voters. Ideas will be put forward and evaluated. Discussion will be had along the lines of 'if you want to ring fence health you have to cut another budget by 40 per cent (or pay higher taxes)'. One of the issues to be negotiated over is the length of working lives. If the retirement age is raised in addition to the currently announced programme of tax rises and spending reductions then the government will give itself more room to manoeuvre. The savings on pensions and other benefits for the retired and the extra tax revenues generated by higher output can be used to increase consumption by our children by paying off the debt stock more rapidly. It can be used to increase consumption of privately provided goods by cutting taxes in five years' time, or it can be used to increase the planned provision of publicly provided consumption (and investment) goods. If we plan to work longer, consumption will be higher; the polity has to choose when it consumes the gains, and what goods are consumed.
Barrell R. and Dury, K. (2003), 'Asymmetric labour markets in a converging Europe: do differences matter?', National Institute Economic Review, 183.
Barrell, R., Hurst, I. and Kirby, S. (2009), 'How to pay for the crisis or macroeconomic implications of pension reform', NIESR Discussion Paper no. 333.
--(2010), 'The macroeconomic implications of pension reform' in Franco, D. (ed.), Pension Reform, Fiscal Policy and Economic Performance, Bank of Italy.
Barrell, R. and Kirby, S. (2010), 'Medium-term prospects for the public finances', National Institute Economic Review, 212, April, pp. F60-7.
Barrell, R. and Weale, M. (2010), 'Fiscal policy, fairness between generations and national saving', NIESR Discussion Paper no. 338, and Oxford Review of Economic Policy, 26 (I), pp. 87-116.
Flemming, J.S (1988), 'Debt and taxes in war and peace: the closed economy case', in Arrow, K.J. and Boskin, M.J. (eds), The Economics of Public Debt, Palgrave Macmillan.
Khoman, E., Mitchell, J. and Weale, M.R. (2008), 'Incidence-based estimates of life expectancy of the health for the UK: coherence between transition probabilities and aggregate life tables', Journal of the Royal Statistical Society Series A., pp. 203-22.
Sefton, J., van de Ven, J. and Weale, M. (2008), 'Means-testing retirement benefits: fostering equity or discouraging savings', Economic Journal, 118 (528), pp. 556-90.
(1) The average effective age of retirement is defined as the average age of exit from the labour force during a 5-year period. Labour force (net) exits are estimated by taking the difference in the participation rate for each 5-year age group (40 and over) at the beginning of the period and the rate for the corresponding age group aged five years older at the end of the period. The official age corresponds to the age at which a pension can be received irrespective of whether a worker has a long insurance record of years of contributions.
(2) The observation that consumption falls at retirement can also be explained by the idea that people face work-related expenses which are counted as consumption. Taxes are collected on many of these.
(3) This depends upon the size of the cohort and the hours people work as they approach retirement, which vary depending on the year in question for the former and the comparison group for the latter. Barrell, Hurst and Kirby (2009) assume those near retirement work around two-thirds of normal hours. There is limited evidence that productivity declines with age, at least before 70, except in strenuous manual occupations, and hence increases in labour supply from those above 60 are likely to be as effective as from other sources.
(4) This is the estimated impact in both the National Institute forecasts in this and recent Reviews and in Treasury and OBR projections.
(5) The most common large shocks to the public finances are wars and financial crises, and these almost always have a negative impact. There are very few shocks that have a large and positive impact on the public finances and hence between wars and crises we need to set policy to pay off debt rather than sit with a constant debt stock waiting for the matching positive surprise. Another way to put this, following Flemming (1988) is that debts are a random walk with upward drift if left to themselves, and hence policymakers have to ensure they drift down in between shocks.
(6) They analyse a situation where retirement is delayed for 1 1/2 years, raising effective working lives by one full-time equivalent year. In a forthcoming report for the Department for Work and Pensions, we look at a variety of scenarios around these issues.
(7) Khoman, Mitchell and Weale (2008) suggest that only 60 per cent of the increase in life expectancy since 1990 has been reflected in healthy life expectancy. Hence we have to choose between keeping the proportion of life in work constant, or the proportion of healthy life in work constant. As we have to raise the resources to pay for increased unhealthy lives we have chosen to concentrate on total life.
Table 1. Average retirement ages 2002-7 Men Women Effective Official Effective Official United Kingdom 63.2 65 61.9 60.0 Sweden 65.7 65 62.9 65.0 Spain 61.4 65 63.1 65.0 Netherlands 61.6 65 61.3 65.0 Italy 60.8 57 60.8 57.0 Greece 62.4 58 60.9 58.0 Germany 62.1 65 61.0 65.0 France 58.7 60 59.5 60.0 Source OECD. 'Official' refers to state pension age.
|Printer friendly Cite/link Email Feedback|
|Author:||Barrell, Ray; Holland, Dawn; Kirby, Simon|
|Publication:||National Institute Economic Review|
|Date:||Jul 1, 2010|
|Previous Article:||The UK economy.|
|Next Article:||Fiscal multipliers to assess consolidation plans.|