ISBN 92-2-112027-9
ISSN 1020-5322
First published 2000

Download the document in Adobe Acrobat's Portable Document Format (PDF)

Contents

Foreword

Because of the decline in fertility rates and the increase in life expectancy in a growing number of countries, the world population will age much faster in the next decades than previously. These demographic developments have important consequences for labour markets. This paper suggests that with a shrinking supply of young entrants, the older workforce will have to remain additional years in the labour market, a consequence, among other things, of the rising financial problems of retirement systems. This requirement to remain active longer periods than before, is casting doubt on the viability of the hitherto pursued early retirement policies. Instead of labour supply reduction by early retirement, labour supply increases by a prolongation of working life are now on the political agenda. The ageing trend will also have to be accompanied by a change of attitude towards the older workforce and proactive policies for boosting training, productivity and decency of work for the older workforce.

The extension of working life in OECD countries, will result in an increase in employment rates and alleviate pressures on pension systems. However, this development should not be seen as a win-win situation only. It means that the baby boom generation will not enjoy the same standards as the generation of their fathers and mothers, if pension systems are not adequately reformed. Working longer and retiring later, while paying higher pension contributions for reduced pensions, can be seen as a departure from a redistributive regime, channelling parts of the fruits of growth to early retirement with sometimes high wage replacement rates. The new welfare state emerging seems to be more and more based on the redistribution of wealth through revenue from actual employment. It implies a general change from "welfare" to "work". While this may be seen as unavoidable, the social and economic implications of this "paradigmatic change" have yet to be analysed.

The present report analyses demographic trends in OECD countries and reflects on their labour market consequences. It shows some of the new approaches in public policies and company strategies to tackle the challenges of ageing and presents some former ILO activities in the field. It ends with a set of proposals aimed at combining the search for productivity with decent work for the older workforce.

Werner Sengenberger
Director
Employment Strategy Department
Employment Sector

"There are many big issues to be faced in scrutinizing the proposals for revising the retiring age. That is a very big issue and I do not want to address it so off-handedly but I am just pointing out how conflicts are often seen when none might exist ... Indeed, the combination of the gut reaction to the effect that the source of the problem of an ageing population is that the old cannot work with the gut reaction that the young must lose jobs if the older people did work is to provide a hopeless impasse which rides just on unexamined possibilities, based on a simple presumption of conflict that may or may not actually exist. I am afraid quite a lot of thinking on labour economics is really governed by presumption of conflicts which have not been thoroughly examined."

Address by Mr. Amartya Sen,
Nobel Laureate in Economics
International Labour Conference, Geneva, 87th Session
15 June 1999

Introduction

Demographic developments such as the decline in fertility rates and increase in life expectancy and their implications seem to cast some doubt on the policies of labour supply reduction trough early retirement which were so important throughout the eighties and nineties. Indeed, the reduced inflow of young people into the labour market and the longer retirement periods going together with increased life expectancy have put pressure on the retirement systems, which have to cope with the trend towards early retirement and the increased inflow of the baby boom generation into retirement in general. The policy of reducing labour supply at the exit side of the labour market through early retirement (parallelled at the entry side by an extension of education) has already been restricted in a number of countries. Instead of labour supply reduction, labour supply increases are high on the political agenda today.

Several factors have played a role in this shift: present and especially future costs of early and regular retirement, expressed in increasing old age dependency rates, reduced inflow of the young in the labour markets, the "juvenilization" of the aged, with large cohorts of healthy, employable retirees, and a growing awareness of companies of a loss of experienced human resources through an indiscriminate retirement policy.

The decline of the number of active people per retiree poses in fact the problem of the sustainability of the policies hitherto undertaken. From a pure cost view it is certainly timely that the trend towards earlier exit is reversed. However, this is easier claimed than done. While it seems that all the factors cited above point towards the necessity to alter former policies, there are many barriers, which do not allow an easy u-turn in labour supply reduction policies. For example, insufficient growth, linked often to a low employment intensity of economic growth, does not allow the expansion of labour markets in order to absorb (or even maintain) older workers in sufficient numbers. Companies, faced with continuous restructuring, are rather inclined to downsize than to expand employment and there is no indication that public employment is on the rise. And in both public and private companies, downsizing is still accompanied by early retirement which give companies adjustment flexibility and secure income and status to former workers. It is therefore not yet clear, how firms could maintain (or hire) older workers without facing redundancy problems. Despite Amartya Sen's important claims that there must not be a trade-off between employment for the old and employment for the young, some substitution between older and younger workers can be expected if there is no sufficient employment expansion.

In addition, in many countries, unemployment is not yet sufficiently low to allow a labour supply expansion. That is also the reason why, despite a reorientation of the debate, the trend towards earlier retirement continues, often supported by the social partners. The support of the social partners for this measure will remain as long as the functional arguments, that the trend of early retirement has led to a loss in human resources, remains unconvincing in the face of companies looking for inexpensive adjustment both in monetary and social terms. This form of coping with the problem is popular among workers who prefer to retire early. Therefore, both companies and their older workforce have to be convinced that longer active periods result in gains.

These problems indicate that the reversal of policies poses a formidable challenge to policy makers, the social partners, company managers and the older workforce. However, while the situation is serious, it is not hopeless. Policy action has been taken in most countries to react to the problems associated with an ageing labour force. Besides policies of increasing retirement age and/or decreasing retirement benefits, they include incentives/disincentives for companies maintaining/dismissing older workers, partial retirement and company plans for the employment of older workers. While public policy can set incentives/disincentives companies must engage in pro-active policies of changes in work organization and working time and training to allow older workers to remain in productive work. The social partners have to provide their constituents with guidance. They could play a decisive role in a reversal of the labour supply reduction strategy.

"Active ageing" has many dimensions such as the general participation of the older population in society and in the economy, in family life and in salaried and self employment, or in voluntary employment. The following paper mostly discusses the issues involved in the participation of the older workforce in employment for the OECD countries. It shows the dimension of the active and non active population, develops the problems of falling retirement age and its implications for the financing of the retirement systems. It shows that countries face the problem of demographic ageing to different degrees, and that not all have had the same policies towards the aged. It raises the question of the policy options available and gives an overview of some of the public and private policies to cope with the challenge of ageing. It ends with some policy recommendation, based on the ILO's labour standards and former work undertaken by the office in the area.

1. Population ageing: Trends

The ageing of populations is a global phenomenon. With the continuation of fertility decline and increase in life expectancy, the population of the world will age much faster in the next decades than previously. Fertility decline is the major reason of population ageing (United Nations, 1998). In practically all developed countries fertility is significantly below the level necessary for the replacement of generations ⁽¹⁾. As table 1 shows, according to United Nations population estimates and projections, the TFR is estimated to be, in 1995-2000, at or below the level of 2.1 children per woman in all the OECD countries (except in Turkey and Mexico) and in 11 OECD countries, this rate is estimated to be less than 1.5 children per woman.

Table 1 Range of total fertility rate in 1995-2000

Source: United Nations Population Division, World Population Prospects: The 1998 Revision.

In 1950 there were about 200 million persons aged 60 and over in the world, constituting 8.1 per cent of the global population. By 2050 there will be a ninefold increase and the world's elderly population is projected to be 1.8 billion people, about 20 per cent of the total population. Countries have different rates of demographic ageing. Developing countries still have a relatively young population while populations in industrialized countries are relatively old ⁽²⁾. As we can see in figure 1, by 2050, the more developed regions will have a very old population, with the proportion of older persons projected to increase to 33 per cent in 2050.

Figure 1

Source: World Population Prospects. The 1996 revision, United Nations

Europe is, and is projected to remain, the area of the world most affected by ageing. The proportion of older persons will increase from 20 per cent in 1998 to 35 per cent in 2050. By then, one in every three persons will be 60 years or above. Southern Europe, with a proportion of older persons of 22 per cent in 1998, is the world region with the oldest population. By 2050, its proportion of older persons will have reached 39 per cent. In 1998, the country with the largest proportion of old people in the world is Italy, followed by Greece, Japan, Spain and Germany. By 2050, the country with the largest share of old people will be Spain, closely followed by Italy. The other areas of the world most touched by ageing are, in decreasing order, Northern America, Oceania, Asia and Latin America and the Caribbean.

Figure 2 shows the prospective demographic transition in nine OECD countries. We can see that there is a large and rapid increase of the old-age dependency ratio ⁽³⁾. Italy is the country with the highest increase in the dependency ratio followed by Spain and Japan. This does not, in all cases, mean that the overall old population is dependent on transfer payments (pensions) because some are still economically active. However, given the trend towards earlier retirement and the arrival of the baby boom generation at retirement age, old age dependency is increasing and this trend will continue. The fact that less and less active people have to support more and more inactive people is an issue of major concern especially for the financing of retirement systems./P>

Figure 2

Source: World Population Prospects, 1996 revision, United Nations

The worker-to-retiree ratio shows a similar picture than the dependency ratio. ⁽⁴⁾ Table 2 shows that these countries will face a decline in the number of employed per retiree. In Ireland and Spain, it is expected that by 2020 there will be an average of only 1.5 employed per retiree.

Table 2 Worker to retiree ratios, selected OECD countries

Source: H.D. Steinmeyer, 1996.

2. Older workers and the labour market

2.1 Participation rates

As figure 3 shows, in most OECD countries, labour force participation rates of older workers (above 55) have declined markedly with the decline being more marked in Europe and in the United States than in Japan. Labour force participation rates of older people vary according to gender, education and the state of economic development of the country. The decline in rates of older workers is associated with the trend towards earlier retirement, influenced by increasing national per capita income. Longer education, shorter working lives and longer retirement periods are all consequences of increased wealth. In addition it was found that urbanization increases this trend.

As we can see in figure 3 included in the Annex, this decline in activity rates has been particularly marked among older men. Instead, female participation rates have been increasing, even among older women. This has partially offset the decline in male labour force participation rates. As the figure shows, the steepest decline in the participation rates of males between 55 and 64 years across the OECD have taken place in the Netherlands, Austria, France, Finland, France, and the United Kingdom. The less significant declines are found in Japan, the United States and Sweden. Japan is the country with the highest older males participation rates (around 95 per cent for the age range 55-59 and 75 per cent for 60-64) and Austria has the lowest (around 61 for the age range 55-59 and 13 per cent for 60-64).

Participation rates of older women are especially high in nordic countries. In Sweden, they have increased continuously since the 1950s and Sweden has now the highest participation rates of older women (around 80 per cent for the group 55-59 and over 50 per cent for the group 60-64). Japan and the United States have also relatively high levels (around 55 per cent for the 55-59 group and around 35 per cent and for the group 60-64). Italy, Spain, Austria, and the Netherlands have the lowest levels (between 20-25 per cent for the group 55-59 and under 10 per cent for the group 60-64). However, since the 1950s, Spain and the Netherlands have experienced some increase for the age group of 55-59.

It is the 60-64 age group which indicates most clearly the extent to which older men have dropped out of the active labour force. Within this age group, differences between countries are more clearly marked. Japan with over 75 per cent has the highest rate, followed by Ireland, Norway, Portugal, Sweden, the United Kingdom and the United States with rates between 50 and 60 per cent. Austria, France an the Netherlands have the lowest rates (around 25 per cent). For the age group over 65, participation rates are low for all countries. Only in Japan, the rate remains over 50 per cent.

If we break down the participation rates in its two components: employment and unemployment rates, the following picture emerges.

2.2 Employment rates

Over the last decade, like participation rates, employment rates for older men have continued to decline in many OECD countries. The decline has been more substantial in some countries than in others. Table 3 shows the decline in employment/population ratios of men and women aged 55-64 in 1980 and 1996. However, the magnitude of the decline varied markedly. The countries with the steepest declines were the Netherlands (-28.6), France (-26.7) and Spain (-21.6).The scale of the decline has been more limited in Japan (-1.6), Luxembourg (-2.2) and the United States (-5).

The employment situation of older women has seen more variability across the OECD area. Some countries have continued to see increases in labour force activity as new cohorts of women with more established patterns of labour force participation reach older ages. In other countries, such as Finland and France, there has been a clear decline in employment participation among older women.

Table 3 Employment/Population ratios for older workers (55-64), 1980-96

Source: Kalisch D. and Tetsuya A. "Retirement income systems: the reform process across OECD countries", 1997
1. For unemployment, data for the age group 55 to 64 refers to 55 and over
2. 1980 data refers to 1983
3. 1980 data refers to 1979
4. Unemployment rate for 1980 refers to 60 years and over
5. 1983 data refers to 1984
6. 1996 data refers to 1995

Table 4 shows a strong relationship between educational attainment of older workers and the employment rate. The higher the education, the higher the employment rate. As educational qualifications are often closely related to earnings capacity and job stability, these figures suggest that those with lower qualifications (and thus lower earnings) have a greater propensity to retire earlier.

Table 4 Employment rates of older workers (55-64) and level of education

Source: EC Employment Report, 1998

2.3 Unemployment rates

Table 5 presents data on the unemployment rates of older workers and the incidence of long term unemployment. We can see that, in most countries, the unemployment rate of older workers is lower than the overall unemployment rate. However, once older workers are unemployed they have a greater risk to be unemployed for a long period of time. Belgium, Finland, France, Italy, Ireland, Spain, Portugal and the Netherlands present long term unemployment rates over 60 per cent.

Table 5 Unemployment rates and the incidence of long term unemployment, 1996

Source: OECD, 1998a

3. Reasons for non employment of older workers

While employment characterises the middle phase of our lives, and retirement characterises the end phase, the transition between the two phases is not uniform. Not all older workers pass directly from work to retirement. A part from retirement, there are a number of non-working situations like unemployment, discouragement ⁽⁵⁾, invalidity or long-term sickness, and looking after the home or caring for a relative.

There is no clear cross-country pattern to explain the reasons why inactive males in the 55-64 year old age group have left their last job. According to the 1995 European Union Labour Force Survey, in 6 of the 15 countries, more than half of inactive men in this group claimed that the reasons were due to early or normal retirement. The country with the highest share is Austria, where almost 80 per cent of the older men left their job due to early or normal retirement. However, in Sweden, Finland, the United Kingdom and Spain, which are three countries that experienced particularly deep recessions in the early 1990s, the reasons were different. In Finland and the United Kingdom, the high rate of involuntary departures was equally split between dismissal or redundancy and disability or illness; in Spain the termination of fixed contracts also played an important role in addition to health and redundancy; whereas in Sweden the high rate was concentrated on dismissal or redundancy. Table 6 presents the results of the survey.

Table 6 Retired males aged 55-64 (1): Main reasons for leaving last job or business in the EU, 1995 (share of total)

Note: 1. Refers to persons aged 55-64 who are not in the labour force, but who had been in the labour force in the 8 years preceding the survey. 2. Other reasons include: education and training, compulsory military or community service and other reasons

Source: The European Union Labour Force Survey, 1995

Queen (1996) reviews for the United States the degree to which the trend towards early retirement reflects voluntary decisions. The results from surveys targeted to men over 65 years of age and taken at approximately 10-year intervals from 1940 to 1980 show that in the early decades, almost nobody suggested that he retired in good health because he wanted to. Health problems and job losses were among the main reasons for early exit. Over time, the proportion of those leaving voluntarily increased, and by 1980, exceeded the involuntary exits. Conclusions are: (i) more people are retiring voluntarily, depending on their wealth and on the financial incentives in the social security and employer pension plans and (ii) the distinction between voluntarily and involuntarily is vague; many older workers face unattractive alternatives in the labour market such as pay cuts or new jobs at lower pay and many leave the labour force.

In Japan (Mitani, 1990 and OECD, 1995a.), while the statutory retirement age has increased over time, a larger proportion of workers have left their firms before reaching this age. According to the Survey on Employment of Older Persons (1988), the proportion of the workers aged 55-59 who leave the firm before the mandatory retirement age increased from 3.3 per cent in 1983 to 4.5 in 1988. This trend was marked in large firms with at least 1000 employees, where the proportion increased from 4.4 to 6.9 per cent. This can be explained by the fact that workers who leave firms before the mandatory retirement age seem to enjoy some advantages. The cited survey shows that the most popular reason for retiring early is "favourable in terms of income including retirement allowance" (47.5 per cent of male retirees), followed by "recommended by firms", "not so physically strong", "better to change job while still young" and "dissatisfied with personnel management or promotion". A part from separation before mandatory retirement age, older workers also leave firms because of transfer ⁽⁶⁾. The number of transfers is increasing, reflecting the shortage of posts due to the ageing of the workforce, and the expansion of company networks due to the diversification or extension of business activities. The age until which employment is secured in the destination company is roughly the same as the mandatory retirement age in the original company.

4. Falling retirement ages

Over the past decades, most OECD countries have experienced a substantial drop in the average age at which individuals retire from the labour market. As we can see in table 7, in 1950, the average effective age of retirement for males was above 65 in all OECD countries except in Belgium and New Zealand. By 1995, that average had dropped to 59 years, with striking differences emerging between countries. Only in Iceland and Japan men continued to work on average well beyond the age of 65. Austria, Belgium, Finland, France, Luxembourg and the Netherlands arrived at effective retirement ages below 60.

Table 7 Estimates of the average age of transition to inactivity among older male workers

Source: Estimates by Scarpetta and Blöndal, 1998.

Table 8 shows that although the effective female retirement age is traditionallly lower than that for men, it followed a similar pattern over the last decades. A decrease of the female effective retirement age was recorded in the 24 OECD countries between 1950 and 1995. Currently, more than half of the OECD countries have a female effective retirement age below 60.

Table 8 Estimates of the average age of transition to inactivity among older female workers

Source:Estimates by Scarpetta S. and Blöndal S., 1998.

There is a direct relationship between older male retirement ages and older male participation rates. In the countries where the effective retirement age is below 60, less than 50 per cent of the male population is participating in the labour force. However, in Japan and Iceland, where the participation rate of the 55-64 years old is above 80 per cent, the effective retirement age is the highest of the OECD area. However, for women aged 55-64 years, the participation rate has remained stable or even increased, while their average retirement age has fallen.

As a result of this trend towards early retirement, few people in most OECD countries, are working after the age of 60 and almost none after 65. Obviously, longer life expectancy and better health has not been accompanied by longer working lives. To the contrary, people are retiring earlier and earlier.

5. Old age pension systems as incentive for early retirement

All OECD countries have established systems to support people in their old age. Typically people contribute to such schemes during their working life in exchange for income support after a certain age. Empirical research (Scarpetta S. and Blöndal S., 1998) shows that expected pension wealth usually falls with continued work after a certain age rather than being independent of it. These disincentives depend on: (i) the statutory and minimum ages of entitlement to pensions, (ii) the pension replacement rate, (iii) the pension accrual profile, (iv) the actuarial adjustment to pensions and (v) the private pensions.

5.1 The statutory and minimum ages of entitlement

As table 9 shows, in more than half of the OECD countries, male workers are entitled to public pensions at the age of 65. This statutory age ranges from a low of 60 in a few countries (Japan, France, Italy) to a high of 67 in some Nordic countries (Denmark, Iceland and Norway). The statutory entitlement age is often lower for females. Since the 1960s, most OECD countries have kept the statutory age unchanged; only a few countries have lowered the statutory retirement age, and a few countries have already started to increase it.

Table 9 Statutory age of entitlement to public old-age pensions

Source: Scarpetta, S. and Blöndal, S., 1998

There are good reasons to think that raising the statutory retirement age will raise participation rates of older workers. However, a rise in statutory age alone does not necessarily entail rising effective retirement age. If we compare tables 7 and 8 with table 9 we see that the effective retirement age has fallen despite unchanged statutory age. Therefore, in most OECD countries, workers have left the labour force before reaching statutory retirement age, showing that there is a strong preference for early retirement. However, the statutory retirement age may act as a barrier for those who remain in employment up to this age and wish to continue working. The opportunity cost ⁽⁷⁾ to continue in employment is usually very high.

Several OECD countries allow, under certain conditions, to access pensions before the statutory age. Many European countries have introduced seniority pensions for those who have been contributing for a long time and who have reached a certain minimum age for entitlement. Other countries allow older citizens to obtain their pension prior to the statutory age, subject to permanent actuarial reduction of benefits. The minimum age varies across countries but on average, it is 60 for males. Working beyond this age has a high opportunity cost due to forgone pensions and continued contributions while working with little or no gains in pensions. Also, the receipt of early pensions is often subject to permanent exit from the labour market. Raising the minimum age for entitlements to pension will have an important effect on effective retirement age only if extended early retirement options are not available any longer.

5.2 The pension replacement rate (PRR)

Table 10 presents gross PRRs that a 55 year old worker could expect to get at standard retirement age if he would continue working until then. We can see that PRR have risen over the last 30 years. However, they differ considerably across OECD countries. At one extreme there are countries where pensions can be expected to be close to the earnings level before retirement (Italy, Luxembourg, Iceland, Portugal, Sweden and Austria), or even to match or exceed earnings from work (Spain and Greece). At the other extreme there are countries (Australia and Ireland) where only about 40 per cent of gross earnings can be expected to be replaced by public pensions. However, for the majority of the countries the PRR indicator is in the range of 47 to 67 per cent.

Comparing tables 7 and 8 with table 10, we see that despite PPRs have risen over the last 30 years, effective retirement ages have fallen. There is no clear relationship between the evolution of the PRRs and that of effective retirement age. For instance in Japan, where participation rates of older workers are very high, the PRR has risen strongly. At the other extreme, France, where effective retirement age has fallen drastically, PRRs have increased modestly.

The structure of PRRs also seems to depend on income levels and the form of employment. The existence of flat pension components, maximum and/or minimum pensions implies that in many countries, replacement rates are higher for those with low earnings than for those with high earnings. This pattern of PRRs may partially ⁽⁸⁾ explain why early retirement is concentrated among those with low earnings capacity. In some countries, PRRs are also lower for the self-employed. For example, the self-employed in Japan are only entitled to basic old-age pensions and not to mandatory earnings-related pensions, which implies that the PRR for a single person with average earnings was only 17 per cent for the self-employed compared with 44 per cent for employees in 1995. This could explain the high proprotion of self employed among the older actives in Japan (30 per cent). The PRR for self-employed in general, and for farmers in particular, is also much lower than for employees at least in some European countries (e.g. Belgium, Germany, Greece). This is in line with the fact that the self-employed tend to work longer than employees.

Table 10 Expected old age gross PPRs: Synthetic indicator ⁽⁹⁾

Source:Scarpetta, S. and Blöndal, S., 1998.

5.3 The pension accumulation rate

The rate at which pension rights accrue has an important impact on the retirement decision. If the pension accrual rate is zero there are no penalties in terms of lower old-age pensions from withdrawing from the labour market. However, if the accrual rate is high there are incentives to continue working. Pension accumulation rates have sharply decreased since the 1960s. However, they differ significantly across OECD countries. In some countries, such as in Australia, pensions are unrelated to employment/contribution records but on length of residence, in which case the accrual rate with respect to employment is zero. In other countries, such as in Germany, the level of pensions increases over the whole working life. In between are countries where full pensions are earned relatively quickly, implying zero pension accrual rates for older workers.

The incentives embedded in high pension accumulation rates have been weakened in some countries by crediting some form of inactivity over working-age years as equivalent to covered employment. For example, special early retirement schemes involve continued contributions for old age pensions; disability benefit recipients do not experience any drop in their old-age pensions due to their non-employment status; and in some countries (e.g. Germany) the unemployed are entitled to pensions in the same way as the employed.

5.4 Actuarial adjustment to pensions and long-serving pensions

While maintaining a statutory age of entitlement, a few countries ⁽¹⁰⁾ allow older workers some flexibility in accessing their pension with some adjustment in the value of the retirement income. The earliest age at which pensions can be accessed is 60 and the adjustment to pensions last at maximum to 70. The most common adjustment is about 0.5 per cent reduced pension per month of early or deferred withdrawal. The adjustment factor is significantly higher in Japan.

In Italy, Greece, Austria and Germany pensions can be accessed before the standard retirement age if workers have contributed for a sufficient number of years. The contribution period required for such early withdrawal is 35 years. However, the earliest age at which such withdrawal is allowed varies considerably across countries ⁽¹¹⁾. A common feature of systems that allow for early pension withdrawal is that access is conditional on ceasing work, at least until the standard retirement age has been reached. In some of the countries which allow early pension withdrawal, the earliest age at which such benefits can be paid coincides with spikes in the probability of retiring: early-withdrawal options have thus clearly encouraged retirement.

5.5 Private occupational pension schemes ⁽¹²⁾

Private occupational pension plans also discourage work at older ages. In recent years, such plans have been expanded considerably and are considered to be part of the answers to financial problems of government pensions systems and to the expected widening pension gap. The role of private pensions varies among countries and their development is linked to that of government systems. In France, Switzerland and Hungary, these schemes are obligatory for employees (except for some categories and the self-employed)

The eligibility age for these schemes is often lower than the statutory age for public pensions. Neither the replacement rates nor the penalty for early retirement rate nor the penalty for early retirement are actuarially neutral and in this sense, their sensitivity to demographic developments is similar to that of public systems. As with public schemes, working beyond the pension age adds little to eventual benefits. This is what happens in the UK, especially for employees whose pension schemes often specify a retirement age lower than that of the public scheme. However, where schemes are of the defined benefit type and specify retirement at an earlier age than the public scheme, and especially where they permit early payment of pensions, the incentives to retire early increase.

6. Women's transition to retirement

In some countries women receive preferential treatment in terms of the age at which they qualify for state pension. As we can see in table 9, in half of the OECD countries, the statutory age of entitlement to public pensions is lower for women than for men. This favourable treatment that women receive may be justified on several grounds (OECD, 1995b). Historically, the retirement pension was awarded on a family basis. The logic behind retirement insurance was that the father would be the family breadwinner and that the mother would care for the home. One of the reasons why women obtained individual pensions at a lower retirement age in certain countries, could be to compensate them to some extent for the years that they had spent outside the labour force caring for children. It may also be based on traditional thinking that older women are physically weaker than older men. The OECD (OECD, 1995b) mentions other reasons such as that it was a way to secure a better match between retirement ages for a couple in which the woman was generally presumed to be younger than the man. It also notes that it may be a way to compensate women since women in earlier generations derived less satisfaction from employment because of the kinds of jobs they were given. However, currently, women are ever more eager to work and are occupying increasingly attractive jobs. It seems that there is no longer a rationale behind different retirement ages for men and women. The EU member states are gradually rising the standard retirement age for women in order to equalise men and women's retirement ages ⁽¹³⁾.

In the past, women's participation in early retirement schemes has been low in relation to their share of total employment. In Belgium, by December 1989 only 22 per cent of the beneficiaries of early retirement schemes were women. In France, women accounted for fewer than one third of early retirees. In the Netherlands, in 1990, 15.2 per cent of males between 55 and 64 were on early retirement. This contrasts with only 2.8 per cent for women (OECD, 1995a). There are several reasons for this. First of all, during the 1970s, the retirement age in many early retirement schemes was generally 60, which was the statutory retirement age for women in many countries. Also, during the 1970s, early retirement programmes were de facto targeted at declining manufacturing industries whose work forces were predominantly male. In other words, even when age conditions did not prevent women from taking early retirement under government schemes, the targeting effect made it harder for women to avail themselves of such schemes. Given the limited number of early retirement avenues open to women, some women may have been using the sickness schemes as a route to early retirement. This has been the case in the United Kingdom and Sweden (OECD, 1995b). However, women are also clearly underepresented in occupational invalidity schemes.

In the EU, there have been important reforms in this area and, as we have already noted, most member states have now equalised their pension age or plan to do so early next century. Equalization has been achieved mainly by raising women's pension age to that of men. However, women still encounter difficulties in accessing pension schemes because their paid and unpaid work patterns do not conform to the pattern of long-term continuous participation in paid work on which entitlement to those schemes is based. Women are disadvantaged by such requirements because many women have a more discontinuous work pattern. Also, in the EU, the numbers of women involved in part time work is very significant. The European Commission estimated that in 1994, 17 million of the 21 million part time workers in the EU were women. In relation to pay, there is a persistence of an earnings gap between men and women in every member state (Bulletin; 1994), and low pay tends to mean lower benefit rates .

A common rule in both state or employer sponsored national schemes and employer sponsored company schemes, is the requirement of a minimum period of membership to be satisfied before rights are acquired. Long qualifying periods disadvantage women. Also, the reference wage on which pension benefits are based can also be a disadvantage for women, depending on their career pattern. For example, if women end careers on a part-time job and the wage of the last job is the reference wage, such a disadvantage arises.

In the EU, there are two rules which compensate women for the difficulties they encounter gaining access to adequate pension provision because they do not comply with pension entitlement rules geared towards a particular pattern of paid work. The first type of compensatory rule focuses on sex or marital status or a combination of both. Provisions falling within this category include different pension ages, survivors' benefits, dependency additions and derived rights generally. The second rule, makes special provision for caring. The rule requires pensions schemes to treat periods of care ⁽¹⁴⁾ as periods of insurance. Also, basic pension systems relax the link between pension entitlement and engagement in full time, continuous relatively well paid work. For instance, basic pension systems provide pension on the ground of residence and nationality rather than on participation in paid employment.

The advantage of the compensatory rules and basic pension systems is that they enable women (and men) without a history of full time, continuos highly paid work to secure an adequate retirement pension income. However, compensation for care provisions does not protect the pension position of workers (women and men) engaged in part time, non continuous, low paid work. The expansion of flexible forms of employment make it likely that the numbers of such workers will increase rather than decrease in future years, thereby increasing the number of people with insecure or inadequate pension rights. Therefore, an adequate response to women's specific disadvantages in regard to pension cannot consist only in an equal statutory retirement age. elements of basic pensions such as guaranteed minimum pension entitlement are required.

7. Are reasons behind early retirement justified?

7.1 Do older workers cost more?

It is