Older People and Transitions from Employment to Nonemployment: International Perspectives and Policy Issues

Abstract

The Organization for Economic Co-operation and Development (OECD) has identified population aging as one of the most important challenges facing OECD countries and has highlighted the need for people to work longer and for job prospects for older workers to be enhanced. This article provides a summary review of a recent OECD report, Ageing and Employment Policies—Live Longer, Work Longer, as a platform to highlight differences between countries in demographic profiles and projections and in patterns of formal labor market participation among older workers. Drawing on selected information from a broader evidence base, it unveils important differences between countries in the scale of demographic and associated labor market challenges. It also explores factors affecting labor market transitions among older workers and age-related and other barriers to paid work among older people, emphasizing the diversity of experience between individuals. Finally, it highlights some strategic challenges for policy.

La Organización de Cooperación Económica y Desarrollo (OECD) identifica el envejecimiento de la población como uno de los retos más importantes que enfrentan los países miembros y ha enfatizado la necesidad que tiene la gente de trabajar durante más tiempo y la conveniencia de mejorar las perspectivas laborales para los trabajadores más viejos. Este artículo suministra una reseña abreviada de un informe reciente de la OECD, Ageing and Employment Policies—Live Longer, Work Longer (Políticas sobre envejecimiento y empleo – Viva más, trabaje más años), como plataforma para destacar diferencias entre países por perfiles demográficos y proyecciones, lo mismo que por los patrones de participación de los trabajadores de más edad en el mercado de trabajo formal. Utilizando información selecta de una más amplia base de evidencia, el artículo deja ver importantes diferencias entre países en la escala de los retos demográficos asociados y del mercado laboral. También se exploran los factores que afectan las transiciones del mercado laboral entre trabajadores más viejos y las barreras relacionadas con la edad y de otro tipo, con la paga por trabajo entre gente de mayor edad, enfatizando la diversidad de experiencia entre individuos. Finalmente, se destacan algunos retos estratégicos en términos de políticas.

Key Words:

• aging
• demographic change
• labor market participation
• policy
• retirement

关键词:

• 老龄化
• 人口统计变化
• 劳动力市场的参与
• 政策
• 退休

Palabras clave:

• envejecimiento
• cambio demográfico
• participación en el mercado laboral
• políticas
• jubilación

Until a hundred years ago, most people in Western economies continued in paid employment until death. Gradually throughout the twentieth century, pensions were introduced as a means of supporting people after they left employment. With the establishment of state pension ages (SPAs, also known as official retirement ages) retirement became a recognizable stage in the life course. However, alongside declining fertility rates and longer life expectancy, from the 1970s a trend toward earlier withdrawal from the labor market became established, such that on average “effective retirement” (i.e., the average age at which workers over the age of forty withdraw from the labor force) occurred prior to the SPA being attained. For public policy this distinction between official and effective retirement age is an important one, especially in terms of ensuring economic support for older people in the context of an aging population and sociocultural and institutional differences in welfare regimes (Esping-Anderson 1990, 1999).

During the 1990s, extending working life in an aging society became an important policy goal, especially in Europe (Taylor 2004). (In the United States, concerns about aging at economy and society levels tend to be less developed, although there have been debates over altering Social Security policies, and at an individual level the association between retirement and crises in funding for health care is apparent.) As well as offsetting the negative potential effects of population aging on public budgets and economic growth, policies supporting this goal were heralded as promoting choice for older people, both in paid employment and in other aspects of their lives. Since the 1990s there is some evidence that early withdrawal from the labor market has been reversed (Moynagh and Worsley 2005): During the period from 1994 to 2004, employment participation rates among older people rose in almost all Organization for Economic Co-operation and Development (OECD) countries. The emphasis on promotion of choice in policies for older people also presages a breaking down of the traditional life course with three clear stages of education, paid work, and retirement, with greater fluidity and flexibility across each life stage.

Given the international scope and significance for public policy of aging and economic participation of older people, in 2001 the OECD Employment, Labor and Social Affairs Committee decided to carry out a review of policies to improve labor market prospects for older workers. The thematic review embraced twenty-one countries and separate country reports were produced by OECD analysts (using OECD and national data sources) and country specialists. Each country report analyzed the main existing barriers to employment of older people, assessed the coherence and effectiveness of policies in removing such barriers, and presented a set of recommendations on policy measures for government and social partners. The separate country reports were synthesised in an overview report entitled Ageing and Employment Policies—Live Longer, Work Longer (OECD 2006).

This article is organized into four sections. In the first section, key features of demographic change and labor market participation across OECD countries are outlined and the challenges that they pose for public policy are highlighted, drawing mainly on material from the OECD (2006) synthesis report. In the second section, the focus shifts to understanding transitions from paid employment to nonemployment at older ages with a view to understanding the barriers to such employment among older people and how these might be tackled. Again, this section draws on the OECD (2006) synthesis report, but supplementary information from other sources is also cited. However, due to space constraints, the discussion is necessarily selective. The emphasis in these first two sections is on contrasts in experience between countries, mainly at the aggregate level (although some contrasts by gender are outlined). In the third section, the importance of considering subnational variations in experience and between different subgroups of the population is discussed, with reference to the OECD (2006) synthesis report and other research. The fourth and concluding section of the article highlights the need for a range of policies to combat possible negative implications of demographic change, making special reference for illustrative purposes to the United Kingdom. This section emphasizes why population aging is an important issue for study by geographers and highlights selected possible avenues for further research beyond the specific concerns of the OECD (2006) synthesis report.

Demographic Change and Labor Market Change: Challenges for Policy

This section summarizes key demographic changes and trends in labor force participation among older workers outlined in the OECD (2006) synthesis report. It highlights the substantial potential that exists to counter smaller working age populations by increasing participation rates among older people in the face of prevailing demographic trends.

All OECD countries are experiencing population aging as a result of two factors working in opposite directions to increase the share of older people in the total population: decreasing fertility and increasing longevity. Population projections for OECD countries suggest that in Italy, Japan, and Spain, over one-third of the population will be sixty-five years or older by 2050, compared with one-fifth or less in the United States, Mexico, and Turkey (OECD 2006, 18).

The relative size of different age cohorts within the population (i.e., the demographic structure) has an important role in influencing the pace and scale of population aging across countries and associated labor market challenges. A key feature of population aging across OECD countries is that young cohorts are getting smaller than older cohorts. In Italy, Germany, and Japan it is already the case that there are fewer people between twenty and twenty-four years old than there are between sixty and sixty-four years old. By 2010 this is also projected to occur in France, by 2015 in Canada, and by 2020 in countries such as the United Kingdom and United States. In demographic and labor market terms, the aging of baby boom generations is especially significant. The baby boomers born after World War II are now reaching their sixties, and so are reaching the upper boundary of the conventional working age group. Over the last forty years or so the baby boomers have added to the size of the working-age population (typically from around fifteen to sixty-four years), but as they move from employment to nonemployment they represent a disproportionate increase in the population in the postretirement age group. At the same time, they serve to disproportionately reduce the size of the working-age population, as the younger cohorts entering the population of working age are smaller than those cohorts moving out at the upper end of the age scale. This has important implications for dependency ratios. Across the European Union (EU), the aged dependency ratio (defined as the ratio of the population aged sixty-five years and over to that aged between twenty and sixty-four years) is projected to rise from around 0.25 in 2000 to 0.4 by 2025 and around 0.55 by 2050. In Japan the projected increase in aged dependency ratios is even starker from a higher initial base, from just under 0.3 in 2000 to just over 0.5 by 2025 and in excess of 0.7 by 2050. In the United States, the increase is less marked but is nevertheless substantial, from just over 0.2 in 2000 to just over 0.3 by 2025 and nearly 0.4 by 2050.

OECD countries embrace a range of economic circumstances and welfare regimes, but from an economic perspective there are universal concerns about the pressures placed by a larger elderly population on public finances and on a reduction in growth in living standards in circumstances when a smaller working-age population has a larger older population to support, albeit in the context of a compensating reduction in young dependents. In some countries (for example, in northern Europe) the elderly population would be even larger if it were not for international migration of older people to southern Europe (Williams, King, and Warnes 1997; King, Warnes, and Williams 1998). However, the negative impact of population aging on economic growth can be offset by policies to influence future trends in the three key factors into which gross domestic product (GDP) per-capita growth can be decomposed: population, productivity, and participation.

Taking the three factors in turn, the first possible method for countering current trends is to seek to change the age structure of the population so that there are more people of working age relative to the younger and older age groups. Possible mechanisms for this are increasing fertility or increasing migration. The latter mechanism is more commonly used than the former. Several countries have adopted proactive migration policies in an attempt to counter demographic change and boost economic growth, as exemplified by the “Fresh Talent” initiative in Scotland (Scottish Executive 2004), which seeks to encourage migrants to settle in Scotland. However, although they are extremely important in economic development terms and also have implications for future demography, fertility and migration policies are not the key focus of this article. A second method to offset the negative impact of population aging on economic growth is to increase productivity. Again, policies to increase worker productivity are not the key focus of attention here, but they are likely to be of value in economic terms. Rather, our emphasis is on the third method of increasing participation in the labor market among older people to increase revenues and decrease some benefit claims (however, this relatively narrow focus does not discount the fact that many older adults who are not employed in the formal labor market make an important, productive, nonpaid contribution to national regional and local economies, as stressed by Droogleever Fortuijn et al. 2006). With a larger share of older people working, there is potential for greater economic growth, at the same time offsetting pressures on public finance by reducing the aged dependency ratio. What is the current picture in terms of labor force participation among older people? What is the potential for increasing participation rates in the future?

As previously outlined, there was a trend toward earlier retirement in the 1970s, such that in many countries effective retirement occurred, on average, prior to official retirement age. Table 1 contrasts effective and official ages of retirement in OECD countries in the period from 1999 to 2004. The countries are ranked in ascending order on the size of the difference (in number of years) between effective and official retirement ages. The data presented highlight a number of key features. First, in the majority of the countries (twenty-one of thirty) for which information is presented, effective retirement age precedes official retirement age. In two countries (Denmark and the Czech Republic), effective and official retirement ages coincide, and in seven countries continuing in employment after official retirement age is the norm. Second, the extent of this difference between effective and official retirement ages varies across countries. In Belgium, Austria, Luxembourg, Italy, and Finland, effective retirement occurs, on average, at least six years prior to official retirement age, compared with two years prior in countries such as the United Kingdom, Canada, and Australia and one year prior in the United States and Ireland, among others. Third, there are differences in the timing of official and, more especially, effective ages of retirement among countries. Sixty-five years is the general norm for official retirement age across the countries listed but is as low as fifty-eight years in Greece and sixty years in countries such as Japan and France, rising to sixty-seven years in Norway and Iceland. Effective retirement ages range from the late fifties in Hungary, France, Belgium, Austria, and Luxembourg to sixty-eight years in Iceland, sixty-nine years in Japan, and seventy years or more in Korea and Mexico, whereas in most countries effective retirement occurs when individuals are in their early sixties. However, as older people's activity rates become increasingly diverse, these averages become less helpful as indicators.

Table 1 Effective age of retirement and official age of retirement, 1999–2004

Country	Effective age ^a (years)	Official age ^b (years)	Difference: Effective age minus official age (number of years)
Belgium	59	65	–6
Austria	59	65	–6
Luxembourg	59	65	–6
Italy	60	65	–5
Finland	60	65	–5
Hungary	58	62	–4
Spain	61	65	–4
The Netherlands	61	65	–4
Germany	61	65	–4
Poland	61	65	–4
Norway	63	67	–4
Slovak Republic	60	62	–2
Canada	63	65	–2
Australia	63	65	–2
United Kingdom	63	65	–2
France	59	60	–1
Switzerland	64	65	–1
Portugal	64	65	–1
New Zealand	64	65	–1
United States	64	65	–1
Sweden	64	65	–1
Czech Republic	62	62	0
Denmark	65	65	0
Ireland	65	66	1
Iceland	68	67	1
Turkey	64	60	4
Greece	63	58	5
Japan	69	60	9
Mexico	74	65	9
Korea	70	60	10
Source: Adapted from Keese (2005, 8).
Note: Ages are presented to the nearest whole year.
^b

This cross-sectional picture disguises trends over time and differences by gender, as well as by educational level, occupation, and industry. Between 1970 and 2000 across OECD countries, the effective retirement age for men was reduced from nearly sixty-nine years to sixty-three years and expected years in retirement increased from eleven to eighteen. For women the average effective retirement age fell from sixty-seven years to less than sixty-two years over the same period and expected years in retirement rose from fourteen to more than twenty-two years (OECD 2006, 34). The increase in years in retirement for both men and women is a function of increases in longevity coupled with changes in effective retirement ages. Figure 1 shows how the number of expected years in retirement has increased between 1974 and 2004 for men (Figure 1A) and women (Figure 1B). In both graphs the countries are ranked in descending order on expected years in retirement for men. The same scale is used on the y-axis for ease of comparison. The first key feature illustrated by Figure 1 is the increase in expected years in retirement over the period in all countries shown. Second, women can expect to spend more years in retirement than men. This reflects both gender differentials in longevity and also differences in effective retirement ages. Third, it is evident that EU countries, especially France, Austria, Italy, Belgium, and Luxembourg, are characterized by among the highest numbers of expected years in retirement, at around twenty years for men and twenty-five years for women. There are marked differences between countries, which are largely a reflection of differing socio-institutional structures and labor market norms. This is illustrated by the fact that in general (but not exclusively), countries categorized by Esping-Anderson (1990, 1999; see also Kofman and Sales 1996) as socialist or universalist, where welfare provision is based on universalism and high taxes (e.g., Scandinavia) or liberal or residual welfare regimes, where welfare provision is means-tested and at modest levels (e.g., the United Kingdom, the United States, and Australia) are characterized by higher effective retirement ages and lower expected years in retirement than those characterized by conservative or social insurance welfare regimes with relatively high benefit levels and a strong family ideology (e.g., Italy, France, Austria).

Figure 1 Expected years in retirement. Note: Data relate to life expectancy at the average age of retirement. Source: Organization for Economic Co-operation and Development (OECD) estimates from OECD (2006, Figure 2.5, 33).

Given general improvements in health over time and increasing life expectancy, the information presented in Figure 1 suggests, at least at face value, that there could be considerable potential for increasing participation in the labor market among older people. Indeed, since the mid-1990s data from the European Labor Force Survey and other national labor force surveys show that workforce participation rates among older people have increased in almost all OECD countries. However, there have been contrasting patterns in labor force participation by gender over the longer term, with a substantial decline in participation rates among men between fifty and sixty-four years old across OECD countries, whereas participation rates for older women increased. In 2004 more than one in four men in this age group were not in the labor force, compared with one in six in 1970. Over the same period, participation rates for women between fifty and sixty-four years old increased in almost all OECD countries, although there are substantial differences among countries, with 35 percent or fewer women of this age in the formal labor force in countries such as Belgium, Greece, Italy, Luxembourg, Spain, and Turkey, compared with at least 65 percent in the Nordic countries and New Zealand (OECD 2006, 28). In part, these differences reflect national variations in gender inequalities and relations (Perrons 1995; Duncan 1996), but they also highlight some of the difficulties of comparative cross-national (and cross-regional) analyses based on secondary data sources when there are important differences in the balance between formal and informal economic activities (Monk and Garcia-Ramon 1996).

On the basis of the information presented so far, it is evident that the same broad demographic and associated labor force challenges are evident across countries, although the scale and pace of trends means that the size of these challenges varies across countries. This intercountry variation is summarized in indicative fashion in Table 2, where countries are categorized into three categories (high, average, or low) on the basis of participation rates of those aged between fifty and sixty-four years and three categories (moderate, large, or very large) according to projected change in the aged dependency ratio over the period from 2000 to 2050 (OECD 2006). Those countries in the bottom right cell (including Greece, Italy, Poland, and Spain) are identified as facing very large projected changes in aging (as captured by the old-age dependency ratio), but low participation rates at the current time mean that there is a substantial opportunity to increase participation. Countries such as Japan (top right cell) also face very large challenges because of population aging, but here participation rates among those between fifty and sixty-four years old are already high, suggesting that there is less scope to increase participation rates than in the former category. Conversely, countries such as Belgium and Turkey (bottom left cell) share with Greece, Italy, Poland, and Spain low participation rates in this age group, but the projected challenge they face in terms of population aging is less pronounced. The countries in the top left cell (including Denmark, Norway, Sweden, and the United States) face the smallest challenges in relative terms, but even here large adjustments are likely to be necessary in the face of demographic and labor market change (see Wan et al. 2005). Denmark and Sweden, both of which are characterized by a universal model of welfare provision and an emphasis on individual autonomy that serves to promote high rates of labor market participation in the older age groups, are the only EU countries in the top row of Table 2, so it is clear that in most EU countries there is substantial potential to increase participation rates among older people in the face of prevailing demographic trends. The next section focuses on factors underlying participation among older workers, associated transitions from employment to nonemployment, and barriers to increasing participation rates among older people, drawing on material from the OECD (2006) synthesis report and broader research.

Table 2 Size of labor supply challenge

	Projected change in the old-age dependency ratio, 2000–2050
Participation rate of 50–64-year-olds, 2004	Moderate	Large	Very large
High	Denmark	Canada	Japan
	Iceland	New Zealand
	Norway
	Sweden
	Switzerland
	United States
Average	The Netherlands	Australia	Czech Republic
	United Kingdom	Finland	Korea
		France	Portugal
		Germany
		Mexico
		Ireland
Low	Belgium	Austria	Greece
	Luxembourg	Hungary	Italy
	Turkey		Poland
			Slovak Republic
			Spain
Source: Organization for Economic Co-operation and Development (OECD) estimates based on data on participation rates from OECD Labor Force Statistics and old-age dependency ratios from the OECD Demographic and Labor Force Projections Database; Table 1.1, 20; OECD (2006).
Notes: Countries have been ranked into the three categories for each variable according to whether they are more than one-half of a standard deviation away from the mean for all countries. Old-age dependency ratio is the ratio of the population aged sixty-five and over to the population aged between twenty and sixty-four years.

Understanding Labor Market Transitions

Traditionally, retirement has been conceptualized as an event, marking the point at which an individual leaves the labor market at the end of his or her working life, with no possibility of return at some later stage. However, research from the United Kingdom contends that it might be more accurate, and more useful in terms of gleaning greater understanding regarding labor force participation decisions, to conceptualize retirement as a gradual trend during which people “pass through a retirement zone” rather than “jump off a cliff” (Centre for Research into the Older Workforce [CROW] 2004; Vickerstaff 2004).

In some countries, such as the United Kingdom and Sweden, relatively large numbers of older men (and women) exit the labor force via disability and other welfare benefits, and in other countries (such as Italy) formal early retirement schemes play an important role in exit from the labor market prior to official retirement age. Hence, as outlined earlier, the socio-institutional context of retirement is important in understanding differences in labor force participation and retirement between countries.

Furthermore, the transition from employment to nonemployment is one of a series of transitions most people experience in their lives from their fifties onward. For instance, changing family circumstances often particularly affect women in their fifties who have been identified as a “pivot” generation juggling care (for parents and grandchildren) as well as work roles (Phillipson and Smith 2005). This underlines the importance of viewing decisions regarding participation and nonparticipation in the labor force in a broader sociodemographic context (Cappellari, Dorsett, and Haile 2005), rather than exclusively in labor market terms.

In this broader context a multiplicity of factors have a role in shaping decision making. Often these factors are conceptualized in relatively simple terms as pull and push factors. Pull factors are those that draw older workers into retirement. Financial incentives play a prominent role here, especially private pension arrangements. More generally, the age at which pensions can be accessed and the value of pensions vis-à-vis the value of earnings from work influence decision making. In general, pull factors are associated with choice on the part of individuals about the timing of transitions. Nonfinancial considerations may also play a role. For instance, it is clear that retirement of one partner in a couple often prompts the retirement of the other (Blau and Riphahn 1999; O'Rand and Farkas 2002). This highlights the importance of the household context in joint (or at least coordinated) decision making and in understanding labor market transitions.

By contrast, push factors are those that make continued employment nonviable or unattractive, such that people are propelled toward retirement. Push factors are far more likely to be associated with involuntary transitions than pull factors. Traditionally, older workers have been conceptualized as fulfilling a buffer role in the labor market, such that they are foremost candidates for labor shedding in a slack labor market. Hence, from a geographical perspective, in slack local labor markets, perhaps characterized by mass redundancies from traditional industries, involuntary nonparticipation in the labor market among older people is likely to be higher than in those areas where local labor markets are more buoyant. Push factors may take the form of negative perceptions of older workers among employers, perhaps relating to concerns about productivity and wage costs, and a preference for younger workers (where available). At the individual level, push factors may include deteriorating health or skills obsolescence, perhaps coupled with a lack of opportunity or unwillingness to invest in retraining to acquire new skills.

The OECD (2006) synthesis report identifies three major barriers to employment among older people: financial (dis)incentives, employment barriers (i.e., labor demand barriers), and weak employability (defined as labor supply barriers, although most definitions would incorporate both demand and supply elements; McQuaid and Lindsay 2005). Each of these is considered in more detail later (with particular emphasis on the second and third factors). However, given the complexity of decisionmaking regarding retirement and of labor market transitions, it is unlikely from a policy perspective that much success will be gained in addressing one type of barrier in isolation from the others.

Financial factors influencing employment and retirement decisions include pension rules, tax allowances, early retirement schemes and options for phased retirement, and out-of-work and in-work benefits. In most countries legislation and practice regarding these issues is determined at the national level, although rules regarding early retirement schemes might vary from one company to another. Pension reform is high on the public policy agenda in many countries and changes to raise SPAs and to outlaw mandatory retirement ages are underway (OECD 2007).

On the demand side of the labor market, employment barriers can be categorized into two groups: employer barriers—including negative attitudes toward older workers (especially in terms of their adaptability and productivity) and higher wage costs (where wages are linked to age; Taylor and Walker 1994)—and the quality of employment. Employer attitudes toward older workers reflect a range of objective and subjective factors (Metcalf and Meadows 2006). A recent study from Scotland (Loretto and White 2006) indicated that employers display somewhat ambivalent attitudes toward older workers: On the one hand, they provided a qualified endorsement of the contribution of older workers to their organizations but on the other showing no marked enthusiasm to wholeheartedly embrace older workers or to formally consider them as part of resource planning into the future. This is indicative of an enactment gap between formal policy and practices accepted in reality. In virtually all countries covered by the OECD (2006) study, some evidence emerged that employers held stereotypical views about the strengths and weaknesses of older workers, typically regarding them to be less flexible and less willing to learn new skills, albeit that older workers were considered more loyal and committed than younger workers. Negative perceptions about the abilities and productivity of older workers tend to be translated into lower hiring rates and lower retention rates of older workers than younger workers. In general, evidence on aging and productivity suggests that although individuals' productivity does decline with age in terms of manual dexterity and speed of processing material, any such decline tends to be gradual and can be compensated for by a greater range of work and life experience on which to draw (for a review of the literature see Skirbekk 2003). Moreover, changes in the structure of employment, notably the shift from manual to nonmanual employment and increased emphasis on communication and customer care skills, are likely to favor employment of older workers, as are trends toward more flexible working—as an opportunity to work reduced hours might be especially attractive to older people as they make a gradual transition out of employment to nonemployment. However, research in the United States suggests that there is a lower incidence of flexible work options for older workers than might be expected given the level of demand for them (Rix 2004).

On the supply side of the labor market, possible aspects of weak employability among older workers include actual or perceived obsolescence of skills, inadequate help for older job-seekers in finding or retaining employment, and poor health (Bound et al. 1998). Irrespective of individual productivity, labor market restructuring might lead to a devaluation of the skills of older workers because the task that they are trained to undertake might no longer be important or necessary. Technological and organizational change means that skills requirements are changing. Across all countries covered in the OECD (2006) review, improving training opportunities and use of those opportunities by older people was identified as one key means for strengthening the employability of older workers. The percentage of workers receiving work-related training declines with age and is lowest among those over fifty (Taylor and Unwin 2001). In part, this might reflect fewer expected years of employment in which employers and workers expect to recoup the investment in training, but from an employer perspective the fact that older workers are less likely than younger workers to move to another job could be an offsetting factor. Indeed, there is evidence that the relative incidence of training for older workers tends to be higher in those countries where workers withdraw from the labor market at an older age (OECD 2006).

Diversity of Experience

The emphasis so far has been on similarities and differences among countries in the size, pace, and scale of demographic aging and the labor market position of older people. However, this national-level focus disguises important variations in experience at the subnational level and between population subgroups. This section summarizes why it is necessary to focus on the diversity of older people, drawing on evidence from the OECD (2006) synthesis and from the broader literature.

For geographers, subnational variations in experience are of particular interest. Within all countries there are marked subnational variations in age profiles. Often interregional variations in demographic structure are less pronounced than intraregional variations between urban and rural areas. The challenges of aging tend to be starkest in rural areas, which are typically characterized by out-migration of younger people and in-migration of older age groups (Champion 1992). Moreover, traditionally international migrants have tended to be attracted to larger cities, further exacerbating intraregional age structure differentials. In general, it is in the most peripheral rural areas that the highest median ages are to be found and where projected increases in old-age dependency ratios are most marked (Lowe and Speakman 2006).

Likewise, within countries there might be marked differences in labor market conditions, with some regions and local areas characterized by higher employment rates than others. There are also variations in industrial and occupational structure, such that the nature of employment opportunities varies to some extent between local areas. Past experiences of economic restructuring help shape local labor market norms and recruitment practices (Hanson and Pratt 1992; Peck 1996), in turn shaping attitudes relating to the transition from employment to nonemployment among older people. Geographical differences in participation rates among older workers are particularly pronounced in countries such as Australia, Canada, Finland, Germany, and Italy.

Within countries, regions, and local areas there are also marked differences in labor market experience between population subgroups. So far, mention has been made only of contrasts by gender. In all countries covered by the OECD (2006) review there are important variations in participation rates among older people (and younger ones) by level of education. Generally, participation rates for people between fifty and sixty-four years old are lower for those with less than upper secondary-level education than for those with education at upper secondary and tertiary levels (OECD 2006, 45). In much of central and eastern Europe, along with Italy, Belgium, and Luxembourg, fewer than 40 percent of those between fifty and sixty-four years old without education at upper secondary level are in the labor force. In countries with relatively high aggregate participation rates in older age groups, the labor market disadvantage of those with lower educational levels tends to be particularly pronounced. In Denmark just over 50 percent of such individuals are in the labor force, compared with well over 70 percent of those with upper secondary-level education and around 80 percent of those with tertiary education. Likewise, in the United Kingdom, less than 60 percent of older people with few qualifications are in employment, compared with nearly 80 percent of those with education to upper secondary and tertiary level. In the United States also, there are clear differentials in participation rates by educational level, ranging from around 50 percent for those with less than secondary-level education to nearly 80 percent for those educated to tertiary level. Even when focusing on differences in experience by population subgroups it is important that the significance of a geographical perspective is not overlooked: Geography matters most for the least skilled subgroups because typically they are more dependent on local employment opportunities than those with higher skill levels (Green and Owen 2006).

In considering variations in experience by population subgroup, as well as looking at attributes such as skill level, it is possible to categorize older people according to the amount of choice that they are able to exercise over decisions about how and when to move from employment to nonemployment. In research in the United Kingdom, which is likely to have broader resonance elsewhere, McNair et al. (2004) made a distinction between “choosers,” “survivors,” and “jugglers.” The choosers are those who are able to exercise the greatest control over the timing and nature of transitions, perhaps opting to combine the benefits of work with the flexibility of retirement in a way that suits their individual and household circumstances. Choosers are likely to be more highly qualified than average and may be more numerous in tight than in slack local labor markets. Conversely, the survivors might find that the timing and nature of transitions from employment to nonemployment are more likely to be dictated by redundancy or ill health than by choice. They are likely to be found disproportionately among those with poor skills and in weak labor markets. In between are the jugglers, those who find that their choices are constrained by labor market opportunities and household responsibilities. Women and residents of rural areas are likely to be overrepresented among this group.

Dimensions of variation other than gender, skill level, and degree of choice exercised over decision-making are also likely to be of importance. For instance, variations by ethnicity are not covered in the OECD (2006) synthesis report but are likely to be significant in many OECD countries given the importance of international migration involving diverse ethnic groups. Such diversity of experience between areas and population subgroups outlined earlier is indicative of the need for policy to be sensitive to different local areas and contexts if challenges of demographic aging and employment are to be addressed successfully.

Concluding Comments

Demographic aging and labor force participation are key challenges facing all Western economies, albeit to varying extents in different countries. The OECD (2006) synthesis report suggests that, to date, Europe and Japan tend to be most advanced in terms of the prominence of aging on the public policy agenda (in Europe, EU policy directives have played an important role in shaping the debate). The absence of action could pose significant economic and social problems. The message of Live Longer, Work Longer is to do what the title suggests: to match longer, and generally healthier, lives with longer working lives. To do this, it will be necessary to adopt a comprehensive approach, taking action on three fronts. The first is to reform the balance of financial incentives to make working longer a more attractive proposition. The second is to change employer attitudes and employment practices to facilitate recruitment and retention of older workers. The third is to enhance the employability of older workers through improvements in labor supply.

Many countries are taking action (as highlighted in the OECD country reports on which the synthesis report is based), but here the United Kingdom is used as an example to illustrate the range of interventions in use. In the United Kingdom, using the strategy “Opportunity Age: Meeting the Challenges of Ageing,” the government, in partnership with employers, trade unions, and civil society, seeks to create an environment to achieve higher employment rates and greater flexibility for those over fifty in continuing careers, managing any health conditions, and combining work with family (and other) commitments; enable older people to play a full and active role in society, with an adequate income and decent housing; and allow individuals to keep independence and control over their lives as they grow older (Hirsch 2003; Department for Work and Pensions 2005). Clearly, the strategy is not about the relationship between older people and the labor market alone, but increasing employment rates for older people is an important element. Within the context of an overall employment target of 80 percent, there is a target to reduce the gap in employment rates between those over fifty and the overall employment rate. Likewise, in recognition of the need to combat geographical differences in experience, there are further targets to reduce gaps in employment rates between the areas experiencing the lowest employment rate and the national average. Older people are the target of both a specific active labor market program (New Deal 50 plus) providing a combination of advice, help with a job search, and financial support to assist nonemployed people aged fifty and over back into work and more general labor market programs. Although advice and guidance are generally effective for older people and financial incentives have facilitated entry into work (Moss and Arrowsmith 2003), it is clear that the impact of supply-side measures may be limited in the absence of general measures to improve the quality of work for older people and open up possibilities for advancement in work. With the introduction of anti-age-discrimination legislation in 2006 greater emphasis has been placed on promoting business benefits of age diversity to employers.

In many countries the national state sets the legislative and socio-institutional context for fiscal and other policies influencing economic growth, but a geographical perspective is important, not only because of differences in the size of the demographic and associated labor market challenges of aging, but because it is at the local level that labor market norms operate and where people (especially older people) formulate their decisions and live their lives (Phillipson and Smith 2005). For geographers, the OECD (2006) synthesis report and individual country reports provide a useful context for more detailed analyses. Key issues and questions for future research stemming from the report include more detailed exploratory analyses of regional and local differences in the scale and size of demographic and labor markets, trends in employment restructuring, and changes in labor force participation (as in the case of Scotland in a broader UK context [see Brown and Danson 2003] and at the subnational level elsewhere, in countries such as Australia, Canada, and Finland [see OECD 2006, 44]). At a more microlevel there is a need for local studies to understand decision-making at individual and household levels and by employers, in different local contexts. In particular, there is a need for more research on how the life course varies for different subgroups in different regions and local areas, and how the employment to nonemployment transition of older people fits alongside other life course transitions. From a policy perspective, there is scope for evaluation studies of what policy initiatives to promote higher employment rates among older people work (or do not work) for whom, where, how, and why.

Despite differences among OECD countries, regions, and local areas in the scale of the challenge and its public policy profile, it is a challenge of universal significance; as noted in the OECD (2006, 14) synthesis report: “the bottom line is that population aging is both a challenge and an opportunity.” However, one key issue that the OECD report fails to consider is aging in a global context. This is an important omission, given that aging has a dynamic relationship with globalization and is “truly a global concern” (Harper 2006, 219). To date, the main focus of debate on demographic aging has been on Europe (European Commission, 2007). The OECD (2006) report extends beyond Europe to some OECD countries, but as Harper (2006) notes, the Asia-Pacific region (which is largely absent from the OECD review) is the most rapidly aging world region. The challenges faced in those parts of the world where the speed of aging is greatest will in turn have implications for the OECD countries that have been the focus of attention in this article.

Acknowledgments

ANNE E. GREEN is a Principal Research Fellow at the Institute for Employment Research, University of Warwick, Gibbet Hill Road, Coventry, CV4 7AL, UK. E-mail: [email protected]. Her research interests include spatial dimensions of economic, social, and demographic change; regional and local labor markets; and issues concerned with migration and commuting.

Notes

^aThe effective age of retirement is the average age at which workers over the age of 40 withdrew from the labor force over the period from 1994 to 2004.

The official retirement age is the relevant state pension age.

* I wish to thank anonymous referees and the editorial team for their useful suggestions for improving this article. Special thanks are also due to Irene Hardill for convening the conference session where this article was first presented and for steering the focus section