1. Introduction to the Review of the Korean pension system

The OECD Reviews of Pension Systems deliver an in-depth analysis of the pension system in selected countries. They focus on the pension system’s capacity to provide adequate retirement income in a financially sustainable way. The reviews examine how demographic, social and economic developments affect pension benefits and pension spending. They cover all components of the pension system, both old-age safety nets and earnings-related schemes, public and private pensions, and special regimes for specific occupations. The analyses heavily draw on OECD flagship publications (Pensions at a Glance and Pensions Outlook) and use country-specific sources and research.

OECD Reviews of Pension Systems: Korea is the eighth in the series, after Ireland (2014), Mexico (2016), Latvia (2018), Portugal (2019), Peru (2019), the Czech Republic (2020) and Slovenia (2022). It is financed by the Korean Government and is jointly produced by the OECD Directorate for Employment, Labour and Social Affairs and the Directorate for Financial and Enterprise Affairs. This review provides policy recommendations on how to improve the Korean pension system, building on the OECD’s best practices in pension design. The review describes the Korean pension system in great detail and identifies strengths and weaknesses based on cross-country comparisons.

This introductory chapter starts by succinctly discussing why now is a good time for an OECD review of the Korean pension system. It then very briefly describes the historical background of the Korean pension system before concentrating on Korea’s broader economic, labour market and demographic backgrounds.

The second chapter describes the main earnings-related schemes that exist in Korea. The general scheme covering private-sector workers is the National Pension Scheme (NPS). Entitlement to the NPS pension requires at least 10 years of contributions,1 and pension can be taken from the age of 62 years in 2020, increasing to 65 for those born in 1969. Early retirement is possible with a penalty, up to five years before the normal retirement age. The chapter then moves to the special regimes, namely the Government Employees Pension Scheme (GEPS), the Private School Teachers Pension Scheme (PSTPS), the Special Post Office Pension Scheme (SPOPS) and the Military Personnel Pension Scheme (MPPS).

The third chapter compares the adequacy of the Korean pension system with that in other OECD countries. This comparison includes safety-net benefits – the non-contributory basic pension. The financial sustainability of the NPS is also analysed taking into account funding projections of the National Pension Fund based on various national sources. The chapter then moves to analysing the financial position of the special schemes for government workers, military personnel and private school teachers. The chapter then concludes with policy recommendations to improve public pensions.

The fourth chapter analyses Korea’s private pension schemes. Korea has a mandatory employer-funded scheme which can either be a retirement pension or severance pay scheme, and a voluntary personal pension scheme as a third tier. The chapter and evaluates these schemes with reference to other OECD countries. It explores coverage, contribution levels, the tax treatment of private pensions, trends in assets under management and investment returns, solvency and funding requirements, the pay-out phase, market structure and competition, supervision, and governance. It then provides policy recommendations to improve Korea’s funded pension schemes with reference to the OECD Core Principles of Private Pension Regulation and international best practice.

The pension coverage of private-sector workers only started in 1988 in Korea. This is very late in comparison to other OECD countries, the majority of which had pension systems in place in the first decades of the 20th century. When the pension system was introduced, the population structure was very young, but has since been ageing at an unprecedented rate.

Ageing pressure and current high levels of poverty amongst the elderly generate major challenges for the retirement system in Korea. To deal with ageing, many countries have reformed their pension system over the last decades, trying to keep up with demographic, financial market and labour market developments. Needed measures taken in response to the COVID-19 crisis may generate additional pressures on public finances over time, especially if interest rates were to increase.

In which way and how strongly have pension system reforms varied across countries? Changes in retirement ages, contribution rates and pension benefit levels have been common. Moreover, some countries decided to introduce automatic adjustment mechanisms into their pension systems, based on demographic and economic developments (OECD, 2021[1]). While these innovations aim at reducing political risks in particular to deal with population ageing trends, their correct design and implementation are challenging.

Korea has been particularly active in reforming its pension system since its introduction in 1988, mainly focusing on improving financial sustainability as initial pension promises could not be maintained over the long term. Among the main reforms since the 1990s (Chapter 2) were: increasing the total contribution rate from 3% to 9%; decreasing the pension promise for a 40-year career from 70% to 40% of past wages; increasing the statutory retirement age from 60 to 65 years; and, introducing a means-tested safety-net pension for the most vulnerable.

Korean pensions consist of a partially funded public defined benefit system, old-age safety nets, an occupational scheme, and voluntary private savings. The defined benefit system has two main schemes: the general social security scheme (National Pension Scheme or NPS) and the civil servants pension scheme (Government Employees Pension Scheme or GEPS). The rules for the latter are also partially applicable for the pension schemes covering the military personnel (Military Personnel Pension Scheme or MPPS), special post office workers (Special Post Office Pension Scheme or SPOPS) and private school teachers (Private School Teachers Pension Scheme or PSTPS). The safety net consists of a means-tested non-contributory basic pension targeted at the poorest 70% of those aged 65+ supplemented by several other means-tested components covering basic livelihood and housing amongst others. Under the Korean occupational benefit system, employers have to either provide a severance pay plan or a retirement pension plan to their employees. In 2018, 27% of workplaces in Korea had retirement pension plans for their employees (Korea, 2018[2]). Individuals can also access voluntary personal pension plans to help them save for retirement.

Now is the time to take stock of where these measures have taken the whole system. There is limited short-term pressure on the finances of the NPS as the number of eligible retirees is still low. However, old-age poverty rates are very high causing social and intergenerational solidarity issues. Moreover, improvements in the long-term sustainability of the NPS are necessary as the pace of ageing will be very fast in Korea, with demographic projections pointing to a sharply decreasing size of the working-age population combined with a rapidly increasing number of retirees. This will create imbalances in pension financing and might ultimately put downward pressure on retirement income. Projections also suggest that public pensions will be transformed from partially funded to pay-as-you-go within 40 years, as the National Reserve Fund may be fully depleted. This Review analyses: whether the parameters of the public pension scheme are set in a way that makes the core of the system well equipped to face both ageing challenges and possibly deep changes in the functioning of the labour market; and, how private pension schemes can be improved to provide complementary income.

Korea did not introduce a mandatory pension system until 1960, and then only for government employees. The following section describes the historical setting of the different schemes that currently exist within Korea. Reform details and current rules and regulations are provided in Chapter 2.

The first mandatory pension scheme was introduced in 1960 and covered only government employees (GEPS). As in many countries, introducing GEPS helped attract talented people to the public service, thereby raising staff retention of civil servants.

In addition to the old-age pension benefit in the GEPS, there was also a disability benefit and lump-sum survivor benefit amongst others. Within the first few years after inception several other benefits were added – medical expense subsidy, childbirth expense subsidy, invalidity allowance, childbirth allowance and funeral service expense subsidy – to expand the coverage and scope of the schemes available to government employees. This led to a wide-ranging pension and employment package that covered all aspects of government employee’s lives, which became increasingly costly to finance. Reforms have therefore concentrated on trying to ensure the long-term financial stability of the system by increasing contributions, lowering entitlements and tightening eligibility criteria (Chapter 2).

In the beginning, the scheme included both public officials and military personnel. However in 1963, a separate pension system for the military (MPPS) was created. In 1975, a pension system for private school teachers (PSTPS) was launched, and in 1982 a pension system was launched for special post office workers (SPOPS) and as with the MPPS previously the structure of both was very similar to that of public officials. The current rules are provided in Chapter 2.

The first retirement benefit scheme – the voluntary retirement allowance – was actually introduced in 1953, but it was only voluntary for employers to contribute. In 1961, it was changed into a severance payment system and made mandatory for firms with at least 30 employees, gradually expanded until including workplaces with at least 5 employees since 1989. Severance payments are made when leaving an employer at any stage of the career and so are not necessarily used during the retirement phase.2

In 2005, the occupational pension system was introduced as an alternative to the severance payment system. Employers could choose to keep the existing severance scheme or to convert it to a tax-advantaged defined benefit or defined contribution plan with the consent of employees. The new system introduced a portable individual retirement account (IRA) for workers who change jobs.

The occupational pension system became available for all firms regardless of their size in December 2010. The IRA system was reformed further in 2011 being replaced by Individual Retirement Pension (IRP) plans. IRP plans allowed workers (full-time employees only) to voluntarily set up their own accounts. Accrued retirement benefits are mandatorily transferred to individual retirement plan accounts (where applicable) at retirement or when changing employment. In 2017, the coverage of IRP was expanded to include the self-employed, part-time workers, public servants, military personnel and private school teachers.

Although the National Pension Act was legislated in 1973 to cover private-sector workers, its enforcement was delayed due to the world economic recession triggered by the first oil shock. After the revision of the act in 1987, the National Pension Scheme (NPS) was implemented in 1988.

When establishing a pension system countries normally make a choice within a range of options, from a Beveridge approach, paying flat-rate benefits to all, to a Bismarckian scheme where benefits are directly linked to past earnings. Many countries actually follow a combination of these two models, as does the Korean NPS system in that half of the reference earnings for pension purposes is based on individual earnings with the other half based on the average earnings of all those contributing. This second component thereby effectively pays a flat-rate benefit based on each year of contribution with no direct link to individual earnings.

In addition, when the scheme is introduced, countries have to decide on how to treat those of pensionable age. The first cohorts of retirees under the new scheme have been unable to contribute during their working lives as there was no pension system in place at the time. Whilst the Korean scheme did not initially provide any compensation to those who were older than the retirement age, a means-tested non-contributory basic pension was introduced in 2007, two decades after the start of the NPS (Chapter 3). Moreover, closely related to the idea of a “gift” being given to the first cohorts of retirees upon the introduction of the pension scheme, those who initially contributed were entitled to a future pension promise that far exceeded the level obtained from compounding their contributions. While contribution rates were raised and future benefit promises were lowered, initial promises for the first generations remain a liability that is contributing to the gradual projected depletion of the National Reserve Fund.

At the beginning, the NPS covered workers in workplaces with 10 or more employees. In 1992, the compulsory coverage was expanded to firms with 5 or more employees. It was further expanded in 1995 to farmers, fishermen and the self-employed who reside in rural areas, and finally in April 1999 to those in urban areas, thereby finally covering the total labour force between the age of 18 and 59.

The NPS started with just over 4 million insured workers in 1988, but coverage expanded quickly, reaching over 16 million by the end of 1999. In principle, the NPS should now cover the entire working population that is not covered by the other mandated pension programmes (GEPS, PSTPS, SPOPS and MPPS). However, there are a large number of employees who are not making regular contributions. The main reason is because they are classified as incapable of making a contribution as either there income is below the earnings floor to contributions, around 10% of average earnings, or because of business failures leading to a temporary suspension of payment. Both of these reasons are particularly relevant among the self-employed and as a result only about half of the urban self-employed population actually contributed to the scheme immediately after they were eligible (Moon, 2001[3]). Coverage has since increased. The current rules and regulations of the system as well as coverage levels, including amongst the self-employed, are discussed in Chapter 2.

Within the next five years Korea’s population is going to start shrinking. The population is projected to peak at 51.3 million around 2024-25, and would then decline to 49.8 million in 2040, 46.8 million in 2050 and to 42.7 million by 2060 (Figure 1.1). The overall decrease is due to a sharp drop in the number of working-age adults and to the still declining number of young people which started in the late 1970s. While there were 12.3 million under-20-year-olds in Korea in 2005, their number fell to 8.9 million in 2020 – a drop of 28% – and is expected to fall further to 6.8 million in 2040 and 5.8 million in 2060 – a further drop of 35% since 2020. As for 20-64 year-olds, there were 32.1 million in 2005 increasing to 34.3 million in 2020, but their number is projected to decline to 26.6 million by 2040 and to 19.4 million by 2060 – a drop of 43% since 2020.

Conversely, the upward trend in the number of people 65 and older has started to accelerate. It stood at 8.1 million in 2020, up from 2.1 million in 1990, and is projected to reach a peak of 17.8 million in 2050 before plateauing. Hence, the combination of the sharp decline in the size of both the working-age and the young populations with the sharp increase in the old-age population triggers very fast ageing in Korea. While there were more than 11 young people under 20 for every person aged 65 or above in 1980, the 65+ will outnumber the under 20-year-olds from 2022 and there will be three people over 65 for every young person under 20 by 2060. Lower mortality rates at older ages and record-low fertility rates are the key drivers.

People who reach 65 in Korea currently have a remaining period life expectancy – i.e. based on current mortality rates – of 20.8 years, placing Korea one year above the OECD average of 19.9 years, but nearly two years behind top-ranked Japan (Figure 1.2). By 2060, it is projected that those aged 65 will have a much higher remaining life expectancy of 25.2 years in Korea against 23.9 in the OECD, with only Japan having more than 0.2 years higher at 26.3 years.

Korea currently reports the lowest fertility rates among all OECD countries, at only 1.11 children per woman of age 15 to 49, with Greece, Italy, Portugal and Spain next lowest at around 1.3, while the OECD average at 1.67 is well below what is needed to stabilise the size of the population in the long run (Figure 1.3). Across all OECD countries, only Israel, Mexico and Türkiye have fertility rates of at least 2.1. While fertility rates in Korea were still relatively high in the early 1980s, they plummeted to about 1.5 children per woman in the early 1990s and have kept shrinking further until today.

Measures of fertility can differ across different organisations. The above figures refer to the estimates from the United Nations Population Prospects data to enable a consistent international comparison from a common source. Statistics Korea data are higher, estimating current fertility at 1.24 in both 2015 and 2020. Statistics Korea forecasts that the fertility level will decrease to 1.14 in 2030 before increasing to 1.27 in 2050 and beyond against 1.09 and 1.39 based on UN projections, respectively.

Those trends lead to a rapid pace of ageing in Korea that is unprecedented in the OECD. In 1980, Korea and Colombia had the youngest population of any OECD country when considering the number of people aged 65+ compared to those aged 20-64 (Figure 1.4 and Figure 1.5). Currently, there are 23.6 people in Korea aged 65 or over for every 100 people aged 20 to 64, still maintaining Korea well below the OECD average of 30.4, within a range of 13.2 in Mexico and 52.0 in Japan. In the next few decades the pace of ageing will accelerate rapidly so that in 2060 Korea will be the oldest OECD country based on this metric at 89.7, just above Japan at 83.3 and over 30 percentage points above the OECD average of 57.8. Korea will therefore have gone from being the youngest country in the OECD to the oldest. While longer lives are undoubtedly a positive development the financial pressure on both Korea’s pension and health care systems will intensify.

Among all OECD countries, Korea is the country with the largest projected decline in the size of the working-age population (in percentage) between 2020 and 2060 (Figure 1.6). The projected 43%-drop in the number of 20-64 year-olds compares with 35% in Japan and 10% on average in the OECD. By contrast there will be an increase in 12 OECD countries, with a maximum of almost 67% in Israel with Australia and Mexico being the only other countries over 20%. The falling number of working-age adults in Korea will have major consequences for the labour market, GDP and pensions.

Economic differences between Korea and the most developed OECD countries have been sharply reduced over the last 20 years, with GDP consistently growing above the OECD average even though recently the difference in real GDP growth rates has been declining (Figure 1.7). Over this period real GDP growth has averaged 3.9% per year in Korea compared to 1.7% across the OECD as a whole.

Based on long-term projections, future GDP growth will be lower than in the past in Korea, with the declining working-age population being an important contributing factor. According to the latest estimates from the Korean Development Institute, real GDP growth will gradually decline to around 1.4% per year in the 2030s, 1.0% in the 2040s, 0.8% in the 2050s and then around 0.5% thereafter. OECD long-term projections are less optimistic, with growth at just under 1.0% per year in the 2030s on average, falling to 0.3% and 0.1% in the next two decades (OECD Economics Department long-term projections, Autumn 2019). This compares with 1.3% on average across the OECD in the 2030s and 1.2% thereafter.

Rapid population ageing in Korea is likely to have an adverse effect on government’s net financial position. Assuming that the NPS legislation remains unchanged, the reserve fund for the NPS is projected to peak in 2034 as a share of GDP, and be insufficient to cover pension spending from 2042. The fund is projected to be depleted in 2057 based on the 4th Actuarial Review (2018) of the NPS (Chapter 3).

The level of public debt in Korea is low compared to other OECD countries, although it has increased over the last few years (Figure 1.8). Before the COVID-19 crisis, general government debt currently stood at around 42% of GDP, with only five OECD countries recording lower levels. By comparison the debt ratio in Japan was nearly six times higher at just under 234% of GDP, with Greece also just under 200%.

There is substantial room to improve employment performance in Korea. In 2020, 70.1% of 20-64 year-olds worked in Korea, against 73.1% in the OECD (Figure 1.9). While the total employment rate is higher than for example in Spain (65.7%) and Italy (62.6%), it is still one of the lowest in the OECD and well below top-performing countries, such as Japan (82.2%) and Switzerland (82.5%).

The employment gender gap is wide compared to other countries, with a 19.6 percentage point gap between men and women against 12.1 percentage point on average in the OECD. Only, Chile, Colombia, Costa Rica, Italy and Türkiye have a larger gap. Employment rates among women in Korea are thus considerably below the OECD average (60.2% against 67.0% in the OECD).

Relatively low employment in Korea is heavily influenced by low rates for younger age groups while the employment rates for older workers are well above the OECD average, at 66.6% and 61.7%, respectively, among the 55-64 age group (Figure 1.10).

Among this older age group, the gender gap is also one of the highest in the OECD (Figure 1.11). Employment rates among women are 22 percentage points lower in Korea than among men compared to an OECD average of 14 percentage points, and with only eight countries having a larger gender gap. Of all OECD countries only Finland and Estonia have higher employment rates for women than for men in this age group.

The employment rate of the 55-64 age group increased by 9 percentage points between 2000 and 2020 in comparison to 3 percentage points for the 25-54 age group. While this is encouraging, the increase for the 55-64 age group was about twice larger in the OECD on average where the employment rate grew from 44.1% in 2000 to 61.7% in 2020 (Figure 1.12).

In Korea, employment after age 65 remains high relative to that in other OECD countries: 49% of people aged 65 to 69 are employed in Korea, second only to Japan at 50% and well above the OECD average of 23% (Chapter 2). For those aged 70 to 74 the employment rate is now the highest in the OECD at 37% against an OECD average of 11%, with Colombia, Israel, Japan, Mexico and New Zealand being the only other countries above 20%.

High employment for older workers in Korea partly reflects individual income constraints triggered by very low old-age benefits, as these older workers have only been able to contribute to the pension system for a short period of time. With the pension system yet to mature, most of today’s older workers in Korea hence need to continue working at a much older age than in many OECD countries. They effectively finally leave the labour market at an average age of 66 years for men and 65 years for women, at least two years above the OECD averages and ranking Korea 9th and 4th highest for men and women, respectively (Figure 1.13).

Furthermore, many workers have “retired” early from their main job (for most before reaching age 55), after which they have had to find alternative employment, often in lower paid employment or self-employment, resulting in lower contribution levels to the pension system. The average age of separation from the main job (i.e. the job the worker has held for the longest duration) for workers aged 55-64 (including the self-employed) was as low as 51.2 years for men and 47.7 years for women in 2021 (Table 1.1). This practice of being “retired” early is covered in more detail in Chapter 2.

Given “mandatory retirement” practices, only 25% of Korean workers aged 55-59 in 2020 were working with the same employer they had five years before – just under half the OECD average of 52% (Figure 1.14). Although hiring rates of older workers are by far the highest of any OECD country, at 31% compared to 8% on average, the quality of the jobs is generally much lower than that of previous employment, as shown below.

Consistent with the decline in job security with age, the level of self-employment amongst older workers is much higher than among the total working-age population and one of the highest in the OECD. In 2020, 35% of those aged 55 to 64 in employment were self-employed compared to 22% on average across the OECD (Figure 1.15). Only Colombia, Costa Rica, Greece, Mexico and Türkiye have higher levels among older workers. For all workers, the share of the self-employed in total employment is equal to 25% against an OECD average of 17%.3 The prevalence of self-employment has, however, been declining in Korea in recent years among older workers, as in 2010 45% of them were self-employed, 10 percentage points higher than currently, with only Chile, Portugal, Slovenia and Türkiye having a larger decline over the same time period.

Moreover, often the new jobs are poor-quality, highly insecure and low-paid jobs. Just over one-third of workers aged 55-64 in Korea hold a non-permanent job, compared with an OECD average of under one-twelfth (Figure 1.16). There has been a downward trend in Korea over recent years as in 2010 the incidence was four out of ten workers, with the OECD average also falling from one-tenth. Precarious work is even more common among older workers, with six out of every ten of those aged 65-69 in non-permanent positions, i.e. roughly three times the OECD average (OECD, 2018[4]).

Results from the OECD Survey of Adult Skills (PIAAC) (OECD, 2019[5]) suggest that older workers in Korea are not well equipped to deal with recent technological advances. This further impedes the ability of older workers to remain in their primary employment position at later ages. While this is partly related to Korea’s rapid economic catching-up, it also indicates that many workers have been unable to learn on the job to a sufficient level. This is supported by the relatively low level of participation in education or training amongst older workers in Korea compared with other countries (Figure 1.17).

Figure 1.17 also suggests that there is a strong willingness to learn among older workers in Korea. Slightly more than 20% of those aged 55-64 report unmet demands for training, more than twice the OECD average for this age group. If all of these training or education demands were met, Korea would rank among the best-performing countries.

The main reason quoted for being unable to participate in training is being too busy at work, which is applicable across all age groups, and accounts for 40% of all response reasons for those aged 55 to 64, i.e. the highest across all OECD countries. With the common practice of companies ending contracts of workers at early ages, there is little incentive for these companies to invest in upskilling their older employees, resulting in them being left behind in terms of new skills and technological advances. This therefore makes it even less likely that they will be able to continue in their primary occupation later in their career.

Being too busy at work is supported by the fact that workers in Korea work much longer hours than in the majority of other OECD countries, across all groups including for older workers.4 Although the number of working hours in Korea is still large, it has declined steadily in recent years. According to the Survey on Labour Conditions, average working hours across all groups were 217 hours per month in 1993, falling to 198 hours in 2006 and 180 hours in 2017.5


[2] Korea, S. (2018), Retirement Pension Statistics in 2018, Statistics Korea, https://kostat.go.kr/portal/eng/pressReleases/11/1/index.board?bmode=read&bSeq=&aSeq=382089&pageNo=1&rowNum=10&navCount=10&currPg=&searchInfo=&sTarget=title&sTxt=.

[3] Moon, H. (2001), The Korean Pension System: Current State and Tasks Ahead, Korean Development Institute.

[1] OECD (2021), Pensions at a Glance 2021: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ca401ebd-en.

[5] OECD (2019), Skills Matter: Additional Results from the Survey of Adult Skills, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/1f029d8f-en.

[4] OECD (2018), Working Better with Age: Korea, Ageing and Employment Policies, OECD Publishing, Paris, https://doi.org/10.1787/9789264208261-en.


← 1. The old-age pension is available with 20 years of contributions with a so-called partial pension being available with a minimum of 10 years of contributions. The design of the NPS was meant to target a replacement rate of 70% initially after 40 years of contribution. This has been reduced to 40% for future generations.

← 2. The severance payment is a lump-sum benefit of one month of the base salary for every year worked for that particular company, and is eligible to full-time workers who have worked for at least one year. The base salary is computed as an average of the prior three months of salary. There is no guarantee of receiving payment if the firm goes bankrupt, as the majority of severance accounts are not funded. The severance payment system also excludes workers in smaller firms (fewer than five employees). The scheme therefore favours workers with a long history of employment within solid companies. Ultimately though the severance payment system cannot be regarded as a retirement account able to guarantee income in later life as even if an individual remains at one company throughout their career they can withdraw funds from the severance account during employment for new home purchases or medical expenses, for example.

← 3. In total across all age groups, the share of self-employment in total employment in Korea is the sixth highest among OECD countries.

← 4. The average number of hours actually worked per worker in 2018 was 1993, well above the OECD average of 1 734 hours. Koreans aged 55-64 years’ work on average 39.2 hours per week for women and 44.4 hours for men. By comparison the OECD averages for older workers are 35.3 hours for women and 40.9 hours for men, with older women in Korea working longer than in every OECD country except for Colombia and Türkiye and men only ranking behind Colombia, Mexico and Türkiye. While accustomed to long working hours, they also choose to work longer to complement for their low hourly wages.

← 5. This decline is a direct result of the reduction in statutory weekly working hours from 48 hours prior to 1990, to 44 hours until 2004 and then 40 hours currently, though an additional 12 hours of overtime is still possible.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2022

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.