Public pensions are one of the largest components of public spending across OECD countries. Demographic shifts are likely to put further pressures on spending. Ageing populations and persistently low fertility rates are leading to a worsening old-age to working-age ratio. However, demographic pressures do not automatically translate into higher pension spending, and policy choices play a decisive role. This chapter sets out some of the current measures being implemented by governments. These fall broadly into a number of categories including tightening public pension eligibility, including though increases in retirement ages and greater means testing; increasing the coverage and scope of complementary pension schemes to reduce pressure on publicly funded schemes; and fostering incentives to work beyond official retirement age. Governments are also increasing reliance on, and expected returns from, private pensions. Many are introducing sustainability mechanisms to automatically adjust pensions without the need for new legislative changes.
Restoring Public Finances
Enabling Effective Government
2. Old-age pensions
Copy link to 2. Old-age pensionsAbstract
Many reported measures under the 2026 RPF Survey relate to reducing eligibility for public pensions, expanding both contribution levels and number of contributors to complementary pension schemes, and offering incentives to work beyond statutory retirement ages. Measures implemented as part of longer-term savings strategies linked to demographic changes are less common, although some respondents have implemented measures, such as automatic pension adjustments, linking retirement ages to life expectancy, and triggering balancing mechanisms to ensure long term sustainability of pension schemes.
Reform initiatives and savings measures
1. Tightening public pension eligibility
Increasing both statutory and early retirement ages, as well as increasing contribution requirements to receive full public pensions.
Tightening of means testing, through the introduction of stricter criteria for both income and asset levels.
Linking public pension amounts to key parameters influencing pension spending, such as asset-liability ratios or growth pension spending.
2. Increasing coverage and scope of complementary pension schemes
Increasing in contribution rates for both employers and employees, as well as increase contribution period requirements.
Expanding of eligibility for contributory systems, including for short-term workers.
Using of incentives to enrol in contributory schemes, including through penalties and automatic enrolment.
3. Fostering incentives to work beyond retirement age
More generous combinations of earnings and pension withdrawal, so that retirees benefit from working past retirement age without loss of income.
Increasing in incentives to work past retirement age through pension deferral bonuses and discounts on taxes and pension insurance contributions.
Tackling job retention issues amongst older workers.
2.1. Recent trends in pension spending
Copy link to 2.1. Recent trends in pension spendingPension expenditure is one of the largest expenditure items in OECD countries, and in many the single largest expenditure item. In 2023, the average OECD country’s expenditure on old age and survivor related programmes (which includes both pension expenditure and related old-age benefits such as long-term care services) was equivalent to 9.4% of GDP, using COFOG data. This represents around 21% of general government spending (Figure 2.1). Such expenditure was not marked by significant increases unlike other areas of social welfare during the COVID years, but remained broadly stable between 2015 and 2023.
Figure 2.1. Pension expenditure is among the largest public expenditure items
Copy link to Figure 2.1. Pension expenditure is among the largest public expenditure items
Note: No data available for Canada, Chile, Mexico, New Zealand, South Korea, Türkiye, and the United States.
Source: Panel A. OECD Public finance by function - government at a glance indicators, yearly updates https://data-explorer.oecd.org/s/4d0. Panel B. OECD Public finance by function - government at a glance indicators, yearly updates https://data-explorer.oecd.org/s/4d0, and OECD National Accounts at a Glance, Chapter 6A: Government expenditure by function https://data-explorer.oecd.org/s/4d1.
During the 1950s-80s, large baby-boom cohorts, combined with rapid growth in the labour force, meant that the worker-to-retiree ratio stayed relatively constant. Major pension system expansions and reforms in the 1960s and 70s were implemented under these favourable demographic conditions, allowing retirees to receive pension benefits at an actuarial value exceeding what they had contributed during their working lives, while also subsidising lower-income retirees that did not have time to build up pension rights. With the extension of life span at older ages, the first phase of the ageing transition had a positive effect on pension sustainability, as longer lives translated in a larger share of the working age population (15-64) in a first phase. However, from the 1990s to mid-2000s, this trend of a growing share of the working age population in the total population reversed, although the speed and exact timing differs across countries. In parallel, fertility rates in most OECD countries continued to fall, below replacement level (around 2.1 children per woman).
Fertility rates below replacement level means that each generation is smaller than the previous one, leading to a higher old-age to working-age ratio in the long term. In other words, a smaller number of people are working and therefore paying taxes, while at the same time ageing populations mean more people are receiving pensions. Almost all OECD countries are expected to have ratios of at least 40% by 2050 (Figure 2.2), meaning there will be at least 40 people aged 65+ per 100 people aged 20-64. Without significant reforms to pension systems, this combination of an ageing population and shrinking workforce will reduce the number of taxpayers while increasing the pension burden.
Figure 2.2. The ratio of older people (65+) to working age people (20-64) is expected to worsen significantly by 2050
Copy link to Figure 2.2. The ratio of older people (65+) to working age people (20-64) is expected to worsen significantly by 2050Old-age to working-age ratio, OECD countries
Note: The old-age to working-age ratio measures the number of people aged 65+ per 100 people aged 20-64 in a given country.
Source: OECD Pensions at Glance 2025 (OECD, 2025[1]), Figure 1.5, which uses data from the Nations World Population Prospects 2024: http://population.un.org/wpp/.
This worsening old-age to working-age ratio creates risks for the sustainability of pension systems organised on a pay-as-you-go basis. This is the case of the vast majority of the compulsory schemes organised with public funding, and hence it creates pressures on public finances. However, it does not mean that such systems will inevitably become unsustainable. Pension systems organised on a pay-as-you-go basis can remain sustainable despite demographic challenges with various policies that help increase the sustainability of such pension systems, thus relieving pressures on public finances. This includes boosting labour market participation – especially among women and those at or near retirement age and improved productivity.
As indicated in Figure 2.3, demographic pressure and projected changes in pension spending do not necessarily align. For example, Japan’s projected increase in the old-age dependency ratio is above the OECD average, yet its projected pension expenditure is below the OECD average. This is likely attributable to the introduction of an automatic mechanism adjusting pensions based on life expectancy and number of contributors. On the other hand, Canada’s projected increase in the old age-dependency ratio is below the OECD average, yet its projected pension expenditure is above average. This is likely due to the expansion of benefits in the Canada Pension Plan and Quebec Pension Plan systems, with an associated increase in contributions and ceilings, which are projected to reduce reliance on the tax-payer-financed Guaranteed Income Supplement (GIS) for low-income seniors. This highlights that policies, and not just demographics, drive cross-country differences in pension system pressures.
Figure 2.3. There is limited correlation between demographic and pension pressures on government
Copy link to Figure 2.3. There is limited correlation between demographic and pension pressures on governmentPercentage points. Blue lines represent average pension expenditure and average old-age dependency ratio
Note: Data for Colombia, Iceland, Israel, Mexico, Türkiye, Switzerland not available.
Source: Adapted from Figure 1.5 and Table 8.4 in Pensions at a Glance (OECD, 2025[1]).
Figure 2.4, using data from the European Commission’s 2024 Ageing Report, shows a similar picture. For example, despite the fact that Italy has one of the highest projected increases in old-age dependency ratio in the OECD, it is forecast to see stable pension expenditure between 2023 and 2050. This can be attributed to a comprehensive set of pension reforms since 1996, which include reductions of replacement rates, and increases in the retirement age, and automatic updating of retirement age in accordance with life expectancy growth.
Figure 2.4. Despite demographic pressures, several OECD countries are predicted to see only small increases or even decreases in pension expenditure, while others can expect significant increases by 2050
Copy link to Figure 2.4. Despite demographic pressures, several OECD countries are predicted to see only small increases or even decreases in pension expenditure, while others can expect significant increases by 2050Projected percentage point change in pension expenditure between 2023-2024 and 2050
2.1.1. Taking into account tax expenditure related to pensions
Tax expenditures, which reflect policy-related tax decisions that depart from “normal” tax legislation, often in the form of deductions, credits and exemptions, play a role in total pension expenditure. One common example of tax expenditures are tax incentives for retirement savings, where countries allow workers to pay reduced taxes on their contributions to private pension plans. Tax expenditures are often not considered in the annual budget process, and their uptake rates can be unpredictable. This means they can have an unpredictable impact on public finances. For example, if a country reduced public spending on public pensions, leading to an increase in private savings for pensions, this could be seen as positive for fiscal sustainability. However, if these savings occurred in accounts which offered significant tax incentives, the cost of these tax expenditures could offset some of the savings from reduced public spending. Overall tax expenditures remain of a second order of magnitude, as they are estimated to have evolved from around 0.1% of GDP in the early 2000s to 0.3% of GDP in recent years (Redonda and von Haldewang C., 2025[2])
2.2. Reform initiatives and savings measures
Copy link to 2.2. Reform initiatives and savings measuresThis section provides detail on the measures that were reported in the 2026 RPF Survey. Figure 2.5 provides an overview of recent policy measures to limit pension expenditure and encourage long-term pension sustainability, drawing on the responses to the Survey. As in most participants, there is a combination of public and private provision of pensions, these measures can either be related to reducing public expenditure on pensions, or increasing private provision of pensions. The measures presented below reflect submissions from respondents in terms of measures that will have an effect in 2025-2026. While in most cases these reflect direct changes as part of national budget submissions, some respondents also reported reforms from years prior to 2025 that have had an impact in 2025-2026. In most cases, this involves increases to retirement ages as part of previous legislation that required incremental increases to retirement age over time or factoring in the broader impact of some past reforms that have had a lagged impact.
The measures reported in the RPF Survey focus on fiscal savings. A full discussion of trends in pension reforms is offered in the OECD Pension at a Glance publication (OECD, 2025[1]).
Figure 2.5. Key reforms and saving measures related to pensions
Copy link to Figure 2.5. Key reforms and saving measures related to pensionsMeasures approved or submitted to parliament for the fiscal years of 2025 and 2026
Note: Results based on 39 RPF Survey responses. Measures reported as “other” in the RPF Survey have been split into the following five sub-categories, based on the qualitative information provided by respondents.: “Increasing reliance and expected returns of private pensions”, “Increasing employment rates for older workers”, “Expanding contributory system coverage”, “Introducing sustainability mechanisms” and “Other”.
Source: 2026 OECD Survey on Restoring Public Finances, Question 1: Public Pensions.
2.2.1. Increasing the statutory retirement age
As part of savings measures, several respondents reported increasing their statutory retirement age. This increases the amount of time that people work (and therefore pay taxes) and reduces the amount of time people are eligible for public pensions, therefore both increasing government revenue and reducing expenditure. Examples include:
Belgium will increase the statutory retirement age from 65 gradually up to 67 by 2030.
Denmark will increase the statutory retirement age from 69 to 70, beginning in 2040.
Czechia, where there is a system of gradual increase of retirement age up to 67.
Latvia, where the retirement age was changed to 65 in 2025.
Norway, where the retirement age is gradually increasing, starting with the cohort born in 1964 as of 2025 (and in addition the mandatory retirement age in the state sector will be increased from 70 to 72 years), which will allow people to continue working longer and will standardised these rules across public private sectors).
Slovenia, will increase the retirement age from 65 to 67.
In Chinese Taipei, civil servants’ retirement ages have been increased in a phased manner and will reach 65 in 2026.
All these savings measures are helpful and part of broader OECD trends to increase retirement ages (Figure 2.6). There is also proof that increases in both early and statutory retirement ages have led to large employment increases. In the Netherlands, increases in statutory retirement age from 65 in 2012 to 66 years 4 months in 2019 led to a significant increase in the employment rate (Rabate, Jongen and Atav, 2024[3]). Similarly, in Austria, an earlier increase led to 9.8 and 11 percentage point increases for men and women respectively. This response was strongest among high-wage and healthy workers, while lower-wage and less healthy workers were more likely to seek alternatives. In Japan, increases in the pensionable age for the Employees’ Pension Insurance from 60 to 65 (for women, this will reach to 65 in 2030) led to a 7-8 percentage point increase in employment amongst affected individuals (Nakazawa, 2022[4]). Research from Denmark, Norway and Switzerland indicates that these changes impact spouses also even if they are not directly impacted, as couples adjust to retire together.
However, in many countries, average effective retirement remains below the statutory retirement age. This can be due to pension systems allowing early exit pathways, often tied to length of contributory period, or labour market circumstances (i.e. those working in hazardous employment). In countries where there are significant discrepancies between statutory and early retirement rates, it can be useful to reduce earlier retirement rates. This can be done by raising the minimum years of contributions required to receive a full pension, and introducing disincentives to retire early (or incentives to retire later) to increase job retention among older workers.
Figure 2.6. The normal retirement age will be rising in half of OECD countries for men
Copy link to Figure 2.6. The normal retirement age will be rising in half of OECD countries for menNormal retirement age for men entering the labour market at age 22 with a full career
Note: The normal retirement age is calculated for an individual with a full career from age 22. “Current” refers to people retiring in 2024. “Future” refers to the age from which someone is eligible to full retirement benefits from all mandatory components (without any reduction), assuming a full career from age 22 in 2024. Educational credits are not included. For better visibility, the scale of this chart excludes the lowest observed value of 52 for current normal retirement age in Türkiye.
Source: OECD Pensions at a Glance (2025[1]), Figure 1.12.
For example, in the RPF Survey, several respondents have early retirement rules in place, which allow people to retire earlier under certain conditions. Some respondents have adjusted their early retirement rules to restrict these:
Austria changed eligibility rules for early retirement in summer 2025, including an increase in the general early-retirement age from 62 to 63.
Denmark’s higher statutory retirement age (mentioned above) will also lead to adjustments in the eligibility ages of certain early retirement schemes.
Two of Italy’s early retirement schemes (one which allowed early retirement for those that had 41 years of contributions, and one that allowed early retirement for women under certain conditions) have not been extended in the 2026 Budget. However, those that had met the requirements within the deadline established by previous legislation remain eligible.
In Slovenia, a person with 40 years of contributions will be able to retire with a full pension from age 62 instead of 60.
2.2.2. Freezing or changing rules for adjusting public pensions
As part of savings measures, some respondents have reduced the pension accrual rate so that each year of work adds slightly less to the public pension entitlement than before:
In Czechia, the accrual rate will change from 1.5% to 1.45%.
Slovenia has adjusted its indexation formula, so that it will gradually depend more on inflation than wage growth. As inflation generally increases at a slower rate than wages, this will lead to slower pension growth.
Some respondents have focused reductions in pension adjustments on those with higher pension levels:
In Austria, only pensions up to EUR 2 500 are adjusted by a percentage amount, whereas those above receives a fixed amount.
In Belgium, pensions up to EUR 5 182 will remain fully indexed, but pensions exceeding this will receive only a fixed amount. As a result, those with high pensions do not receiving large adjustments, reducing pressure on government finances.
There are also several examples of adjustments to the calculation basis for retirement pensions:
In Belgium, the pension for government employees is currently calculated based on the best 10 years of retirement. This will be gradually increased until it reaches 45 years in 2062, effectively the concept of “best of” is abolished. As salaries are generally higher later in life, this is likely to reduce the pension expenditures.
In Czechia, currently 100% of earnings below a threshold (equivalent to around EUR 800 a month) are considered in the calculation formula used to determine how much of a person’s earnings are counted towards their pensions. This will slowly be reduced, reaching 90% by 2035. Above this threshold, 26% of earnings are considered, and above the second threshold earnings are not considered.
In Ireland, there will be a phased removal of the Yearly Average Method for calculating state pension entitlements from 2025. This will be replaced by the Total Contributions Approach under which pension benefits are directly proportional to the total number of contributions made over a person’s working life. This is expected to moderate future pension costs.
In Slovenia, the pension base reference period has recently changed from the 24 best consecutive years to 40 years minus 5 worst years.
The Slovak Republic will freeze the “13th month pension,” between 2026 and 2028. The 13th month pension is an annual bonus benefit equal to the average monthly pension received the previous year.
In the Netherlands, a supplementary payment on top of the basic state pension has been abolished.
2.2.3. Increasing contribution rates and time periods
As part of savings measures, one of the most common changes to the design of pension plans is an increase in contribution rates as well as time periods. These reforms are among the most common changes to policy parameters and are aiming to put some of the publicly funded pay as you go schemes on a more sustainable fiscal path. This is also a reflection that life expectancy has been increasing in ways that were not factored into the financial balance of the systems. Therefore, increasing the length of time that individuals must contribute towards their scheme, is one way to both achieve savings and to promote greater labour force participation at older ages (which is also part of a separate set of measures discussed below).
As with increases to contribution time, increasing contribution rates increases the amount individuals have in their contributory pension plans and so reduces their reliance on public pensions. Examples of respondents mentioning such saving measures involving changes of design in pension plan include:
Ireland increased the contribution rate towards the State Pension for both employers and employees by 0.1 percentage points in 2024 and will increase it by a total of 0.7 percentage points over five years, with the aim of supporting the long-term sustainability of the Social Insurance Fund.
In Japan, the upper limit for monthly salary used as the basis for pension insurance contribution level calculations is being increased from JPY 650 000 to 750 000, increasing contribution amounts for those in this salary range.
Korea plans to increase its national pension scheme contribution rate, currently 9%, to 13% by 2033 in annual increments.
In Luxembourg, the total contribution rate rose from 24% to 25.5% of gross salary in 2026, and will remain in place until at least 2032.
Other changes to pension plan designs include:
Ireland introduced a contributory scheme for employees in 2026. This will involve automatic enrolment for employees earning above EUR 20 000 and not already in a scheme, with contribution rates gradually increasing over 10 years.
In the Netherlands, the Future of Pensions Act replaces the former defined-benefit system with a contribution-based system
Czechia has increased incentives to take up pension insurance by halving the penalty for early retirement for those who have at least 45 years of pension insurance.
2.2.4. Increasing employment rates for older workers
As part of savings measures, several respondents have introduced incentives for older workers to continue working past statutory retirement age. There are often reduced job opportunities at older ages. In 2022, 9.5% of employees aged 55-64 were newly hired, compared to 15% of workers aged 35-44 (OECD, 2025[5]). As such, it is important to improve labour market opportunities and actual participation for older workers (see Chapter 3 in (OECD, 2025[5]), and also (OECD, 2019[6])). This usually takes the form of financial benefit for those who work at least part time when retired.
In terms of concrete examples of saving measures from the RPF Survey:
In Austria, the government has introduced the option to reduce work hours and receive partial pension benefits. The Austrian Older Workers’ Employment Package (Älteren Beschäftigungspaket), which aims to increase employment amongst older workers by tackling employability and job retention issues (Box 2.1).
Belgium is abolishing a system which gave a lump-sum bonus for working past early retirement. This will be replaced with a bonus-penalty system, with progressive reductions (up to 5%) in statutory pension for people who retire early, and progressive increases (up to 5%) for people who take their statutory pension after the statutory retirement age.
In Czechia, old-age pensioners who receive a full old-age pension and work can claim a discount on their pension insurance contributions.
In Luxembourg, the introduction of a partial pension allows employees to work part-time while receiving part of their pension. Furthermore, a tax allowance is available for those who qualify for early retirement but voluntarily continue working until the statutory retirement age.
Ireland allows individuals to defer drawing down their state pension up to four years past the pension age. This allows people to continue to work and potentially improve their contribution record for the State Pension Contributory part to receive a higher payment upon retirement. The government has now increased the benefit level if claiming is deferred, incentivising people to do so.
Box 2.1. Austria’s incentives for older workers
Copy link to Box 2.1. Austria’s incentives for older workersAustria’s pension reform has made several steps to incentivise people to continue working beyond retirement, In recognition of the difficulties that older people often face in gaining and maintaining employment, the government has also introduced measures which aim to get more people over 60 into employment and keep them employed. This will include increasing monitoring of older workers’ labour market situations, sending companies targeted information and support to enable them to provide training for older employees and implement preventive measures to help older employees remain healthy.
In addition, the government is planning to introduce a tax-free allowance for additional income during retirement up to EUR 15 000 per year, and to abolish employee contributions to pension insurance for employed persons of regular retirement age (so-called Aktivpension).
Source: 2025 Restoring Public Finances Survey; (Bundeskanzleramt, 2025[7]); (Vienna.at, 2025[8])
2.2.5. Harmonising civil service and private pensions systems
As part of savings measures, several respondents have identified measures to integrate or harmonise fragmented public pensions:
In Belgium, some preferential pension arrangements for civil servants are being adjusted to be more in line with private sector retirement plans. This includes:
Coefficients which allowed the early retirement conditions to be reached more quickly, which are being reduced.
The abolishment of the statutory sickness pension for civil servants.
The raising of preferential early retirement ages, that exist for certain categories of civil servants.
The abolishment of automatic adjustments to civil-servant pensions based on wage developments, with revaluations instead aligned with general pension indexation.
In Norway, the mandatory retirement age for government employees will be increased from 70 to 72 years, as mentioned above. This will harmonise the upper age limit in the public and private sector.
In Italy, public pensions for the armed forces, the police and fire personnel will see an increase of three months in age requirements, bringing the requirements more in line with other pension schemes.
2.2.6. Increasing reliance and expected returns of private pensions
Several respondents are trying to achieve savings in public funds through encouraging greater reliance on private pension funds and increasing the returns to private pension funds.
For example, q number of respondents are introducing plans to promote higher returns from the investment of pension funds, through increases in competition and changes in regulation.
In Chile, the Law no. 21.735 of 2025 expands a mechanism to 10% of the affiliates, which was previously restricted to new entrants, whereby they can benefit from subjecting their Pension Fund Administrators (private companies responsible for the management and investment of pension funds) to a competitive public tender every two years. This will help to lower the management fees and to reduce administrative costs of investment management, thus promoting higher returns for beneficiaries. Employer pension contributions are also set to rise from 1.5% to 8.5% by 2033.
In Costa Rica, a reform to the Mandatory Supplementary Pension Scheme (ROP) has been introduced, aiming to improve risk management and investment returns. The adjustment is expected to provide greater protection for members nearing retirement and optimise returns for younger members. If approved, it could lead to savings, with higher net returns leading to reduced public pension pressure.
Finland’s pension reform in 2025 has similar aims of improving the investment returns of pension funds. It plans to do so by changing the regulation of investment activities of private-sector pension institutions.
Latvia has introduced stricter criteria for the setting of pension management fees, introducing a cap on the permanent part of the fees charged. This aims to reduce lower long-term costs for participants so that they retain more of their investment returns. It is also adjusting how contributions to its pension system are split until 2028. 5% of earnings will go to the funded pension pillar and 15% to the unfunded state system, meaning slightly more contributions are now going to the state scheme.
In Thailand, the government has introduced a lottery policy for the National Savings Fund. Under the scheme, people can purchase lottery-linked savings products, where the full amount paid for the lottery ticket is contributed into the purchasers’ individual retirement accounts and invested by the Fund. The total accumulated fund, including returns, will be paid into individual retirement accounts when the lottery purchasers reach age 60. This allows the government to reduce the budget on elderly welfare support to households.
In Latvia, an individual must have 20 years of insurance to be eligible for the old-age pension from 2025 onwards.
In Luxembourg, to bring the effective retirement age closer to the statutory age, mandatory contribution periods will be extended by eight months by 2030.
2.2.7. Introducing or applying stricter criteria for means testing
Means-testing can be helpful to adjust access to public pensions and benefits. In many cases, this has focused on income, but there are also options to use asset-based mechanisms. This is particularly useful amongst older populations, where accumulated wealth can lead to large differences in how well-off different households are, even when income differences are small (OECD, forthcoming[9]).
RPF Survey responses include the following examples:
Australia uses what is known as a deeming rate to determine what proportion of assets should be used in determining pension eligibility. From 2025, this deeming rate will increase from 0.25% to 0.75%, and from 2.25% to 2.75% for assets above AUD 64 200, increasing the amount of assets considered in determining pension eligibility. This will reduce the number of people eligible for certain pension programmes, reducing state outlays.
Finland introduced stricter criteria for its pension housing allowance. The new criteria will see income and assets reduce the amount received more than they do at present, particularly for the wealthiest beneficiaries.
2.2.8. Introducing sustainability mechanisms
As part of longer-term savings and fiscal rebalancing strategies, some respondents have introduced sustainability mechanisms which adjust pensions according to set thresholds. These adjust pension amounts based on economic or financial variables, to ensure that expenditure on pensions is within a country’s means. These mechanisms are part of broader automatic adjustment mechanism which involve predefined rules that adjust key pension parameters in response to demographic, economic or financial indicators, without the need for legislative changes (see Box 2.2 below).
Box 2.2. Automatic adjustment mechanisms
Copy link to Box 2.2. Automatic adjustment mechanismsAutomatic adjustment mechanisms were implemented in response to sustainability issues following the 2008 financial crisis. About two-thirds of OECD countries employ some form of AAM in pension schemes One type of AAM involves linking changes in the retirement age to life expectancy. Currently, the countries with the highest future retirement age all link retirement age to life expectancy, including Finland, Portugal and the Slovak Republic. Different approaches are possible. In some cases, countries link their statutory retirement age one-to-one to life expectancy, meaning a one-year increase in life expectancy leads to a one-year increase in statutory retirement age. In other cases, the statutory retirement age is increased by two-thirds of the increase in life expectancy.
Some AAMs adjust pension benefits based on demographic trends. Such mechanisms can be effective. For example, Portugal’s Sustainability Factor, which adjusts early retirement pensions according to life expectancy changes, is estimated to reduce early pensions by 30% by 2066, while Finland’s Life Expectancy Coefficient, which adjusts initial benefits according to life expectancy), is projected to reduce new pensions by 13% for cohorts entering the labour market in 2020 by 2066. Other examples tie pension benefits to total contribution levels, GDP levels or demographic ratios.
Another subset of AAMs help ensure pension schemes maintain balanced budgets. These generally reduce pension levels, reduce pension growth levels, or increase contribution rates if the ratio of pension assets to liabilities reaches a defined level. For example, in Finland, if reserve funds fall below 20% of expected PAYG pension expenditure in the coming year, the contribution rate is automatically increased to the level required to meet the 20% threshold.
In Sweden, if pension liabilities exceed assets, indexation of pension benefits is automatically reduced. Each year, a balance ratio is calculated, considering the balance between contribution assets, buffer funds (where surplus is invested) and pension liabilities. If the balance ratio is more than 1, it means the government’s assets are sufficient to cover its liabilities. If it is less than 1, it means liabilities exceed assets, triggering the balancing mechanism. This is a financial safeguard which adjusts the indexation of both pension entitlements (for those still working) and current pensions. A balance index is calculated for each year thereafter until it reaches at least the same value as the income index. In 2025, the Swedish Pensions Group agreed to introduce a surplus mechanism to this system, enabling excess funds to be distributed to pensioners. This is based on a principle that the surplus within the pension system should not be greater than what is needed to finance pensions long term, and will only be distributed when assets in the system are at least 15% higher than its liabilities.
AAMs can be more sustainable over time than discretionary changes, as changes do not require legislative approval once the initial mechanism is approved. However, it is important that their value is communicated effectively to the public. This can be done by making the case that such mechanisms reduce uncertainties, are less erratic, and are more equitable across generations than discretionary changes .While they are relatively underrepresented in the recent pension reforms reported in the RPF Survey, their ability to adjust automatically in response to demographic changes means they are highly advantageous for ensure pension sustainability in the medium to long term, with overall implications for public finance sustainability.
The RPF Survey includes the following examples:
Austria has introduced a sustainability mechanism, requiring the federal government to take measures in the pension sector if a certain budget path for public pensions is not adhered to by 2030.
Finland’s 2025 pension reform includes an inflation stabiliser, with a cap on pension indexation, that limits index increases to earnings-related pensions if the pension index rises faster than the wage coefficient over a two-year period.
2.2.9. Expanding contributory system coverage
As part of savings measures, other respondents are extending eligibility for contributory systems, resulting in a de-facto revenue increase. In Japan, for example, part-time workers (those who work 20 hours or more per week) historically only need to join employee pension plans if they work for a company with 51 or more employees and earn a monthly wage of at least JPY 88 000 (equivalent to EUR 520). The government plans on abolishing both these requirements.
In Belgium, certain periods of unemployment could previously under certain conditions be equated to the normal notional wage. In the most recent reform, this calculation will be changed so that they count less from 2027 onwards. Socially necessary periods such as sickness and parental leave do not change.
2.2.10. Other savings measures, including adjusting tax relief
Incentivising private savings can also be a useful tool to reduce dependence on public programmes. By shifting part of retirement financing to private savings, governments can more easily limit future pension liabilities without excess adverse consequences for future retirees. Private savings can be incentivised through specific vehicles, including those that allow for a guaranteed income stream at retirement, such as annuities, or those that unlock liquidity stored within housing, such as reverse mortgages (OECD, forthcoming[9]). This involves a combination of regulatory as well as tax relief measures.
Many countries already have systems in place that offer reduced taxes on savings put aside specifically for retirement. Such incentives can be beneficial, but their cost through revenue foregone should be considered. As a result, participants considering savings may also envisage adjusting tax relief, as illustrated through the RPF Survey.
Several respondents also reduced tax relief related to pensions, as reported in the RPF Survey:
Finland has abolished the tax deduction for voluntary pension contributions
France’s 2026 Finance Bill replaced the 10% income-tax allowance currently granted to retirees with a flat rate of EUR 2 000.
Korea has reduced eligibility for the tax-free comprehensive savings account. While it was previously accessible to all those aged 65 or older, it will now only be available to those eligible for the basic pension (around 70% of the elderly).
New Zealand’s KiwiSaver programme provides incentives to save privately for pension purposes. The government has removed the government contribution for all KiwiSaver members with income over NZD 180 000 per annum, and halving the rate of the government contribution to 25 cents per dollar to a maximum of NZD 260.72.
The United Kingdom has targeted winter fuel payments so that only those in receipt of pension credit or certain other income-related benefits will receive them.
In addition, several measures were reported which do not fit into the above categories:
Canada’s 2025 budget adjusts net benefit accrual rate to 2% for federal public sector workers as well as members of the Canadian Armed Forces and of the royal Canadian Mounted Policy, to account for enhancements in the Canada Pension Plan and Quebec Pension Plan. This is forecasted to save about CAD 1.1 billion over five years.
Estonia has discontinued the allowance for pensioners living alone for recipients of 24-hour-care services.
Lithuania has abolished auto-enrolment in its defined contribution pension scheme, moving back to voluntary coverage. It is also possible to withdraw own contributions (exempt from personal income tax). The portions of these funded by subsidies will flow to the social insurance fund and be converted into supplementary pension points in the points-based pension scheme, boosting the state pension base without additional cash outlay. This will generate savings in the future, when the government will no longer need to pay contributions to those who have withdrawn from the second pension pillar.
References
[7] Bundeskanzleramt (2025), Erwerbstätigkeit im Alter und Überstunden werden entlastet [Employment in old age and overtime are relieved of burdens.], https://services.bundeskanzleramt.gv.at/newsletter/bka-medien-newsletter/innenpolitik/bka-medieninformation-17-12-2025.html.
[13] European Pensions (2025), Surplus mechanism to be introduced to Swedish income pension system, https://www.europeanpensions.net/ep/Surplus-mechanism-to-be-introduced-to-Swedish-income-pension-system.php.
[11] FIAP (2023), Automatic Adjustment Mechanisms: Role and Lessons for the Pension Systems, https://www.fiapinternacional.org/wp-content/uploads/2023/03/NDP70-nota-de-pensiones-eng.pdf.
[14] Finnish Centre for Pensions (2025), 2025 Pension Reform, https://www.tyoelake.fi/en/themes/2025-pension-reform/.
[4] Nakazawa, N. (2022), “The Effects of Increasing the Eligibility Age for Public Pension on Individual Labor Supply”, Journal of Human Resources, Vol. 60/1, pp. 102-128, https://doi.org/10.3368/jhr.0421-11627r1.
[5] OECD (2025), OECD Employment Outlook 2025: Can We Get Through the Demographic Crunch?, OECD Publishing, Paris, https://doi.org/10.1787/194a947b-en.
[1] OECD (2025), Pensions at a Glance 2025: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/e40274c1-en.
[10] OECD (2021), Pensions at a Glance 2021: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ca401ebd-en.
[6] OECD (2019), Working Better with Age, Ageing and Employment Policies, OECD Publishing, Paris, https://doi.org/10.1787/c4d4f66a-en.
[9] OECD (forthcoming), Efficient Fiscal Policy to Ensure Income For the elderly.
[3] Rabate, S., E. Jongen and T. Atav (2024), “Increasing the Retirement Age: Policy Effects and Underlying Mechanisms”, American Economic journal: Economic Policy, Vol. 16/1, pp. 259-91, https://doi.org/10.1257/pol.20210101.
[2] Redonda, A. and A. von Haldewang C. (2025), Global Tax Expenditure Database (GTED) 1.3.1), https://doi.org/105281/zenodo.13332268.
[12] Riksdag (2010), Socialförsäkringsbalk (2010:110) [Social Insurance Code (2010:110)], https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/.
[8] Vienna.at (2025), New Savings Package: An Overview of All Measures, https://www.vienna.at/new-savings-package-an-overview-of-all-measures/9399758.