Over past decades, much of the policy focus in OECD countries to deliver longer working lives has been on increasing incentives to continue working at an older age. Countries have been reforming old-age pensions and restricting the use of early retirement schemes and other passive benefits. Some reforms have also sought to make it easier to combine income from work and the pension as well as reduce financial penalties to working longer. In many countries, the statutory age of retirement has been raised to improve the financial sustainability of the pension system. However, higher retirement ages must be accompanied by greater employment opportunities at an older age to avoid hardship for older workers. Good‑practice measures towards a more inclusive labour market for older workers should facilitate phased retirement, with better possibilities for combining work and retirement while preventing welfare benefits being used as alternative pathways to early retirement.
Working Better with Age
3. Strengthening incentives to build up longer careers
Abstract
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.
Infographic 3.1. Stronger incentives needed to build up longer careers
Key policy recommendations
Enhance incentives to continue working at an older age by:
Ensuring that the old-age pension system encourages and rewards later retirement in line with increased life expectancy while ensuring adequacy and sustainability of pension payments.
Encouraging longer and more satisfying careers through more flexibility in work retirement transitions, including by promoting phased retirement, better balancing work and care and permitting a combination of pensions with work income.
Restrict the use of publicly funded early-retirement schemes which encourage workers to leave employment while they are still in good health and able to work.
Ensure access to social benefits, such as unemployment and disability benefits, for all workers, irrespective of their age and status, and monitor that these benefits are used for their original purpose and not to incentivise early retirement for those still able to work.
Individual decisions to leave, remain or return to work at an older age are strongly affected by old-age pension schemes and other social-benefit provisions for older workers, such as unemployment or disability benefits. Not surprisingly, reforming old‑age pensions and restricting access to formal and de‑facto early retirement schemes have been at the forefront of policy initiatives to strengthen financial incentives to work longer and promote supportive social norms about economic activity at an older age.
Pension reforms have been a key policy lever for promoting longer working lives
Over the past decades, the great majority of OECD countries has been reforming the pension system to encourage longer working lives and increase the financial sustainability of the system. These reforms have essentially included closing early retirement pathways, raising pension ages and making a stronger link between earnings and pensions; either by transforming traditional defined benefit schemes into notional defined contribution (Italy, Latvia, Norway, Poland and Sweden) or point (Estonia, Lithuania and the Slovak Republic) schemes, or, like in Austria, France, Portugal and Spain, by calculating pension benefits based on a larger part of a worker’s career (OECD, 2019[1]).
Raising pension ages has a pronounced impact on participation. Recent analysis suggests that participation rates drop by about 5 percentage points at the minimum retirement age - when pensions from mandatory schemes become first available - and another 5 percentage points at the normal retirement age. A one-year increase in both minimum and normal retirement ages is associated with a gain in the overall participation rate of the age group 55-74 by about 0.8 percentage points (Geppert et al., 2019[2])
Over the past two decades, statutory retirement ages have increased in many OECD countries for both men and women (Figure 3.4). Recent reforms have made “age 67” the new “65” and several OECD countries are going even further. In 2019, age cut-offs ranged from 58 (for women in Turkey) and 59.3 (in Slovenia) to 67 in Iceland, Israel and Norway.
Based on legislated measures, for those entering the labour market now and with a full career from age 20, the normal retirement age will increase by 1½ years for men and 2 years for women on average across the OECD. The future normal retirement age varies enormously from 59 years in Turkey (women only) and 60 years in Luxembourg and Slovenia to an estimated 74 years in Denmark. Only France, Luxembourg, Poland, Greece and Slovenia will have a normal retirement age below 65 years. In Denmark, Italy and the Netherlands, the future normal retirement age will be higher than 70 years.
Figure 3.1. Retirement ages have been rising and will increase further in nearly half of OECD countries
Note: Future retirement age according to latest legislation for those entering the labour market in 2016.
Source: OECD Dataset: Pensions at a Glance, https://stats-2.oecd.org/Index.aspx?DataSetCode=PAG.
Among the non-OECD G20 countries, the retirement age will gradually rise in Indonesia from 56 to 65, bringing it in line with OECD countries. The retirement will also rise in the Russian Federation over a relatively short time frame from 55 to 60 for women (55 with 37 years of contributions) and from 60 to 65 for men (60 with 42 years of contributions). Elsewhere, statutory retirement ages have either remained frozen at 65 for men (e.g. Argentina) or remain comparatively low, partly reflecting lower life expectancy even though longevity has been increasing (e.g. 60 in China for men, 58 in India and 55 in Brazil for men). In these countries, population ageing will also accelerate and so raising statutory retirement ages in line with increasing longevity will need to be considered.
As pension reforms to deal with population ageing are often unpopular, a key challenge for many countries is to withstand pressures to backtrack on reforms and to implement those that are already legislated. Indeed, Canada chose not to implement a planned increase in the retirement age to 67 for the basic and means-tested pensions; the Czech Republic will no longer increase the pension age beyond age 65; and Poland reversed the planned increase to 67, with retirement ages now set to drop back to 65 for men and 60 for women (OECD, 2017[3]). Several other OECD countries including Slovak Republic, Netherlands, Italy and Spain have also backtracked on earlier reforms on extending retirement ages (OECD, forthcoming[4]).
Will higher pension ages reduce time spent in retirement?
In response to pension reforms and the closing of early retirement pathways, the average effective age at which older workers exit the labour force has been increasing over the past 10 to 15 years in nearly all OECD countries.1 This has partly reversed a steep decline that occurred in all countries since the early 1970s. Across the OECD area, the average effective age of retirement for men has risen by 2.3 years between 2000 and 2018 after falling by 5.7 years between 1970 and 2000. For women, it rose by 2.8 years after a decline of 5.6 years. Does this mean that today’s retirees will spend less time in retirement than those of yesterday? In fact, because life expectancy has continued to rise most countries, the expected time spent in retirement (i.e. the life expectancy of persons at the average effective age of retirement) has either risen by a small amount over the past 10 years or remained stable for men in most countries. For women, there have been some small declines in a number of countries. However, because of longer life expectancy and a slightly lower average effective retirement age, women can still expect to spend considerably more years in retirement than men – 22 years on average across the OECD versus 17 years for men. There are also considerable differences across OECD countries, ranging from 16 to almost 27 years for women and from 13 to almost 23 years for men. In general, despite later retirement, both men and women can still expect to spend close to a quarter of their lives or more in retirement.
Figure 3.2. Later retirement does not mean fewer years in retirement
Source: OECD Database on Average Effective Retirement Age, https://www.oecd.org/els/emp/average-effective-age-of-retirement.htm.
Pension reforms should avoid hardship for economically disadvantaged groups
While later retirement on average has not translated one-for-one into a lower expected number of years in retirement, there are differences between groups of workers and across countries. High-income earners tend to live longer than low-income earners and generally receive pensions for longer periods. It has been estimated that lower life-expectancy for low earners of three years at age 65 relative to high earners reduces their pension wealth by about 13% (OECD, 2017[5]). Increasing retirement ages can therefore potentially exacerbate socio‑economic inequalities. However, if inequality in life expectancy is stable, increasing retirement ages given longer lives has a quantitatively small impact on this pension-wealth inequality (OECD, 2017[3]).
There is no easy way to account for inequality in life expectancy within pension systems. The first best solution is to focus policy action on actively reducing socio-economic differences in health. Shorter lives for retirees from disadvantaged socio-economic backgrounds imply that a pension system that seems to be actuarially neutral is in fact not neutral: it is regressive. This strengthens the weight that should be placed on redistributive components of pension systems.
Pension policy measures to take account of socio-economic differences in life expectancy could target the benefit formula (granting higher accrual rates for low earnings, as applied in Portugal) or the contribution‑rate pattern, e.g. increasing with income (such as in Brazil) or including a higher wage ceiling for contributions than for pension entitlements (as in the notional defined contribution scheme in Norway). In funded defined contribution schemes, the annuity factors for conversion of assets into a pension benefit could be set in ways that increase pensions for people with low pensionable income (who die earlier on average) while people with high pensionable income (who die later on average) would receive lower benefits.
Portugal, in particular, provides a good example of how governments can strike a balance between equity objectives and the need to ensure the financial sustainability of public pensions (Box 3.1). If all OECD countries adopted a similar approach by 2020 (except for countries where the currently legislated increases of the retirement age are even larger), employment rates of older people and long-term GDP per capita would increase, alleviating the stress on public finances (Geppert et al., 2019[2]).
Box 3.1. Improving financial sustainability whilst offsetting inequalities
In Portugal, the statutory retirement age rises by two thirds of life expectancy gains, which aims to improve the financial sustainability of the pension system while keeping the ratio of adult working life to time spent in retirement roughly constant. In addition, people with very long careers can retire slightly earlier. A worker with a 43-year career can retire 1 year before the standard statutory retirement age. For someone who started working at age 20, this means that the retirement age in effect rises by half of life expectancy gains rather than two thirds.
At the same time, higher accrual rates are granted for lower earnings, offsetting the pension disadvantage for low-income earners from lower life expectancy.
There are several advantages to this approach:
Linking the retirement age to life expectancy removes the need for recurring political wrangling that accompanies ad hoc initiatives that seek to keep pension parameters in line with demographic realities;
The two-thirds increase maintains a broadly constant ratio of working life to life‑in‑retirement. This may be politically more acceptable than a one-to-one link of retirement ages to life expectancy;
The provision included for long careers represents a progressive element that is beneficial for economically disadvantaged groups but which nevertheless maintains a link to life expectancy for everyone.
Source: OECD (2019[6]), OECD Reviews of Pension Systems: Portugal, https://doi.org/10.1787/9789264313736-en.
Increase in effective retirement ages will be limited if incentives to retire early are not tackled
Access to early retirement in the main old-age pension or in special early retirement schemes has become more restrictive in almost all OECD countries. These early-retirement pathways were initially opened up to free up jobs for the young and reduce youth unemployment but proved to be costly failures as young people cannot be simply substituted for older workers and the increase in taxation to pay for these schemes can harm labour demand. In some countries, the possibility for early retirement has been restricted by an increase in the minimum years of contributions that are required to receive a full pension (e.g. Austria, Italy and Spain) or through large penalties (e.g. Portugal through the sustainability factor). Some countries have abolished early retirement options for the private sector entirely (e.g. Netherlands and Poland, including for public-sector workers) while others have sought to make early retirement more costly for employers (e.g. Belgium).
While these measures go in the right direction, early retirement options continue to exist in both private and public pension systems in many OECD countries even after recent reforms (Annex Table 3.A.1, Column A). For example, in the United States it is possible to claim Social Security benefits at age 62 subject to an actuarial reduction. Many older US workers, especially low-educated males, still opt to take early retirement even though it results in a lower pension benefit whereas working beyond the normal retirement age attracts a pension bonus (OECD, 2018[7]).
Among countries where early retirement is possible, the earliest age of retirement increased in only seven between 2002 and 2016, while the early retirement age has been kept unchanged in the majority. As a result, the average gap between the normal and early retirement age has remained constant in half of the OECD countries (Figure 3.3). Even though early retirement ages are expected to rise in some countries, thanks to the link between the retirement age and life expectancy, pathways into early retirement should be restricted to those older workers facing hardship in continuing to work.
Figure 3.3. The gap between early and normal pension ages remains large in some countries
Note: Ages refer to labour market entry at age 20, with the normal retirement age being the earliest point to be eligible for all pension components without deduction.
Source: OECD (2017[3]), Pensions at a Glance 2017: OECD and G20 Indicators, Table 2.1, https://doi.org/10.1787/pension_glance-2017-en.
Currently, many OECD countries have special retirement rules for arduous and hazardous jobs. A key component of special retirement rules for workers in these jobs is the possibility of earlier retirement than other workers and without penalties. As people in manual jobs not only have less expected years in retirement but also may not be physically able to work longer, some countries have explored compensating undesirable inequality outcomes in retirement by introducing more granularity in retirement eligibility based on people’s skills and the physical requirements of their jobs. However, these schemes also tend to introduce administrative complexity for firms and workers and may create distorted incentives. For instance, it is difficult to account for mixed careers with periods of arduous and non-arduous work and it is not easy to define arduous work unambiguously. Instead, more effort is needed to prevent negative effects arising from hazardous or arduous working conditions in the first place while enhancing job mobility via upskilling and reskilling of worker and facilitating greater choice in working-time arrangements (including through greater flexibility in combining income from work and pensions). A well-designed disability benefit system can also take better account of individual differences amongst workers in their capacity to continue working rather than a rule-based pension scheme for arduous work.
However, welfare programmes such as disability and unemployment schemes have not always been used for their original purposes and have served as de facto pathways into early retirement. Overall, official and de facto early retirement pathways have contributed to the situation where at least a quarter of older workers withdraw from the labour force at 60 years of age or below in the majority of countries even where the official age of retirement is 65 or higher (Figure 3.4).
Early retirement remains widespread despite efforts in many countries to tighten access to early retirement schemes. In 26 OECD countries, a quarter of men retired below the age of 60 in 2018, dropping to as low as 53 or 54 in Turkey and Luxembourg. The corresponding age is often substantially lower for women.
Figure 3.4. In most countries, large shares of workers retiring do so well before the age of 65
Note: The effective retirement age refers to the weighted average age of labour market exit for persons who were in the labour force in 2013 and aged 40 and over but no longer in the labour force in 2018. Data refer to the period 2005-2010 for China and 2012-2017 for Saudi Arabia.
Source: OECD calculations based on the Average effective age of retirement dataset (unpublished).
In several countries, the older unemployed can receive unemployment benefits for a longer period than younger ones and sometimes until the retirement age. Unemployment Insurance (UI) rules that link benefit duration to age remain particularly favourable for older workers in Chile, Japan, Korea and many EU countries (Figure 3.5). While for some older workers, longer UI durations might be justified by their greater difficulties in finding a new job, the same case could be made for other groups of jobseekers as well. In general, countries should move away from entitlement criteria that are explicitly based on age, as these rules can reinforce damaging stereotypes and social norms about older workers and do not reflect the diversity of labour market situations facing them. Instead, activation and employment-support policies should seek to address the specific employment barriers of all jobseekers, regardless of age, by actively tailoring job-search obligations and labour-market re-integration approaches more closely to their individual needs and employment prospects.
Finally, while early retirement pathways via disability benefits have been tightened, disability benefit programmes remain relatively generous and easy to access across many OECD countries (OECD, 2010[8]). It is difficult to strike the right balance between ensuring adequate income support for people with disability and encouraging a quick full or partial return to work for those with remaining working capacities. But the evidence suggests that, in general, restricting access to disability benefits to those unable to work will be more effective if combined with a strong focus on preventive and active measures (OECD, 2015[9]). Research has demonstrated repeatedly that keeping people with reduced work capacity in employment is much easier and more promising than trying to help them back into employment after years of health‑related inactivity. Good practice in this field includes: i) moving from medical to work capacity to asses eligibility; (e.g. Austria); ii) fostering a more co-ordinated approach for rehabilitation and early intervention (e.g. Portugal); iii) boosting the reintegration of those with capacity to work through active measures (e.g. the United Kingdom); and iv) increasing the incentives for employers to adapt work organisation and workplaces in order to prevent long-term sickness and disability (e.g. the Netherlands).
Figure 3.5. Older unemployed receive longer unemployment benefits in some countries
1. The maximum duration is capped at 60 months for Belgium and Finland. The actual maximum duration might be longer or even unlimited (e.g. in Belgium). In the Netherlands, persons over 41 are capped at 28 with the maximum duration gradually reducing by 1 month per quarter to 24 months in April 2019.
2. The chart only shows insurance benefits. Assistance benefits can be available as follow-up support in some countries. Australia and New Zealand operate only unemployment assistance benefits which are means-tested but not limited in duration. In some cases, particular rules exist for older unemployed (65 or under) that are not fully taken into account in the database. This is the case for Finland, France, Lithuania, Luxembourg, Portugal and Spain where it is possible under some conditions to extend benefits to the full-rate pension age. For country details, see Latest Policy descriptions at www.oecd.org/social/benefits-and-wages/.
Source: OECD Tax and Benefits Policy Database: OECD Indicators, www.oecd.org/social/benefits-and-wages/.
Flexible combinations of part-time work and pension receipt can promote longer working lives
In recognition of both financing challenges and diverse working preferences among workers, more flexible forms of retirement have gained growing policy interest in many OECD countries. Older people often report that they would continue to work if they had the possibility to reduce working hours. According to (Eurobarometer, 2012[10]), for two-thirds of EU adults combining a part-time job and partial pension is more appealing than retiring fully.
Flexible retirement provides older workers with the option to continue working, often part-time, while drawing at least a partial pension, or to choose when to retire with pension rights adjusted in an actuarially neutral way. Combining work and pensions is possible in most OECD countries but the conditions for doing so vary. All countries allow pensioners who have fully retired to engage in paid work but earnings from this employment can affect pension payments in different ways. These depend on the design of a pension system and its individual components, as well as tax rules and rules governing possible withdrawal of pensions once earnings from work reach a certain level (OECD, 2017[3]).
Conversely, claiming pension entitlements early is frequently only possible for some parts of the pension package, typically the (defined contribution) earnings-related portion e.g. United Kingdom and Ireland; flat‑rate or safety-net benefits are only available at the normal retirement age, almost by definition. Australia provides so-called Transition-To-Retirement Pensions (TRIPs) that let workers move from full‑time to part‑time work and complement their earnings with pension (superannuation) entitlements. In a further eleven OECD countries, work and early pension receipt can be combined under their defined benefit or point schemes (Annex Table 3.A.1, column B).
Flexible retirement systems can facilitate the choice of phased retirement, i.e. switching from full-time to part-time work at older ages, by smoothing income from work and pensions for different choices of working time. This can be particularly beneficial for extending the working lives of older people in poor health, with care responsibilities or in arduous jobs (Eurofound, 2016[11]).
However, phased retirement schemes should be designed with care to prevent subsidisation of part-time work or even early withdrawal from work, which may reduce rather than raise the effective labour supply of older people. In Germany, for example, a subsidised phased retirement scheme (“old-age part-time”) was introduced in 1988 and, initially, it was mainly used as a “block-work model”. Workers preferred to work full-time for half of the period up until eligible for the full old-age pension and then not work at all for the second half of the period, i.e. effectively, they were able to retire earlier on a full pension. This was reformed in 2009, but the block-work model of phased retirement still exists in some companies and sectors on the basis of collective agreements (e.g. in the metal and electrical industries).
Pensions systems should be designed to reward, not penalise, longer working lives
Pension systems in many OECD countries already allow combining work and pensions after the normal retirement age. Yet, in practice, there can be significant constraints or disincentives that discourage people from staying active after retirement.
In most countries where the combination of pension receipt and work is possible before the normal retirement age, there is an earnings threshold, which if exceeded, pensions are cut (Austria, Belgium, Finland, France, Germany) and therefore limits the incentives to work.
No OECD country requires workers to stop working entirely at the normal retirement age but several of them limit how much pensioners can earn. Seven OECD countries apply limits to post‑retirement earnings, above which pension benefits are reduced (Australia, Denmark, Greece, Israel, Japan, Korea and Spain). In France, working retirees do not earn any additional pension rights even though they are liable to pay pension contributions, which then act as a pure tax on continuing to work. Removing such obstacles is important to make combining work and pensions a genuine attractive option for most, or all, older workers. More generally, in order to efficiently promote more gradual forms of retirement, conditions to withdraw partial pensions should not depend on the amount of work and labour income after the normal retirement age.
The adjustment of entitlements for either postponing retirement or retiring early should be based on actuarial neutrality to avoid these distortive effects. This amounts to a bonus estimated at about 5.5% on past entitlements on average in OECD countries for each year of deferral (OECD, 2017[3]).
Options for flexible retirement must come with renewed efforts to foster financial literacy and transparency
In spite of the wish for greater retirement flexibility, surveys confirm that combining work and pensions is still uncommon and available options are either poorly understood or unattractive for retirees. In Europe, only 10% of individuals aged 60-64 or 65-69 combine work and pensions (Figure 3.6), and take-up of partial pensions is also rare in countries where it is legally possible, e.g. in the United States and France. In addition to barriers in the pension systems themselves, labour-market practices and cultural norms regarding part-time work or working past retirement still limit individuals’ freedom to choose an employment and retirement pattern that works best for them.
Only a few employers have programmes in place to support a gradual exit from employment, even if they support flexible work arrangements in principle. Only 39% of employers offer flexible time schedules in the United States (TCRS, 2016[12]). In Europe, 78% of older workers cited a lack of gradual-retirement opportunities as an important reason to stop working altogether (Eurobarometer, 2015[13])
Figure 3.6. Very few workers combine work and pensions in the OECD
Note: A person is on retirement when he/she receives an old-age pension (statutory scheme, occupational scheme, personal scheme or unknown scheme).
Source: OECD (2018[7]), Ageing and Employment Policies: United States 2018: Working Better with Age and Fighting Unequal Ageing, Figure 1.17, https://doi.org/10.1787/9789264190115-graph17-en.
Whether people will benefit from combining work and pensions depends on their capacity to make well‑informed choices to avoid jeopardising their final retirement incomes. Financial literacy plays an important role in that respect. People need transparent and reliable information on the benefits that they can expect to receive under different scenarios concerning when and how they retire (completely or partially) in order to help them plan ahead. Above all, genuine flexibility also requires reassuring older workers that they will be able to review and adapt their choices as their employment prospects, health situation or family circumstances change before or after retirement.
References
[13] Eurobarometer (2015), “Special Eurobarometer 437 DISCRIMINATION IN THE EU IN 2015 SUMMARY”, http://dx.doi.org/10.2838/325154.
[10] Eurobarometer (2012), Special Eurobarometer 378 Active Ageing REPORT, http://ec.europa.eu/public_opinion/index_en.htm (accessed on 19 July 2019).
[11] Eurofound (2016), Extending working lives through flexible retirement schemes: Partial retirement European Foundation for the Improvement of Living and Working Conditions, http://dx.doi.org/10.2806/004233.
[2] Geppert, C. et al. (2019), “Labour supply of older people in advanced economies: the impact of changes to statutory retirement ages”, OECD Economics Department Working Papers, No. 1554, OECD Publishing, Paris, https://dx.doi.org/10.1787/b9f8d292-en.
[6] OECD (2019), OECD Reviews of Pension Systems: Portugal, OECD Reviews of Pension Systems, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264313736-en.
[1] OECD (2019), Will future pensioners work for longer and retire for less?, http://www.oecd.org/pensions/public-pensions/OECD-Policy-Brief-Future-Pensioners-2019.pdf (accessed on 19 July 2019).
[7] OECD (2018), Ageing and Employment Policies: United States 2018: Working Better with Age and Fighting Unequal Ageing, Ageing and Employment Policies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264190115-en.
[3] OECD (2017), Pensions at a Glance 2017: OECD and G20 Indicators, OECD Publishing, Paris, https://dx.doi.org/10.1787/pension_glance-2017-en.
[5] OECD (2017), Preventing Ageing Unequally, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264279087-en.
[9] OECD (2015), Fit Mind, Fit Job: From Evidence to Practice in Mental Health and Work, Mental Health and Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264228283-en.
[8] OECD (2010), Sickness, Disability and Work: Breaking the Barriers: A Synthesis of Findings across OECD Countries, https://www.oecd-ilibrary.org/docserver/9789264088856-en.pdf?expires=1563823607&id=id&accname=ocid84004878&checksum=1D64006530BDAA9A916345CE8BF99567 (accessed on 22 July 2019).
[4] OECD (forthcoming), Pensions at a Glance 2019: OECD and G20 Indicators, OECD Publishing, Paris.
[12] TCRS (2016), “All About Retirement: An Employer Survey: 17th Annual Transamerica Survey”, Transamerica Center for Retirement Studies, https://www.transamericacenter.org/docs/default-source/employer-research/tcrs2017_sr_employer_research.pdf (accessed on 9 August 2019).
Annex 3.A. Flexible retirement provisions
Annex Table 3.A.1. Summary of early retirement option and rules for combining work and pensions
Note: Ages refer to labour market entry at age 20, with the normal retirement age being the earliest point to be eligible for all pension components without deduction.
Source: OECD (2017[3]), Pensions at a Glance 2017: OECD and G20 Indicators, https://doi.org/10.1787/pension_glance-2017-en.
Note
← 1. For people aged 40 and over, the average effective age of retirement is calculated as a weighted average of the declines in the participation rate over a 5-year period for people in each 5-year age-group at the beginning of the period and those aged 5-years older at the end of the period.