EUROPEAN COMMISSION

Brussels, 20.11.2025

SWD(2025) 367 final

COMMISSION STAFF WORKING DOCUMENT

Accompanying the document

COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE EUROPEAN COUNCIL, THE COUNCIL, THE EUROPEAN CENTRAL BANK, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS

Enhancing the capacity of the EU supplementary pension sector to improve retirement income and supply long-term capital to the EU economy

{COM(2025) 839 final}

Table of Contents

Glossary

1.Introduction: Political and economic context

2.Problem definition

2.1 What are the problems?

2.1.1    Small contribution of private savings to retirement income    

2.1.2 A low contribution of savings channelled via capital markets to investment and productivity    

2.2 What are the problem drivers?

2.2.1 Information gap in a complex decision environment    

2.2.2 Behavioural bias towards procrastination and time inconsistency    

2.2.3 Lack of attractive products    

2.2.3.1 High fees leading to low returns    

2.2.3.3 Conservative investment by pension providers    

2.2.3.2 Many small providers serving a fragmented landscape    

2.3 Out-of-scope drivers    

2.4 How likely is the problem to persist?

3.Why should the EU act?

3.1 Legal basis

3.2 Subsidiarity: Necessity of EU action

3.3 Subsidiarity: Added value of EU action

4.Objectives: What is to be achieved?

4.1 General objectives

4.2 Specific objectives

4.2.1 With respect to increasing participation in supplementary pensions to ensure adequate income in retirement    

4.2.2 With respect to improving pension funds’ capacity to direct households’ savings into productive and innovative investment    

5.What are the available policy options?

5.1 What is the baseline from which options are assessed?

5.2 Description of the policy options

5.2.1 With respect to increasing participation in and transparency of supplementary pensions to ensure adequate income in retirement    

5.2.2 Fostering access to attractive pension products    

5.3 Options discarded at an early stage

6.What are the impacts of the policy options?

6.1 Recommendation on pension tracking, pension dashboards and auto-enrolment

6.1.1 Pension tracking    

6.1.2 Pension dashboards    

6.1.3 The possible impact of auto-enrolment    

6.2 IORP Review

6.3 PEPP Regulation

7.How do the options compare?

7.1 Pension Tracking Systems (PTS)

7.2 Pension Dashboards

7.3 Auto-Enrolment (AE)

7.4 IORP review

7.5 PEPP review

8.Preferred option

8.1 The preferred policy measures

8.2 Interactions between the different policy measures

8.3 Expected effects of the pension package

9.How will actual impacts be monitored and evaluated?

Annex 1: Stakeholder consultation

Annex 2: Who is affected and how?

1.Evidence on the impact of introducing Pension Tracking Systems

2.The impact of a pension dashboard on different stakeholders

3.Evidence on the impact of auto-enrolment

4.Quantitative implications of the initiative

5.Summary of costs and benefits

Annex 3: Analytical methods

1. A primer on the pension statistics used

2. Principal considerations on the relationship between pay-as-you-go and supplementary pensions

3. Replacement rates

4.Low and unequal access to supplementary pensions

5. EU investor base in comparison to the US

6. The fees of pension funds and products

7. Profitability and scale effects

8. Investment destinations and home bias

9. Data Gaps in Pension Dashboards

Annex 4: Country-specific information

Annex 5: Summary of National Tax Treatment PEPP

Glossary

·Introduction: Political and economic context

Across the EU, ageing populations increasingly put the sustainability of pension systems and the adequacy of retirement income at risk. Moreover, upon continuation of past trends, investment, innovation and therewith productivity growth in the EU would be insufficient to allow a declining working age population to cater for an increasing share of retirees. The EU needs to address these challenges so that people can maintain adequate living standards in the face of unfavourable demographic trends and subdued economic growth and productivity progress. The way the EU addresses these issues will have broader implications, shaping its ability to manage the unprecedented transformative challenges such as the green and digital transitions and geopolitical challenges in the decades ahead.

The pension challenge refers to the problem of how an increasingly larger share of the population can maintain adequate living standards after retirement. Principle 15 (“Old age income and pensions”) of the European Pillar of Social Rights, states that “workers and the self-employed in retirement have the right to a pension commensurate to their contributions and ensuring an adequate income. Women and men shall have equal opportunities to acquire pension rights” ( 1 ). Although many Member States have carried out pension reforms, the working-age population will be under increasing pressure from the ‘double burden’ to finance the growing number of pensions directly through higher contributions or indirectly through higher tax transfers to the statutory pension system, and, in addition, to save for their own retirement. Reallocating part of their savings from low-yielding short-term financial assets into supplementary pensions (occupational and personal) and accomplishing higher returns, which are inherent to long-term savings, would help individuals achieve a higher retirement income (see Box 1 for the definition and delineation of supplementary pensions).

The Draghi and Letta reports ( 2 ) started by pointing to the changing demographic context, stressing that in the face of a shrinking workforce, the EU needs much stronger productivity and investment growth to maintain current living standards. In that vein, the Draghi report states: “The EU must also better channel households’ savings to productive investments. The easiest and most efficient way to do so is via long-term saving products (pensions). To increase the flow of funds into capital markets, the EU should encourage retail investors through the offer of second pillar pension schemes, replicating the successful examples of some EU Member States” ( 3 ). This important role attributed to the supplementary pension sector is motivated by its key role in helping convert long-term private savings into productive investments and channelling the returns generated from productive investment back into households’ retirement savings.

This approach is largely in line with that taken by the European Council. In its April 2024 conclusions , the European Council called for advancing the work on developing pensions and long-term savings products. It further echoed the importance of greater private capital participation to support the digital and the green transitions, and to stimulate European businesses growth potential, and called on the Commission to advance work on greater retail participation in capital markets, including via pension products, in its 20 March 2025 conclusions .

In 2019, a high-level group of experts on pensions, with the participation of social partner representatives, provided policy advice on how supplementary pensions could contribute to the adequacy of old age incomes and how the supplementary pensions market could be developed in the EU ( 4 ). In 2025, the European Court of Auditors concluded that the EU was “not effective in strengthening occupational pensions and establishing a pan-European personal pension product”, justifying its findings by the relevance of these pension schemes in contributing to adequate retirement income for EU citizens in the future ( 5 ).

The design of the overall pension system is under the competence of Member States. They determine the role of public pensions, which in turn define in practice how much scope there is for supplementary pensions. This does not allow for one-size-fits-all policy proposals at EU level. Many Member States have introduced reforms of their pension systems or are planning to do so. However, this entails in several cases lower public pension replacement rates ( 6 ), a shift from defined benefit to defined contribution schemes, and/or an increase of the pensionable age. In this context, the role of occupational and personal pension schemes in supplementing public ones becomes increasingly relevant to ensure the adequacy of the overall system for future incomes of retirees.

While supplementary pensions largely fall under national law, EU frameworks such as the Institutions for Occupational Retirement Provision (IORP) II Directive ( 7 ), the Pan-European Personal Pension Product (PEPP) Regulation ( 8 ) and the Directives on the Safeguarding of Supplementary Pension Rights ( 9 ) and on Acquisition and Preservation of Supplementary Pension Rights ( 10 ) establish important common standards across the single market and facilitate the free movement of workers. Occupational and personal pension offer has nevertheless remained predominantly national. In addition to PEPP, there are about 200 types of national pension products and schemes in the EU. Institutions providing personal pensions are not covered by IORP II. Many of them are insurance corporates that fall under Solvency II ( 11 ), others under national rules ( 12 ). The table below shows a breakdown of available pension products and schemes by applicable EU law.

Table 1: Number of pension products by applicable EU law/regulation

Others are UCITS, AIFMD, CRD, i.e. rules that apply to investment funds and banks.

Source: FISMA calculations with EIOPA’s pension product database.

The new EU initiatives on supplementary pensions aim, above all, to encourage and to support the efforts of Member States for addressing the pension challenge. Policies on supplementary pensions at Union level are complementary to Member States’ measures and respect Member States’ competence to decide on the organisation of their multi-pillar pension systems, including the role of fiscal incentives, the autonomy and prerogatives of social partners.

Jointly with the Member States, the Commission has monitored the adequacy and sustainability of pensions for more than a decade through the ageing and the pension adequacy reports ( 13 ). Legislative action at EU level has been limited to the few selected cases presented above. In her political guidelines 2024-2029 , President von der Leyen announced that the Commission would propose a European Savings and Investments Union. She has also tasked in her mission letter to Commissioner Albuquerque, with working on the potential of private and occupational pensions to help EU citizens with their retirement and channel their savings into the economy. When the Commission set out its strategy on the Savings and Investments Union (SIU) in the 19 March 2025 Communication , it noted the key role supplementary pensions could play in converting private savings into productive investments and channelling the returns generated from productive investments back into households’ retirement savings. The Commission announced that it would prepare both legislative and non-legislative measures to further develop the supplementary pension sector with a view to developing capital markets and facilitating pension fund investment into equity and alternative asset classes by pension funds.

2.    Problem definition

2.1 What are the problems?

The potential of supplementary pensions to address the pension challenge has so far remained unfulfilled. While supplementary pensions can complement public pensions in supporting future retirement income and in intermediating finance into productive investment, this pension sector is currently too small, and the offered products earn mostly too low returns due to predominately conservative investment strategies.

§Small contribution of private savings to retirement income

Demographic trends are one of the key causes of the pension challenge ( 14 ). The projected fall in the ratio of working-age to pension-age individuals is from 2.7 workers per pensioner in 2024 in the EU to 2.2 by 2035 and 1.7 by 2070. This puts a growing strain on economic growth and social welfare policies. If left unaddressed, it will challenge the sustainability of public pension systems and the adequacy of the income that they provide to retired persons. Measures to stabilise the public pension system through higher contributions to public pensions, higher retirement age to boost the number of contributors to the pension system or labour migration are out of the scope of this policy initiative. For the comparison of statutory with supplementary pensions, it is however important to note that increasing contributions to the public pension system levied on the younger generation bears the risk of disincentivising their labour supply if these contributions are considered a tax and not as payment to an actuarial fair insurance scheme. In contrast, contributions to supplementary pensions are more likely to incentivise young workers to contribute, as they see direct and personalised benefits from their participation (see Annex 2.4).

The latest Ageing Report shows that public pension benefits will become less generous in many Member States. It projects that the public pension benefit ratio (average pensions to average wage) will decline from 43.8% to 35.6% and the gross replacement rate (pensions relative to pre-retirement wage) will decline from 43.4% to 38.2% in the EU by 2070 (Figure 2). Future replacement and benefit ratios are projected to be lower in most Member States, though the national situations vary widely ( 15 ).

Note: The benefit ratio describes the average public pension to average wage

Source: Ageing Report 2024.

The projected fall in replacement rates will affect different parts of the population differently. The risk is that the demographic pressure on pay-as-you-go (PAYG) pensions will reinforce today’s challenges to balance the objectives of providing actuarially fair payments to all contributors and delivering adequate pensions to those with special needs, for example those with low wage income or career interruptions due to caregiving activities. Almost 20% of people above 60 years old are at the risk of poverty. The gender pension gap stands at 25%, indicating that women’s average retirement income is 75% compared to that of men. The Commission’s Pension Adequacy Report demonstrates trends and developments at granular level and different dimensions of adequacy across Member States.

According to the 2024 Pension Adequacy Report 16 , in the case of a standard, 40-year career, overall pension replacement rates are set to fall over the next four decades in most countries. Compared with the 2022 values, the 2062 values are lower in 21 countries, while being higher in only seven countries. One way of looking at this is to say that a 40-year career is less likely to lead to a higher pension for standard career in 2062 than it did in 2022.

Figure 3: Net theoretical replacement rates (TRR 17 ), base case (40 years to statutory pension age), average earner, 2022 and 2062, men, pp, EU-27 Member States and Norway

Citizens are increasingly aware of the pension challenge, and consequently sceptical about the adequacy of their retirement income ( 18 ). In a 2024 Eurobarometer, 28% of respondents replied that they were 'not too confident' that they will have enough money to live comfortably throughout their retirement years and a further 25% feel 'not at all confident' ( 19 ).

Access to supplementary pensions improves confidence. The share of respondents being not confident about their retirement income is at 58% among those that do not have supplementary pensions, compared to 43% that are enrolled in occupational pensions and 33% that have both occupational and personal pension products. Among those that own pension products, 40% of those with occupational pensions said that these are important or very important part of their retirement income and another 43% that these are rather important. For personal pensions, the corresponding numbers are 34% and 50%” ( 20 ).

However, only a small share of the population uses supplementary pensions. Participation in supplementary pensions is still at a low level in the EU, notwithstanding that it has increased over the last two decades ( 21 ). The coverage varies across Member States and data sources (for variation across Member States, see Table 44 in Annex 3.4).

–Eurostat reports data from the Labour Force Survey, according to which 83% of pensioners (81% of males, 85% of females) had no other pension than a statutory pension in 2023 (see Figure 4). A further 15% of pensioners received a statutory pension in combination with an occupational and/or personal pension. 2.1% received an occupational or personal pension and no statutory pension ( 22 ).

–In EIOPA’s 2024 Eurobarometer, 20% of surveyed EU citizens report participating in an occupational pension and just 18% own a personal pension ( 23 ). Ownership of pension products was smaller among retirees at 17% respectively 11%. The larger number for the population at large than for retirees suggests that supplementary pensions have become more important over time.

–When the participation rates are measured as the number of members of pension funds relative to population, for which data is available in EIOPA and the ECB, the rate is much higher for some Member States (Table 42 in Annex 3.4). The higher number is likely due to a sizeable number of people being member in more than one pension fund, for example in two different occupational pension funds after a change of job.

Figure 4: People receiving an old-age pension by type of pension in % of pension receivers, EU 2023

Source: Eurostat (labour force survey).

Quantifying the actual contribution of supplementary pensions to retirement income is possible for only a few Member States. Table 43 in Annex 3.4 shows data on the amounts saved in supplementary pensions for all Member States from different statistical sources. 11 EU Member States provide a breakdown of their old-age benefit ratio between public and supplementary pensions (Figure 5) for the Ageing Report, yielding a significant contribution of the latter to retirement income for only Denmark, Netherlands and Sweden. Using a different approach, namely by calculating how much the theoretical replacement would increase if supplementary pensions were added, OECD data allows for a more comprehensive coverage of Member States. Mandatory public pensions deliver a replacement rate of 49.5% for a person that earns the average income in the EU-27. Adding the contribution from supplementary pensions (both mandatory and voluntary) to the public replacement rate increases the gross replacement rate by 8%-pts to 57.8% in the EU ( 24 ). While there are notable differences across Member States, these figures highlight an uptake of supplementary pensions which is insufficient to effectively reduce pension gaps.

The somewhat higher share of pension products measured in monetary terms than participation rates is due to an overproportionate use of supplementary pensions by households with higher income. The 2019 High level group of pension experts on supplementary pensions observed that women, lower income groups, and people in non-standard forms of work have less access to supplementary pensions. Lack of inclusiveness is therefore also a problem along the following dimensions for which Annex 4 gives additional information.

While supplementary pensions are underused by households in many Member States, households have ample leeway to save more in pension products. EU households save a lot of their income on average ( 26 ), but not in a form that fully supports their retirement. The sum of household financial wealth ( 27 ) is almost at par with their statutory pension claims (see Figure 6). Overall, however, households save mostly in bank deposits. Official statistics show that 30.6% of the EU households’ financial wealth, excluding accrued statutory pension rights, was in the form of bank deposits or cash in 2024. They save a slightly higher share of their financial wealth in long-term financial products that offer higher returns. Claims against occupational pension funds accounted for 13.5% in 2024, against life insurers and personal pension funds for 13.1%, their investment fund shares for 10.4%.

Thus, while many households may not afford to save more, there is scope for most of them to save in better ways than in bank deposits. Bank deposits have offered a low return over the last few years, below the rate of inflation ( 28 ). While housing is an attractive long-term investment that can be used for accommodation when old, it offers no means for risk diversification and is difficult to liquidate. (See Section 2.2.3 for the return of pension products).

2.1.2 A low contribution of savings channelled via capital markets to investment and productivity

Next to demographic trends, a second source of the pension challenge is the lack of productivity increase. Productivity growth needs to play a larger role to ensure a level of economic growth that is capable of supporting a larger share of retirees when the labour force is shrinking. Over the last decade, employment growth contributed to about two-thirds of GDP growth, and productivity growth around one-third. The low contribution of productivity growth will not be sufficient to keep up EU living standards in face of low population growth and a declining working-age population. In order to accomplish higher productivity growth which is able to offset the deteriorating contribution from employment, productive investment would need to increase stronger than observed over the last decade.

Table 2: EU-27 growth rates in %, average 2015-2024 in constant prices

Note: GDP in chain-linked volumes 2015, domestic employment, productivity in persons employed, investment is gross fixed capital formation in chain-linked volumes 2015, national population.

Source FISMA calculations with Eurostat data.

Empirical work by the IMF notes that an absence of business dynamics is an important factor behind Europe’s weak productivity performance ( 29 ). It found this absence to be visible in both large leading firms innovating less and younger firms being less disruptive than their peers in the US. Lack of market size, human capital and equity financing were identified as the main determinants.

Lack of scale and competitiveness of EU capital markets (relative to other important players, such as the US and the UK) goes in tandem with an underdeveloped mobilisation of household savings and their channelling to capital markets. A further SIU initiative aims to support market consolidation and supervisory integration to foster more integration and scale on EU capital markets. Better mobilisation of household savings for productive uses via long-term investment would be instrumental in enabling the future smaller labour force to generate products and services that allow a decent living for the future larger class of retirees. In this vein, the Draghi report points to the underdevelopment of the EU pension funds sector as one of the main reasons of the under-supply of long-term capital in the EU relative to other major economies. It therefore designs it as indirect cause of the productivity slowdown of the EU economy and its eroding competitiveness. This necessitates efforts to strengthen the efficiency of EU capital markets, and foster the role of pension funds in the allocation of capital in line with the SIU strategy.

Pension funds, alongside investment firms and insurance companies, represent a vast pool of long-term capital and thus are a significant force in financing the real economy. Pension funds, in particular, have the potential to act as key enabler of transformative change towards green, digital and innovative growth needed to boost productivity and at the same time be instrumental in addressing the pension challenge. Their capacity to gather households’ savings in large patient capital pools that can be allocated in well diversified ways to long-term oriented investment projects, including via equity funding to innovative entrepreneurs, makes them one of the main actors for advancing the SIU agenda. Through their function as anchor investors in domestic markets and by contributing to the development of EU capital markets, a stronger role for supplementary pensions could provide important private sources of financing for the green and digital transition (that is likely to rely more on equity financing than loans), as well as productive investments that strengthen the EU’s competitiveness ( 30 ).

Confirming the important role of pension funds and capital market development, the academic literature documents a robust correlation between the size of pension savings and the development of capital markets. Some researchers are suggesting a causal effect from the development of a pension sector to capital market ( 31 ). EU Member States with well-developed supplementary pension vehicles (specifically DK, NL and SE ) are also home to large institutional investors and have well-developed capital markets ( 32 ).

One of the most striking differences between the EU and US investor base is that the relatively small market share of pension funds as investors in private equity and the stock markets in the EU. EU pension funds are about 15% of the size of US pension funds in total assets, and their holdings of listed shares amount to 7% of what US pension funds are holding (Annex 4). The differences in the role of pension funds as investors between the EU and the US might therefore be a central structural element determining the different paths in capital market activity and therefore an indirect reason behind differences in competitiveness and economic adjustment.

In fact, a number of empirical studies underpin the notion that supplementary pensions are an important driver of financial development ( 33 ). Those studies show a capital market deepening following pension reform. They also show an increase in US venture capital and R&D investment after pension funds were permitted to invest there.

A recent analysis with Danish data found that pension funds’ equity investment in firms improves their productivity by a substantial amount, estimated at 3-5% (see table below) ( 34 ). Complementary research by the IMF found evidence that supplementary pension vehicles make it easier for firms to raise long-term capital via their investment in stock markets. It emphasised that pension funds’ investment broadens the access to capital market instruments for households, which means that more households could benefit from higher returns on capital market investments ( 35 ).

Table 3: Empirical results about the productivity effect of pension investment on the productivity in Danish firms

Source: International Centre for Pension Management (ICPM) ( 36 ).

2.2 What are the problem drivers?

There are many reasons why households do not make better use of their savings for retirement and why pension funds contribute so little to the capital allocation in the EU economy. However, the importance of the different reasons is likely to be different across Member States. In any case, it is undisputed that long-term saving decisions are complex for individuals. They need to overcome the information gap between what they know and want to know before they take a decision (driver 1). Behavioural economists point to delaying decisions as a wide-spread human strategy to deal with complex decision making (driver 2). Households have also objective reasons to stay away from pension products if they do not find them attractive enough (driver 3). Reasons for an unattractive product offer are linked to high fees and costs charged by financial intermediaries and low returns generated for savers (driver 4). Low returns are a consequence of risk-averse investment policies by pension funds (driver 5), which interact with their small scale (driver 6) and lead to both a low share of pension savings allocated in riskier asset classes and the generation of low returns for savers. With little demand for supplementary pension products and both demand and regulations fragmented across Member States, many pension funds cannot reach the scale economies necessary to offer attractive pension schemes and thus to access investment opportunities that fund innovation and productivity.

Figure 7: Problem tree

2.2.1 Information gap in a complex decision environment

Information problems can weigh severely on the individual choice of a pension product, considering that such long-term investments amount to a long-term commitment that requires a good understanding of the personal needs and available options. Asked about what features of insurance and pension products they consider most important, 59% of respondents to the 2024 Eurobarometer survey quoted an extensive or accurate coverage at first place. This was followed by low costs (37%) and good returns (34%). These preferences translate into a number of parameters a decision on a pension product needs to take into account ( 37 ). For an illustration of such factors, Mercer/CFA (2024) lists the following parameters as relevant for investments into defined contribution pension plans:

1.Which pension provider to choose;

2.The amount of pension contributions on top of contributions to mandatory pensions and the frequency of payments;

3.Their risk attitude and, dependent on that, the type of investment strategy;

4.The planned age of retirement and/or statutory retirement age;

5.Whether retirement savings are paid out in a lump sum, or via annuities until a fixed date or until death; and

6.The magnitude of pay-outs and when they start.

Information is an important issue. In the 2024 Eurobarometer survey, a quarter of the respondents said that it is difficult to find information about pension products and a third that information is difficult to understand. 16% replied that the main reason why they do not own a private personal pension is that it would be difficult to find a good pension product ( 38 ). Responses of those individuals that have no personal pension product suggest that lack of information about long-term saving opportunities is a driving factor ( 39 ). 10% of respondents said that they are not aware of any such product with sufficient protection being available and another 10% are not aware that such products are available at all. One in five persons indicated not having thought of pension products yet, which suggests that awareness or procrastination may also matter (see next section). Among those that do not have a personal pension product, 28% said they cannot afford it. While the share of people not being able to afford a pension product is significant, it is much lower than the 47% share of population saying they cannot afford buying investment products.

The interaction between education, income and affordability is also evident in the Eurobarometer responses. The share of households indicating having neither pension products nor comparable investment products is negatively correlated with education levels: 18% for highly educated versus 34% for those with lower education levels; it is also much higher for females than for males (31% versus 23%), consistent with lower levels of financial literacy among women than men ( 40 ). The Eurobarometer replies are consistent with results of academic studies that analyse why personal pension products are little used in Germany: low educational attainment, uncertainty and judgement errors about future retirement levels, and an underestimation of life expectancy ( 41 ). A Swedish study points to education attainment, complexity of pension decisions and procrastination behaviour to be major determinants ( 42 ). A number of studies found that trust is an important determinant of participation in pension schemes and a recent Dutch paper was able to establish that there is a causal link from trust to participation ( 43 ).

Figure 8: Responses to the question about the main reason why people do not own a private personal pension

There is a large variety of EU and national requirements for pension fund providers to inform their members on the performance of pensions funds (e.g. the IORP Pension Benefit Statement), but these provisions are inconsistent and not harmonised across different categories of pensions products (e.g. occupational versus personal) and regimes (EU versus national), or even within the same pillar (IORP occupational pensions are pension fund regulated under the covering Institutions for Occupational Retirement Provisions II Directive, here “IORPs” versus Solvency II occupational pensions, insurance backed pensions, regulated under the Solvency II Directive), and they differ across Member States.

Furthermore, even for those individuals who are saving in supplementary pension products, a lack of transparency and an incomplete overview of their entitlements can hinder career and retirement planning and, by implication, pension adequacy. In modern social protection systems, individuals' awareness of their pension rights is increasingly important to enable them to make informed decisions about their career and savings needs and thus to secure adequate pensions. With growing professional and geographical mobility of workers ( 44 ), an increasing complexity of pension systems and shifting of responsibility for ensuring adequate pensions towards individuals ( 45 ), workers face the challenge of following pension entitlements in different schemes to which they were affiliated during their career. For their part, pension providers need to provide information to their members and former members as they move job and change address over very long periods. Countries that monitor unallocated forgotten pension claims found that the amounts can be in the millions of euro ( 46 ).

2.2.2 Behavioural bias towards procrastination and time inconsistency

Analysis by the OECD shows that procrastination on the decision to save for retirement has a large impact on the amount savers can earn. It suggests that people “delay saving into their 40s or 50s”. Since compounded return accumulates over time, an earlier start yields higher amounts by the time the retirement age is reached. In particular, equity investments tend to generate much higher returns the longer the periods are ( 47 ). Short contribution periods reduce the benefits savers can draw from equity investments and increases their exposure to stock price volatility.

Behavioural economists refer to the delay of taking a decision on pension savings as the “procrastination effect” ( 48 ). This effect is independent from the effect of a lack of information. Consistent with this distinction, research with German data found that more education will not undo it ( 49 ). Briere (2025) notes in this context, that 46 % of financial advisors in the US, i.e. those with good knowledge of financial issues, do not have a retirement plan.

Procrastination reflects time-inconsistent behaviour of savers, reflected for example in the discrepancy between widespread concerns about future retirement income in the Eurobarometer surveys quoted above, which reflects that people are aware that they would need to act, and a small share of respondents acting and either enrolling in pension saving schemes or acquiring long-term saving products. In an often-quoted US survey from 2001, most respondents said that it was important to save for retirement. However, only 3% stated they were actually doing so; 24% replied they planned to start in the “near future” and 68% that they were saving too little ( 50 ). A survey by the Irish statistical office revealed that for those that had no pension coverage, 45.2% indicated they “had never got around to organising it”; compared to 40% arguing they could not afford it ( 51 ). Almost half of those without pension coverage are expected to rely on public welfare pensions.

Most empirical analysis about procrastination has been with US data. Nevertheless, a small number of studies find a similar effect in EU data. For example, Elinder et al. find evidence of procrastination on pension decisions in Swedish data, Krijnen et al. (2016) in Dutch data and Piotrowska (2019) in Polish data ( 52 ). A telling analysis about the large share of retail investors not reacting to clear advantages was done by Chetty et al. (2013) ( 53 ). They identified that only 15% of the addressees reacted to a tax incentive, leaving the remaining 85% passive.

Despite passiveness being adversely correlated with financial literacy, education and information provision is unlikely to fully undo the effect of procrastination as behavioural bias. Behavioural economists find it important to distinguish between time-inconsistent behaviour as determined by preferences on the one hand, and information problems determined by lack of financial education or understanding on the other hand. The distinction is useful because suitable policy measures to address them differ. Rules to commit undertaking specific actions are suitable to address the former problem, while financial education and information provision is for the latter ( 54 ).

2.2.3 Lack of attractive products

Pension products are unattractive if they do not correspond to the needs of those that are supposed to buy them and especially if they are costly (section 2.2.3.1) or yield little returns (section 2.2.3.2). A lack of attractive pension products is evident from the low take up of personal pension products in many Member States. Less than 20% of households reported to hold a personal pension product in the EIOPA Eurobarometer quoted above. Examples from two Member States highlight the issue that low net returns discourage participation.

(2)Germany subsidised contributions to personal pension products (‘Riester-Rente’) in the early 2000s, but has since observed that households have stopped or reduced their contributions in reaction to low net returns on their investment ( 55 ). While the Riester-Rente allowed and subsidised investments via savings plans, fund savings plans and insurance plans, the largest share of the discontinued contracts were insurance based, with both expensive capital guarantees and an expensive distribution system as cost drivers. This resulted in severely subdued and sometimes even negative net returns despite public subsidies ( 56 ). The government in July 2025 announced plans to reform this system ( 57 ).

(3)In 2002 Estonia introduced a second pillar pension scheme in which younger workers were automatically enrolled with an option to opt out, with savings managed by Estonian pension funds. After years of criticism over low long-term real returns and high fees from these pension funds, Estonia passed a reform that took effect in January 2021, making participation voluntary and allowing withdrawals. In the first wave following the reform, some 19% of participants exited from the scheme, taking €1.3 billion with them. As of 2024 this figure has risen to 30% of participants ( 58 ).

Since most product features can be adjusted to the features demanded, low net returns are the main determinant of unattractiveness. Better Finance researchers identified four reasons for low returns: (i) inflation, (ii) tax treatment, (iii) the fees financial intermediaries charge, and (iv) the asset allocation of pension funds ( 59 ). A further critical determinant of the returns of pension products in their riskiness. Pension products with guarantees have a significantly lower return than products without guarantees. In an analysis with Danish data, Sondergaard (2025) found a return difference of about 200 basis points between guaranteed and non-guaranteed occupational pension schemes 60 . While surveys consistently reveal that savers have a strong preference for safety, the costs of this safety is not transparent to them.

The energy price shock of 2022/23 had a sizeable impact on average inflation figures, leading to low real returns of many investment products even over a 10-year average. Tax treatment is analysed in Annex 2. High fees (driver 4) and low returns can be due to a lack of competition and transparency. They can also be the result of the costs of product features such as provision of advice, guarantees or flexibility clauses attached to a pension product. They are also determined by two further drivers that have a direct effect on the contribution of pension savings to the financing of the economy. The lack of scale of pension providers (driver 5) augments costs that lead to higher fees. Conservative investment policies of pension providers (driver 6) show up in their asset allocation and yield low returns that can be distributed to members. Lack of scale and conservatism of investment policies are interrelated. Large funds have both better opportunities to exploit scale economies on the cost side and better access to investment opportunities that promise higher returns.

2.2.3.1 High fees leading to low returns

Even well-informed savers, conscious of the need to avoid procrastination, may choose not to save in retirement products. The 2024 Eurobarometer survey referred to earlier provides evidence that a lack of attractive products could discourage a significant proportion of savers. When asked why they did not buy a personal pension product, almost 20% replied that this is because of high costs and fees, and 10% because of poor performance and limited coverage. While these percentages do not look high, they need to be seen in in conjunction with 62% unallocated responses ( 61 ). That means that half of those that gave a response to this question considered costs and fees as too high.

Analysis using different methodologies and data point to overall low returns of pension products for savers in the past decade.

·Frequently quoted research by Better Finance showed a median annual return of 16 occupational pension products corrected for inflation of 0.9% over the 10-year period 2014-2023, that of 22 personal pension products was negative at minus 0.1% (category 1 and 2, respectively, in Figure 9.

·Calculations of an implied return of households’ occupational pension claims with Eurostat data yielded an average return of 2% across the EU Member States over the same period. While this is still lower than inflation over this period, it was about 1 percentage point higher than households’ return on their holdings of bank deposits and bonds, implying a 100 basis points in foregone revenues to hold a liquid asset.

·EIOPA’s cost and performance report paints a marginally better picture with nominal returns of IORPs ranging between 1.3% and 6.1% in nominal terms averaging 2017-2021. This translates into inflation adjusted returns of -0. 4 to 4.2%.

Low historical returns mean that citizens do not experience the long-term value of contributing to such schemes, which ultimately weakens the overall uptake of pension products.

Figure 9: Real returns on pension products, 2014-2023

Source: Implied returns on pension products are FISMA calculations replicating a Eurostat method ( 62 )and adjusted for HICP inflation. Better Finance numbers are from the 2024 publication. The breakdown between category 1 and 2 differs across Member States and is for BE: ORPs vs assurance group and pension saving plans, DE : #N/A vs life insurance, EL : P2 vs P3, FR: corporate DC vs life insurance CG, CY: mandatory vs voluntary, LV: contractual vs open pension funds, LU: mandatory vs voluntary, HU: P2 vs P3, AT: occ vs life insurance, PL: employee vs voluntary pension funds, RO: mandatory vs voluntary, SK: pension funds vs supplementary pension funds ex PEPP, SE: AP7 Såfa vs other premium pension funds.

The data from Better Finance reveals substantial differences between Member States. Data collections from Better Finance, EIOPA and OECD suggest that fees for pension products amount to about 1% of assets (Figures 10 and 11). By contrast, Swedish or Dutch pension funds charge less than 0.05% ( 63 ). As any fee generated by the provider is paid for by the ultimate savers, it reduces the net return they received by the same amount. Thus, a 1% fee deducts 100 basis points from the returns the pension managers generate though their investments. Low fees are therewith key determinants of the income from pension savings in the pay-out phase.

Pension funds’ costs are lower than those of actively managed investment funds. This is evidenced by the comparison of the 1% quoted above with ESMA’s data on investment funds. ESMA identified annual costs for equity and mixed UCITS at 2%, bond UCITS at 1.5%; though costs for ETFs are considerably lower at 0.5% ( 64 ).

Costs are generally lower for occupational pension funds than for personal pension products due to largely mandatory participation, scale and significantly lower distribution costs. Comparisons of personal pension products suggest that annually accumulating distribution costs consume a significant part of the return on capital investments and hence depress the net return. Employers and social partners as sponsors of occupational pension funds are equipped with more bargaining power vis-à-vis financial intermediaries than individual savers, which may be another reason for the lower fees for occupational pension funds. This bargaining power is particularly strong for the very large pension funds, who can decide to manage certain strategies in-house or to outsource them  65 . Annex 3 gives more details, including on the breakdown between returns and fees of occupational and personal pensions and other observations for some Member States. The statistical annex to EIOPA’s technical input shows a breakdown across cost categories and cost differences between types and size of IORPs.

Country-specific empirical research for Austria and Germany has documented that administrative costs and onboarding fees for personal pension products are high in these Member States. Cost structures are complex, and portability is costly. Moreover, restrictive contract designs and guarantees are costly for providers ( 66 ). Since replies from Austrian or German respondents to the Eurobarometer surveys are not different from those from other Member States, it is likely that if such research is undertaken in other Member States, it would find comparable results ( 67 ).

Aside from scale effects, one reason for the magnitude of fees charged by pension funds and other financial intermediaries are the costs of advice to households. Most financial institutions signalled during the consultations that the combination of mandatory advice and a cap on fees was an important reason why they did not offer PEPP. When comparing with other financial products, it is notable that many financial institutions increased their offer of execution-only-products to retail customers or offered robo-advice, which can be interpreted as adjustment strategy to reduce on-boarding costs. The higher costs of advice when the product is offered to individual households compared to occupational pensions, which do not require individual onboarding, is an explanation why fees are lower for occupational pensions than for personal pension products.

Cost structures differ between pension funds that offer defined benefits (DB) and those that offer (DC) schemes ( 68 ). Costs for investment strategies are higher for the former, since their portfolios are larger and more complex, and the investment strategies often make use of active management and derivatives. Administrative costs are higher for DC schemes because member accounts and individual choices need to be managed. A further cost driver are guarantees that cater for households’ perceived primary interest in safety ( 69 ). Occupational pensions limit guarantees to the necessary minimum as they drive up prices.

High costs and fees may also be a reason why many employers do not offer occupational pension schemes to their employees, unless a member state has a strong tradition of occupational pension schemes. Employers face costs from directly contributing to occupational pensions, which is part of their labour costs. Other costs relate to the set-up of the scheme, administration and regulatory compliance costs, and fees paid to providers. A study comparing pension schemes in eight EU Member States quantified that employers finance 50% of both public and occupational pensions, with large variation across Member States. Employers contributed the most in Sweden and Finland, whereas employees contributed most in the Netherlands ( 70 ). Employers’ contribution to occupational pensions reached between 3 and 4% of GDP in 4 Member States (DK, FR, NL, SE) and less in DE, FI and IT (see Figure 12). The analysis did not try to estimate to what extent the employers’ pension contributions are offset by lower wages paid out to employees, nor the extent they can be deducted from corporates’ tax bill. See Annex 4 for more details on tax treatment. Information about other costs to employers is known from the UK pension regulator, suggesting low amounts even for micro firms ( 71 ). Fees to pension providers are included in the fees paid on occupational pensions presented above, i.e. between 0.2 and 1.5% according to EIOPA data.

2.2.3.2 Conservative investment by pension providers

Asset allocation and returns

EU pension funds’ low returns over the past decade reflect their conservative investment strategies. OECD data points to average of nominal returns earned by pension funds of 4.4% over the ten years 2014-2023 which corresponds to 1.4% in inflation-adjusted terms in the EU (weighted average of the EU Member States, weight with pension assets). The figures below show that pension funds’ returns are volatile, especially when compared to households’ income from pension claims reported in section 2.2.3. Under the spell of the energy price shocks of 2022, pension funds in almost all EU Member States recorded strongly negative yields in 2022, many in the double digits. They also suffered substantial losses in 2008 during the financial crisis. The similarity of the time profile of returns across Member States shown in Figures 13 and 14 suggests that common cyclical factors are a more important determinant of returns than country-specific conditions.

The European Commission’s 1997 Green Paper on supplementary pensions ( 72 ) already concluded that any comparisons of the rate of returns across asset classes are sensitive to the period chosen. The sizeable impact of the losses recorded during the 2008 financial crisis and the 2022 energy price shock on 10-year returns underpin the notion of the time horizon to be the most critical determinant of the return on retirement savings. The ideal benchmark for pension funds would take account of pension funds building up asset positions for their members over an average work life and remaining life expectancy. Statistics suggest the average work life is currently 37 years in the EU and the remaining life expectancy of people at age 65 to 67 is 18-20 years ( 73 ). Unfortunately, the longest available time series of pension funds’ investment returns goes back only to 2002 for some EU Member States 74 .

Pension funds invest in three broad asset classes: equity, fixed income and other. The analysis of the actual breakdown is made complicated by pension funds’ use of investment funds for a large part of their asset allocations. They are not unique in this, also insurance companies and many investment funds do so. The OECD and EIOPA have different means to look through the asset allocation via investment funds. The former asks national authorities to provide a look-through of their pension funds’ data and many, but not all do. EIOPA provides a detailed breakdown by types of investment funds, albeit limited to IORPs ( 75 ).

Benefits and returns of equity investment

The following analysis focuses on the share of equity investment, taking equity investment as a proxy for “risky investment”. In reality, equity investment includes listed shares, unlisted shares, private equity and infrastructure, with the latter two perceived as riskier than listed shares. Recent research found that an increase in the expected risk-adjusted returns of private equity relative to listed equity led to higher investment of US pension funds into this and other ‘alternative’ asset classes ( 76 ). It also concluded that the perception of higher risk-adjusted returns was predominantly shaped by consultancy reports. There was limited empirical support for other factors that could explain an increase in pension managers’ risk appetite.

The geographical dimension of pension funds’ investment and their home bias is analysed in Annex 3.

OECD data suggests that EU pension funds invest a smaller share in equity and a larger share in debt instruments than in other OECD countries. The equity share is 28% for the average of EU Member States, 22% for a size-weighted EU aggregate. This compares to 30% in the UK and Switzerland, 35% in the US and Norway. Among the EU Member States, the equity share is particularly low in Germany, Czechia, Slovenia and Portugal, high in Estonia, Lithuania and Poland (see figure below).

Figure 15: Allocation of pension assets in selected investment categories in EU Member States and OECD countries, preliminary data end 2024.

Source: OECD Pension in focus, June 2025.

Detailed data from EIOPA for IORPs that allows the allocation of investments in investment funds (look-through) shows a somewhat higher equity share of about 30%. Only half of the equity share is direct investment in common equity, the other half was derived from the share invested via investment funds. 28% of IORPs assets was listed equity, 3.3% in unlisted equity, which is largely in private equity ( 77 ). Apart from equity, IORPs invest in sovereign and corporate bonds, real estate and a number of other asset classes. 

Regarding specific asset classes that are considered as requiring asset managers to engage in search, information and monitoring efforts, IORPs’ investment in private equity in 2024 was at 3.7% of their exposure, into infrastructures at 2.2%. This is however caused by the relatively large Dutch investment in these riskier asset classes, which some Dutch pension fund can enter due to their large size (4.5 and 3% respectively). Without the Netherlands, the investment share for all other EU Member States covered by EIOPA statistics broadly halves to 2.2% respectively 0.9%. That some Member States allow only investment into equity traded on regulated markets is another reason for the low numbers for these asset classes ( 78 ).

The superiority of equity investment over bonds in the very long-term is demonstrated in an academic paper that analysed the performance of asset classes from 1870s to 2015 ( 79 ). Both equity and housing had average real returns of 7% per annum, bonds of 2.5%. When the sample is limited to the post-World War II time to account for the very low real returns of bonds in war times, the general conclusion does not change. Equity yielded above 8%, bonds 2.7%. Housing appeared as profitable as equity and less volatile, the authors however note that diversification and liquidation is much more difficult to realise with house ownership.

The table below displays the actual returns on different EU asset classes taken from either benchmark indices or for the three non-listed equity classes from an industry benchmark publication ( 80 ). All numbers are illustrative for the broad performance and sample bias cannot be excluded ( 81 ). Yields on sovereign bonds and rated corporate bonds were relatively low. Investment in listed equity would have earned more than twice than investment in sovereign bonds over the last 20 years. Investment in unlisted equity would have earned a higher return than on listed equity, though with a small difference for venture capital and infrastructure funds over the 20 years period, which includes the years of the great financial crisis. The 10-years period is characterised by a low-interest-rate environment, visible in low returns on bond investments.

Table 4: Actual average returns on different EU asset classes over 10 and 20 years holding period

Sovereign bonds are the average of EU Member States with 10 years maturity, Corporate bond indices along rating categories, maturity around 5 years, * averages over 15 years as longest available time period, listed equity is annual change in % of Eurostoxx 600 performance index, GCF (Growth capital funds), VCF (Venture capital funds), PIF (private infrastructure funds) are returns on a sample of funds that were liquidated, i.e. returns realised.

Source: FISMA calculations with ECB, Eurostat, Bloomberg and Invest Europe data.

Their special position of sophisticated institutional investors with a very long-term investment horizon should make pension funds well placed to allocate a larger share of their investment into equity compared to other market participants. When analysing pension funds’ asset allocation and performance across member countries over the period since 2002, the OECD (2024) found that a higher equity share delivered higher returns. It also added that the higher returns in the long term come at the expense of short-term volatility. The OECD (2024) undertook analysis to account for the sensitivity of the returns on equity from the period chosen. Scenario analysis was performed that looked at historical returns since 1900 and derived which returns households would have obtained if a defined contribution pension plan had existed. For most age-cohorts equity performance would have outperformed bond investment. Even for the minority of age cohorts which would have accomplished a higher return with bond investment, the return difference was modest, amounting to 0.3 to 0.4 times the accumulated value of contributions. Additional model simulations yield a 90% probability that a saver with pay-outs in the form of annuity until death accomplishes a higher replacement rate through equity than through bond investment.

The lessons drawn by the OECD from the analyses concern the importance of starting as early as possible to save for retirement. Saving for 40 years instead of only for 20 years makes a significant difference and the more so, the higher the returns on equity investment because of the effect of compounded returns. See also the evidence on the interaction between the size of pension funds, their asset allocation and returns by Bikker and Meringa (2024) in Annex 3.

Determinants of conservative investment strategies

The existing governance arrangements between members, sponsors and pension managers reward risk-averse behaviour. Many industry representatives stressed their interest in generating a stable return for their members and cautioned against risk-taking in their replies to the consultation and in exchanges with Commission staff. Several managers from the pension industry repeatedly said they consider their entities following a social-protection strategy rather than a profit-maximising motivation. The incentive for sponsors of pension funds, for example, employers to ensure that the fund generates high returns for its members should also not be taken as a given. Since the benefits accrue only in the long term, they may be more interested that the fund manager produces a stable return and avoid risks.

Prudential policy has an impact on both pension funds’ investment policy and their size. The IORP investment rules are complemented by often detailed national asset allocation rules, such as asset class limitations or allocation quotas, promulgated by either legislation or supervisory guidance. Article 21 of IORP II allows Member States to introduce a variety of investment restrictions and quotas for various asset classes (see table below). Some (but not all) of these national rules could restrain the levels of portfolio diversification necessary to achieve adequate investment returns for pension scheme participants. Evidence that regulation is a determinant of equity investment stems from the comparison of a 27% average equity share in countries with quantitative investment limits and 32% on average in those countries without ( 82 ).

Table 5: Investment restrictions and quotas consistent with the investment rules and the prudent person principle in IOPR 2’ Article 21

It would appear that national prudential rules are (partly) driven by the need to cater to the reduced risk-taking capacity of often small size of European IORPs. Since smaller pension funds have weaker risk bearing and managing capacity than larger ones, prudential authorities seem to impose more stringent constraints the smaller the pension funds in their jurisdiction. This may not only take effect through legal rules but also through supervisory guidance on the interpretation of the prudent person principle enshrined in the IORP prudential framework.

2.2.3.3 Many small providers serving a fragmented landscape

The low returns that many pension funds generated over the last years could be a consequence of their small size. Scale economies are present in many functions that pension funds cover, from the administration of members’ contributions and claims to the scope of their investment policies. The figures below show that IORPs costs decline with the size of the provider. The relationship between size and the returns that pension funds generate through their investment policy is more complex and not visible in cross plots. The theoretical arguments and available evidence is discussed in this section.

Most EU pension funds are small

The landscape of pension funds is fragmented across EU Member States and within most EU Member States, leading to EU players that are much smaller on average than peers in non-EU jurisdictions. Although different structures of pension systems and in particular different roles of public pensions mean different size of supplementary pension markets and therewith size of pension funds acting on those markets, a comparison of pension funds’ actual size can be informative. Such a comparison reveals that only the largest EU supplementary pension funds equal their peers in the US, Canada or Australia. According to the numbers of the P&I global league table used for Table 6, six EU pension funds have a balance sheet of more than USD 100 billion, compared to 15 in the US. While the two largest EU pension funds have a size that makes them significant players on global markets, the third to twentieth largest pension funds in the EU have a balance sheet size that is about half of that of the third to twentieth largest US pension funds. The fiftieth largest US pension fund has more assets than the sum of all 347 occupational funds in Spain.

Table 6: Assets under management of the largest pension funds across jurisdictions, in billion USD, 2023Q3

Source : FISMA calculations with data of AuM of individual pension schemes from The P&I/Thinking Ahead Institute World 300: Largest retirement funds, download 25 June 2025.

Data from ECB and EIOPA allows to calculate the average size of pension funds, with EIOPA data being limited to IORPs as a subset of pension funds across Member States. The about 1400 IORPs have about EUR 2,720 billion assets under management, yielding an average size of less than 2 billion. 80% of the IORPs have less than EUR 1 billion, and a third less than EUR 25 million. The table below shows how average sizes vary between Member States, being the largest in SE, FR and the NL. In some Member States, the average size is less than EUR 1 billion.

Table 7: Average size of pension funds and number of IORPs, 2024

Note: B stands for billions, M for millions of total assets. Pension funds include IORPs. In several Member States, small IORPs are not covered by EIOPA statistics, which leads to an upward bias in the average size. Member States with IORPs but incomplete data availability for the breakdown by size categories are indicated with ‘NA’. CZ, EE, HU, LT and RO do not operate IORPs; this is indicated with a dash (‘-‘). * Irish 2023 data from EIOPA 2024 factsheet.

Source: FISMA calculations with ECB for euro area pension funds and national statistics for non-euro area pension funds, EIOPA occupational pension statistic for IORPs.

Scale economies for pension funds

The relatively small size of pension funds in Europe implies that they cannot exploit important scale advantages, whereas larger pension funds can benefit from scale economies related to their investment strategies, fee negotiation power, their cost efficiency and risk pooling. Annex 4.7 reviews the economic literature on pension funds’ scale economies.

Compared to smaller funds and in particular individual savers, larger pension funds have better access to asset classes that are information intensive due to uncertain magnitude and timing of payouts, but that can deliver higher return over the long-term, such as infrastructure, or private equity. In addition, a small size impairs the capacity to run sophisticated investment, diversification and risk management strategies. These are elements for which scale economies imply substantial efficiency gains. Smaller funds usually rely strongly on investment in standard financial instruments such as bonds and listed equity ( 83 ), while in some cases they take complex investment decisions despite the lack of expertise, thereby exposing pension savers to undue risks ( 84 ). Even if they had the expertise to invest in private equity, the small amounts they can commit make them unattractive counterparts to scale-up firms ( 85 ). The subsequent problem driver zooms in on actual investment strategies and their profitability.

Larger funds have also better possibilities to manage processes and investment in-house to avoid paying fees to external providers. Small scale curtails the capacity to attract a high-skilled work force and to implement smart governance and control practices. This often translates into conservative investment policies, a low ratio of pension funds’ assets invested in equity and low returns for members.

The larger a pension fund is, the better its possibilities to diversify investment risks by spreading investment into numerous projects. A large member pool enables pension funds to run low buffers for individual members’ positions. Individuals’ investment strategies need to incorporate sudden liquidity requests and the possibility of living far longer than average. Pension funds can pool these risks, therewith providing implicit insurance to their members. EIOPA statistics show a much larger variation of returns for small than for large IORPs, suggesting that smaller pensions funds are less diversified ( 86 ).

There is also the perception of a lack of professionalisation of management in some smaller funds, which led Ireland to introduce requirements for small funds to invest via larger entities and Ireland and the UK to create masters trust to stimulate market-driven consolidation.

Annex 3 reports on studies that investigate whether larger pension funds are more profitable. A series of Dutch papers diagnosed that the pension market in the Netherlands, which is the largest in the EU, has consolidated to a stage where scale effects are no longer present ( 87 ). In all Member States except France and Sweden, the average size of pension funds is much smaller than in the Netherlands, suggesting ample scope to realise scale effects.

Reasons for the dominance of smaller pension funds

An important economic reason for the smaller size of pension funds in the EU may be the smaller size of private pension markets. The largest pension funds in the EU are in those Member States with a large use of supplementary pensions, suggesting that lack of demand is limiting funds’ size. The growth potential of pension funds is also curtailed by the alternative use of retirement savings in either life insurance contracts, as corporates’ book reserves or by households’ preference to save cash or invest in property . Offers of pension products or close substitutes from providers governed by different rules and regulations have an ambiguous impact on competition. In some Member States, this segments the market, allowing smaller providers to remain in niches and specialise rather than to grow 88 .

In those Member States with many pension funds populating a small national market, it is notable that some occupational funds originated from social protection schemes for specific professions or workers in industrial sectors 89 . Their capacity to combine pension savings with other targeted insurance services for these members creates a secure market niche but complicates mergers with other schemes. This might explain why many pension funds have a long history, but limited membership and relatively few assets under management. A further obstacle to scale are the IORP rules on mergers among pension schemes sketched in the table below, which have often delayed or rendered (cross-border) mergers of occupational pension schemes impossible to undertake.

Table 8: Regulatory requirement discouraging mergers and acquisition of pension schemes

The absence of cross-border pension schemes is another reason for the lack of scale. However, differences in languages and cultural attitudes among members present obstacles to the creation of such schemes. Differences in taxation, labour and social law are important reasons why most pension funds remain national ( 90 ).

·Different tax incentives, rates, and regimes can complicate the cross-border administration and attractiveness of pension plans.

·Compliance with different social and labour laws in each member state requires complex administration.

·Other consolidation obstacles that matter in the cross-border context are differences in guarantees or other design elements linked to the product offers, governance rules and consultation obligations. Different levels of consumer protection regulations in Member States can lead to discrepancies in how pension products are offered, impacting their standardisation. the consolidation obstacles that matter especially in the cross-border context are differences in guarantees or other design elements linked to the product offers, governance rules and consultation obligations. .

While several multinational enterprises offer occupational pensions to their members, they do so with different schemes and entities managing them in the different EU Member States ( 91 ). This reveals that the above-mentioned obstacles for cross-border pension offers are material. Although IORP II facilitated cross-border activity for IORPs, only 28 IORPs in the EEA actually operate in more than one Member State. More than half of them are domiciled in one Member State (BE) ( 92 ).

2.3 Out-of-scope drivers

Demographic trends and low productivity growth were singled out above as important determinants of the pension challenge. The decline in the share of working age population would be less of a challenge in an environment of high productivity growth. While this initiative aims to address the issue of low productivity growth in the EU by strengthening the role of pension funds on capital markets and therewith the role of capital markets in channelling funds to the most productive uses, demographic trends, other determinants of the weak productivity environment are considered out-of-scope drivers in this initiative.

Public pensions are a key component of households’ retirement income. The design of public pensions is an important determinant of citizens’ interest to save privately for retirement. The calibration of public pensions is considered an exogenous factor. The promotion of supplementary pensions should not be accomplished by curtailing the adequacy of public pensions. Boosting the use of supplementary pensions also does not intend to create leeway for public bodies to disengage from the provision of adequate public pensions. The creation of additional channels to save via supplementary pensions is not meant to undermine the benefits households expect to receive from public pensions. It is intended, however, to cater for the anticipation of many households that the claims they accumulate through their contributions to the public pension system may not be sufficient to ensure the income level they aspire for when in retirement.

2.4 How likely is the problem to persist?

If no action is taken, the pension gap will persist, and the supplementary pension sector will neither contribute meaningfully to adequate retirement income for citizens nor support the EU economy by mobilising long-term savings into productive, growth-enhancing investment. In the absence of EU action, Member States may either not act or not sufficiently. The PEPP and IORP framework would remain inefficient in helping address the problem drivers. Without targeted reform, the supplementary pensions framework will contribute little to retirement income. Thus, the pension challenges will not be successfully addressed.

3.    Why should the EU act?

3.1 Legal basis

The EU has the possibility to use binding and non-binding legal acts or a combination of them to support the development of the supplementary pension system. The use of non-binding recommendations is envisaged where there is a need to adjust the measures to country-specific factors to make them as effective as possible. The value added of effective rules set at Member States’ level tailored to different starting positions and the design of national pension systems outweighs the need for harmonised rules at EU level. Binding legislative measures are foreseen in those areas where EU legislation is already existing and the benefits from a common approach had been recognised when these were agreed. They cover the IORP II Directive and the PEPP Regulation. This initiative aims to make these frameworks more effective.

The review of these existing EU legislative frameworks for supplementary pension provision aims to strengthen the internal market for supplementary pensions. With a view to the single market dimension, legislative measures can be based on Articles 53 and 62 of the Treaty on the Functioning of the European Union (TFEU), which provide a legal basis for the adoption of measures to coordinate the provisions laid down by law, regulation or administrative action in Member States on establishing and providing services, and on Article 114 TFEU, the objective of which is the establishment and functioning of the internal market by enhancing measures for the approximation of national rules. These legal bases would allow the review of the existing EU legislative frameworks for asset-backed pension provision, including the IORP II Directive and the PEPP Regulation since these aim at strengthening the internal market for supplementary pensions.

Non-legislative measures can take the form of recommendations to Member States without binding legal character based on Article 292 TFEU. They aim to encourage Member States to consider measures in their remit to improve incentives for households to participate in supplementary pension schemes, for providers to offer such schemes or to help overcome information problems and behavioural biases. Due to their non-binding character, the use of recommendations would not force Member States to act or prescribe the way how to act but rather present the recommended best practices based on tested experiences. The formulation of recommendations would also need to reflect that they do not affect the competences of Member States to organise and design their national pension systems, as well as the autonomy and prerogatives of social partners, in light of their key role in occupational pensions schemes.

3.2 Subsidiarity: Necessity of EU action

The organisation of the pension system is a national competence and one of the main elements of Member States’ social protection policy. Supplementary pensions exist in all Member States though with large differences in their design and use. Over the last years, many Member States have introduced reforms aiming to strengthen supplementary pensions. A few Member States have also introduced auto-enrolment policies to enhance take-up of supplementary pensions.

EU action is necessary to accelerate progress with pension reform. Taking the Mercer pension index as gauge ( 93 ),only few Member States have already accomplished a top-performer status. Moreover, when projecting the past change between their first observation and 2025 observation in the Mercer index, most Member States will take still many years until they catch up to the best EU performers. On a positive note, none of the EU Member States covered in the 2025 Mercer report received a rating worse than C. But only two Member States received an A rating and two a B+, i.e. the pension ratings are worse than the ratings of their public debt. Member States with the lower level in their first observations tend to show somewhat larger increases in their general pension index, suggesting slow convergence towards the better performers.

Table 9: EU Member States in the Mercer/CFA global pension index

Rating classification: A > 80 includes also Iceland, Singapore, Israel; B+ 75-80 Australia and Norway; B 65-75 UK, Switzerland, Canada; C+ 60-65 USA and Malaysia, C 52-60 Brazil, China, Japan. Other countries are a selection of 52 countries covered.

While Member States can, in principle, enact appropriate measures at national level that could possibly address some of the problems identified, action at EU level is warranted to address the problem drivers that led to the low use of supplementary pensions and thereby to their small contribution to additional retirement income and to capital market development. In the face of accelerating demographic change and growing fiscal pressure of first pillar pensions, supplementary schemes must be significantly scaled up to secure future income adequacy. Coverage and transparency of supplementary pensions are common challenges for most EU Member States with the EU in a central place to promote the learning of good practices among Member States. At the same time, underdeveloped capital market with hereto linked implications for the return opportunities of supplementary pensions is a common EU problem that EU Member States alone are not in the position to address. The EU needs a deeper and more integrated capital market that can fuel and harness long-term retirement savings. Failure to act, will result in a fragmented, inefficient status quo and squander the opportunity to deliver long-term value for both European citizens and the real economy.

The shortcomings of national policies in fostering supplementary pensions are visible in occupational and personal pension offers having remained predominantly national. Despite the EU wide regulatory regimes for occupational pension funds (IORPs) and pan-European pension products (PEPP) ( 94 ), the lack of scale in conjunction with low degrees of competition on national markets impairs the attractiveness of personal and occupational pension products for individuals and employers. Regulatory divergences and differences in supervisory practices continue to constrain the development of cross-border supplementary pension provision and prevent economies of scale. These issues limit the availability of attractive, cost-effective pension products and larger supplementary pension providers, and reduce the capacity of supplementary pension institutions to contribute to long-term investment in the EU economy. These objectives cannot be sufficiently achieved by Member States’ individual initiatives, but can rather, by reason of the scale and effects of the action, be better achieved at Union level. Moreover, old-age income adequacy and fiscal sustainability of national pension systems are crucial to the stability of the Union as a whole. 

3.3 Subsidiarity: Added value of EU action

The key value added from action at EU level originates from the long-term nature of the pension problem. Costs and consequences of policy measures will materialise in most cases beyond electoral cycles that often determine policy action at national level. The EU action offers a forward-looking long-term perspective to the pension challenge, thereby allowing stability of objectives and consideration given to the long-term effect of policy measures. While some of the measures will need to be implemented by Member States, their support through an EU framework that finds application in other Member States with comparable measure adjusted to national conditions influences the policy discourse and can facilitate their public acceptance.

At the same time, all Member States face the pensions challenge and a growing pension sector in one Member State will have positive externalities on other Member States. Leveraging on existing best practices and joint efforts by Member States on auto-enrolment, pension tracking systems, and pension dashboards would also be more effective than individual initiatives to achieve larger scale, coverage, and transparency of supplementary pensions, and ultimately adequate pension income across the EU population. Joint efforts would in particular help boost capital market development in smaller Member States, which would not only benefit from a larger domestic institutional investor base but also from access to a larger institutional investor base in other Member States.

4.    Objectives: What is to be achieved?

4.1 General objectives

The general objectives of the supplementary pension package follow those formulated in the SIU Communication: (i) Ensure adequate income for citizens in retirement; and (ii) Enhancing the capacity of pension funds to channel households’ savings into productive and innovative investment. Both general objectives are complementary and mutually reinforcing.

Increasing citizens’ participation in supplementary pensions to ensure their adequate income in retirement requires the creation of conditions for households to consider their participation by providing a suitable information base for them to decide and the setting of incentives to act. A larger share of household savings in supplementary pension schemes can complement public pay-as-you go pensions in supporting future adequacy of retirement income. An increased coverage of supplementary pensions would help achieve a more comprehensive retirement income and therewith allow that citizens can maintain an adequate lifestyle after retirement and benefit from a secure financial future. Participation in asset-backed pensions depends not only on incentives and product design but also on trust in the institutions managing retirement savings. While pension institutions in the EU are generally well regulated, maintaining and reinforcing this trust is essential. Enhancing consistency in core prudential safeguards across private pension providers could support long-term confidence and may justify a proportionate review of the existing rules and its scope.

Enhancing the capacity of pension funds to direct households’ savings into productive and innovative investment requires setting conditions and incentives for pension fund providers to allocate a reasonable share of pension savings to funding for corporations, including to innovative firms that require equity financing. The traditional investment focus of EU pension funds in sovereign and high-quality corporate bonds has only an indirect effect on productive investment via the transmission of lower interest rates on these markets into lower funding costs for the corporate sector. Enabling pension providers to make stronger use of (public and private) equity markets would have a more direct effect on availability of funding and their costs. The higher long-term profitability of equity investment and better diversification opportunities could generate higher returns for retirement savers, and it would yield a stronger and more immediate effect on competitiveness of the EU economy.

Both objectives are mutually reinforcing. Higher returns for members will make it more attractive for citizens to enrol in supplementary pensions schemes. The take up of additional pension savings will allow pension funds to realise scale effects and to direct an increasing share of their portfolio into the real economy. With their potential to significantly boost the pool of patient capital available, supplementary pensions could play a key role in fostering the EU capital markets for long-term investments.

The objective of adequate retirement returns can be quantified through the assumption that individuals want to accomplish the current public benefit ratio when they retire, i.e. pension payments relative to average wage ( 95 ). The ageing report projects how this ratio will decline over the next fifty years and individuals can be assumed to offset this decline through supplementary pension savings. They could accomplish this through saving more or through earning a higher return on their savings. The figure below shows for the EU-27 aggregate which combinations of higher participation in long-term pension savings and the returns on these pension assets could offset the expected decline in the benefits that individuals can expect to receive from public pensions ( 96 ). The higher the targeted returns, and with that the risk of the investments, the less the person would need to save. If a pension fund is able to generate higher returns, the additional saving needs are lower and more of income is available for consumption.

4.2 Specific objectives

4.2.1 With respect to increasing participation in supplementary pensions to ensure adequate income in retirement

This objective can be decomposed into the following sub-objectives:

- Improving information for citizens about their future pension levels and pension products available to them: Awareness and being able to take informed decisions is conditional on individuals having access to clear, comprehensive and easily accessible information about their entitlements in different pension schemes and the expected benefit amounts. Most people in the EU receive pension information from various pension providers and need to aggregate this information for themselves. Most do not have access to pension tracking systems that cover all pillars and provide such information in aggregated form.

- Providing incentives for overcoming procrastination: Employers and for savers make little use of pension products. Employees’ participation in occupational schemes often suffers from inactivity bias when voluntary, and from the adverse perception of labour income being taxed if participation is mandatory. For ensuring that participation in supplementary pensions increases among groups where it has traditionally been lower, for example, women in lower-paid or part-time jobs and taking longer career breaks, for self-employed, requiring flexible contribution patterns to take account of volatile income streams, younger people working in precarious and “non-standard” employment and lower-income workers in general.

4.2.2 With respect to improving pension funds’ capacity to direct households’ savings into productive and innovative investment

This second objective can be broken down into the following issues:

- More efficient management of retirement savings and supervision. While retirement savers are not in the position to monitor pension providers’ activity and request corrective measures, supervisors can. Supervisors should have sufficient information, including on costs and fees, and the capacity to put pressure on weak performing entities to take action, for example, to review their investment policies, pool assets with external asset managers or other pension funds, or consider merging with other pension schemes, thereby compensating for the lack of internal managers’ incentives to give more priority to returns.

- Promotion of scale and competition forces. Building up scale reduce costs that are charged on pension savers and could lead to more diversified investment strategies. It would strengthen risk management capacity and access to more profitable, illiquid asset classes. The absence of a level playing field due to the differences in national legal frameworks is reducing competitive pressure on pension providers therefore risks subpar investment policies and high costs. Few occupational pension providers are active across borders. EIOPA identified obstacles to cross-border activity that impair a more competitive framework, more scale and more efficient investment outcomes.

- Streamlining explicit investment barriers for all private pension providers: Certain pension institutions, including IORPs, but also pension funds outside the scope of EU legislation, often face regulatory limits as regards the eligible investment universe, with fewer possibilities to diversify and create returns for their members. A review of national capital requirements and quantitative limits to investments would strengthen pension funds’ capacities to finance productive investment ( 97 ). Since scale determines their capacity to invest in sophisticated, riskier and illiquid financial instrument, the last two specific objectives are interrelated.

- Ensuring cost-efficient pension products: Given the low returns on many pension schemes and products in the European Union, it is essential to enhance their cost effectiveness. Fostering large scale pension products, possibly at EU scale, could lower costs, fees and lead ultimately to higher returns for members. Providers of pension products need to be enabled to offer cost-efficient products and to compete with incumbents, which necessitates ensuring transparency and competition on fees and costs and optimising the regulatory framework for these schemes and products. This can create the conditions to deliver better returns and increase the perception of supplementary pensions as a valuable long-term investment for retirement, ultimately boosting participation and engagement in these schemes and products.

- Reducing regulatory fragmentation to encourage market-driven consolidation and the cross-border offering of services. Many: pension institutions operating funded pension schemes, where the level of benefits offered to members depends on the amount of contributions paid and on the performance of the institution in managing assets, are currently not captured by any EU prudential legislation ( 98 ), therefore subject to diverse national rules, including different supervisory processes. In several Member States, employment-related retirement provision is provided by funded institutions which are excluded from the scope of Directive (EU) 2016/2341 pursuant to its Article 2(2), or are otherwise not covered by that Directive without being subject to any other Union prudential legislation. Taken together, such institutions excluded from the scope of Directive (EU) 2016/2341 hold very substantial assets and constitute an important component of the overall employment-related retirement landscape alongside IORPs. The absence of a minimum harmonisation EU framework for those funded pension institutions results in significant disparities in the protection of members and beneficiaries across Member States. In addition, differences in pension schemes across Member States impair the portability of pension claims, which makes it unattractive to participate in national supplementary pension schemes for citizens intending to move jobs and particularly so if they plan to move into other Member States. Relevant for pension funds and savers are instances of tax discrimination across the EU Internal Market due to national tax legislation that enforcement by the Commission could address. Fragmentation of pension savings along national markets prevents pension funds from exploiting scale economies and reduces competitive pressure to generate returns. The cross-border provision of services would help increase the scale and efficiency of the sector, thereby enhancing pension incomes.

Figure 19: The link between problem drivers, specific policy objectives and possible policy levers

5.    What are the available policy options?

5.1 What is the baseline from which options are assessed?

In a no-change scenario, the supplementary pensions landscape across the EU will remain complex, fragmented, unable to deliver adequate returns and underdeveloped, failing to deliver on its full potential. Market participation will remain limited, cross-border provision negligible, and disparities between Member States persistent.

Replacement rates will develop along the projections of the 2024 Ageing Report and the proportion of European citizens with access to supplementary pension arrangements would stay at its currently low level. People with low incomes and other disadvantaged population groups will remain reliant on public pensions whereas high-income earners continue to benefit over-proportionally from the existing offer of personal pension products. Workers in atypical jobs, in industries dominated by small firms, and in sectors or countries without role for social partners will remain deprived from access to occupational pension schemes.

Without measures to increase participation, pension funds will not be able to build up scale economies so that measures to promote pension funds’ investment strategy would lead to a diminished increase in returns. Without a stronger supplementary pension sector, productivity developments in the EU will depend more strongly on other policy initiatives that aim to enhance innovation, competition and structural change. The role of capital markets in funding productive investment and financing transformation change will be smaller without the boost to capital market development expected from a larger pension sector. Finance will therefore act less as an amplifier of other productivity-enhancing policy measures.

Without measures to increase participation, pension funds will not be able to achieve sufficient scale so that measures to promote pension funds’ investment strategies would lead to a diminished increase in returns.

5.2 Description of the policy options

5.2.1 With respect to increasing participation in and transparency of supplementary pensions to ensure adequate income in retirement

5.2.1.1 Pensions tracking systems to back informed individual choices.

A way to incentivise more engagement with the pension question and thus improve coverage and future adequacy is through the development of pension tracking systems. Pension tracking systems allow individuals to obtain an overview and an estimate of their future retirement income in one place, thereby identifying shortcomings and reacting accordingly early on. The value added of such a system consists in centralising information from different pension schemes in one place for easy overview of individual accrued pension entitlements. Annex 2.2 presents evidence on the use and economic effects of pension tracking systems

Online platforms that provide access to individual information about pension entitlements exist in the majority of Member States, however, their scope, design and governance vary widely. According to the 2021 EIOPA technical advice on pension tracking systems ( 99 ), 19 Member States provided digital personalised information on at least one type of pension. According to a survey carried out in 2024 by the European Tracking Service, a national tracking service (using a slightly different definition) existed in 17 Member States. At the same time, most of the national tracking services cover only one or two pillars of the pension system, thus falling short of the objective to provide a comprehensive overview of individual pension rights through a single-entry point. Furthermore, the existing tracking services used very different solutions with regards to digital identification tools (national, private sector or tailor-made digital ID; eIDAS compatible or not), information storage (live access or central data storage) and governance (public, private, or public-private partnership ), limiting their development potential and compatibility.

In the context of increasingly dynamic careers and multi-pillar pension systems, resulting in savers accumulating pension rights in different ‘pension pots’, there is clear demand for pension tracking to boost pension awareness and adequacy. This is illustrated by the fact that a growing number of Member States are setting up some type of a pension tracking service (e.g., Germany and Ireland in recent years). Member States that already have comprehensive platforms reported positive experiences in a survey to the Commission in 2024. At the same time, many Member States still lack any national tracking service, and most of those that have been set up fall short of the complete coverage of their national pension system. The existing digital information tools generally cover statutory pensions; however, the coverage of occupational and in particular personal pensions remains much more limited. While the benefits of more complete information to individuals are obvious, setting up a comprehensive and compatible national tracking service requires solutions to address a number of technical, legal and organisational challenges.

To facilitate access to pension information for mobile workers, a cross-border European Tracking Service (ETS) is being developed ( 100 ), with the support of the Commission. As of 1 July 2025, two Member States ( 101 ) are fully connected to the ETS, with at least 5-7 connections planned by 2029. Since the ETS operates by connecting national tracking services, its long-term development, as well as its ability to cover a critical mass of the mobile workforce in the EU, remains conditional on the existence of comprehensive and compatible national tracking systems.

EIOPA technical advice on the development of pension tracking systems ( 102 ), prepared in 2021, mapped existing pension tracking services in the EU, EEA and third countries and identified best practices with regard to design, governance and implementation. This technical advice can serve as the point of reference and valuable knowledge base for further action at national and EU level.

Figure 20: Main lessons learned from existing pension tracking systems

Source: EIOPA.

The ETS can also engage with national administrations and stakeholders to provide guidance and facilitate exchange of experience in the development of national tracking services.

The options for action at EU level are the following.

Option 1: no measures at EU level

The first option would be not to undertake any new action at EU level and to continue relying on national initiatives (coming from the public sector, industry, or both) to develop pension tracking services across the EU. Member States wishing to develop a national tracking service could use the existing EIOPA report and/or refer to the ETS for exchange and guidance.

Option 2: A recommendation to Member States to introduce pension tracking systems

A second option is that the Commission issues a recommendation that calls on Member States to address the challenges and introduce pension tracking systems, making use of the most relevant good practices identified in the EIOPA advice and following certain minimum features. These features would pertain both to the standard of services offered to the national citizens and technical compatibility allowing connection to the ETS.

Pension tracking systems should incorporate harmonised reporting under the IORP II Directive, the PEPP Regulation and the Insurance Distribution Directive to ensure consistent, comparable and comprehensive information on pension entitlements.

Option 3: A legal obligation for all pension institutions to provide information through pension tracking system accompanied by a recommendation to use best practices

A third option would introduce a legal obligation to provide information through a pension tracking system, accompanied by a recommendation outlining best practices. Such obligation could be addressed to supplementary pension providers, replace existing legal obligations (such as IORPs requirement to produce pension benefit statements and to send them to their members) and eventually also to Member States to design their pension tracking systems. To help overcome individuals’ procrastination in making long-term saving decisions, pension tracking could be connected to platforms that present information about pension products and other information. This could foster individuals’ financial literacy in a moment when they show interest in learning about financial matters and even could nudge individuals into simple default pension products.

5.2.1.2 Pension dashboards to assess the effectiveness of Member States pension systems.

Another way to incentivise more engagement with the pension question and thus improve coverage are pension dashboards. Unlike pension tracking, which provides personal data to individuals, pension dashboards show country-wide information on pensions with the objective of highlighting gaps in the sustainability of pension systems and in their capacity to deliver adequate pensions to the population. Thus, pension dashboards help Member States assess the performance of their multi-pillar pension system. Since some countries and the OECD refer to “pension dashboards” when meaning tools that track individual pension entitlements, there is some confusion about the term and the possibility for other names could be explored to underpin its notion of a surveillance and monitoring tool ( 105 ).

Dashboards are a tool to create a political setting that allows for appropriate peer pressure to be exercised, so that Member States identify and address shortcomings at their level and are incentivised to learn from best practices. They have been applied in the EU policy context in different policy areas such as for the single market monitoring, the macroeconomic imbalance procedure, the digital economy and society. The Commission already compiles data on pension adequacy and sustainability every three years, covering all EU Member States. While public pensions are well covered, information about supplementary pensions is scarce for many Member States.

The CMU High-level forum advocated the creation of pension dashboards to provide a more comprehensive view on gaps in pension sustainability and adequacy. It advocated comprehensive cross-pillar indicators on pension adequacy and sustainability across the EU including target levels, i.e. including supplementary pensions. The Eurogroup invited the Commission to develop a pension dashboard in collaboration with EIOPA in 2024. In reaction to the recommendation by the High-level forum, the Commission had already asked EIOPA for advice on pension dashboards and the advice, delivered in 2021, identified large data gaps.

The motivation for the monitoring of pensions

Many Member States have reformed their multi-pillar systems or are planning to do so. While the need of pension reform is widely acknowledged, surveys document an entrenched opposition to pension reform in large parts of the population ( 106 ), leading to passionate and often contentious public debates. The provision and communication of fair and comprehensive information about reform needs and its impact may not be a sufficient condition for reform proposals to get public acceptance, but it is a crucial precondition to obtain it. Without reliable data, an objective and informed debate about pension reforms is hardly possible.

Communication about reform needs and the impact of pension reform would benefit from data and tools. While a lot of data, statistics and methods is available to analyse public pensions, there are still large data gaps with respect to supplementary pensions in many Member States, which impairs a systematic monitoring of the overall adequacy and sustainability of their multi-pillar pension systems. The variation of participation rates in supplementary pensions across the various data sources documented in Table 30 in Annex 2.4 highlights inaccurate data on a basic parameter. Annex 5.1 investigates ways to fill data gaps. Less than a dozen Member States project expenditure on supplementary pensions as part of the Ageing Report. While the focus of this report on the sustainability of public pensions can justify that the other Member States do not do these projections, it is notable that information on the contribution of supplementary pensions to replacement rates is missing for broadly the same group of Member States in the OECD pension statistics.

This suggests that few Member States seem to be able to base policy decisions about supplementary pensions on information that is sufficiently forward-looking into the long-term over which pension reforms are designed. They also lack a tool that allows them to compare national pension performance with that in other Member States as benchmark and inspiration to look for good practices from other Member States.

Option 1: no EU action on pension dashboards

In a survey of Member States only two reported they had a dashboard at national level. Statistical offices or other public bodies in several others have public websites that show information about use, trends and performance of the national pension system and in some there are also reports and websites with information on supplementary pensions. This suggests that the transparency on pension performance is beneficial for public authorities and the political debate in identifying challenges early on also at national level.

Under this option, the Commission would continue to work with Member States on the pension adequacy report and the ageing report, which provide detailed country-specific information and projections every three years. EIOPA would collect data on insurance companies and IORPs for prudential purposes, while the ECB, Eurostat and the OECD would maintain their statistic coverage ( 107 ).

Option 2: A recommendation to Member States to produce pension dashboards at national level

The objective of a recommendation would be to entice Member States to compile data on pension coverage, contributions, expenditure and adequacy in form of a national dashboard at a central place. This data should include public pensions, which is already available to public authorities, and also existing data on supplementary pensions. The benefits of such a dashboard would be greater if it included long-term projections in addition to the most recent observations and past trends and if it provided a granular breakdown of adequacy across different population groups and dimensions of pension adequacy.

In addition to covering standard indicators on pension adequacy, it would be warranted if pension dashboards covered possible adverse effects of rising supplementary pensions on other dimensions of pension adequacy and sustainability. Many Member States provide fiscal incentives for supplementary pensions, which adds to the sustainability dimension. Another important dimension of pension adequacy is the reduction of poverty risks. Supplementary pensions are likely to cement income distribution ( 108 ). If wealthier individuals can make better use of the new opportunities, the income distribution may even widen. It also seems that wealthier individuals can afford to retire earlier. If the rise of supplementary pensions leads to a decline in employment, this needs to be spotted early on. Member States could be recommended to identify existing data or indicators to cover these dimensions in the dashboard.

Option 3: A common approach to a European pension dashboard

The starting point for a pension dashboard would be the data compiled by the Commission and Member States on pension adequacy and sustainability. These cover all Member States and are updated with the help of pension experts every three years. The focus of these reports is on public pensions since only few Member States provide data on supplementary pensions. A first step would be to recommend to Member States to increase the coverage of national data on supplementary pensions for the Pension Adequacy Report and to those of them that do not yet do so, to add projections on supplementary pensions in the Ageing report.

A common approach aims to ensure comparability of data across Member States. Common definitions and methodologies would make national data more comparable. EIOPA suggested the steps towards such approach. The first step would be a mapping of pension products and pillars. For the second step, the Commission could create a data template and request Member States to collect data to fill the template.

Figure 21: Basic steps to develop a dashboard

Source: EIOPA.

5.2.1.3 Auto-enrolment and its best practices

A powerful method to overcome procrastination of workers would be to oblige them to join supplementary pensions schemes where such pension schemes are offered by their employer. Some Member States have made the use of occupational pensions mandatory or quasi-mandatory for workers such as BG, DK, NL and SE.

Auto-enrolment means that employees are enrolled by default in occupational pension schemes offered by their employers but are given the choice to opt out. The possibility to opt out is an important element of auto-enrolment and makes it different from mandatory enrolment. It also differs from the standard approach in voluntary occupational pensions where employees are asked whether they want to participate and need to enrol actively. Due to the possibility to opt out, which is an essential component of such a system, it is less intrusive than obliging employees to join such scheme, which is the main reason behind the high take up of occupational pensions in DK, NL and SE.

Behavioural economists have proposed the use of auto-enrolment as a commitment device to overcome people’s procrastination and bring them into pension schemes. As means to overcome low voluntary saving for retirement, they recognised that it has a paternalistic element but was not imposing participation upon people ( 109 ). In legal advice provided ahead of the Irish introduction of auto-enrolment, the possibility to opt out was noted as a central element to ensure compliance with employees’ property rights in the European Convention on Human Rights (ECHR) and the Irish Constitution, since it minimises the weakening of these rights while allowing to address a pressing social concern ( 110 ).

Countries that introduced auto-enrolment like New Zealand or the UK have observed that it led to a significant increase in the coverage of occupational pensions. Auto-enrolment or elements of it already exist in some EU Member States (DE, IT, PL) or are being introduced (IE, SK) in occupational pensions. Auto-enrolment helps overcome the bias in decision-making of remaining inactive. Studies identified design features that determine its success. The European Commission financed a study on the experiences and best practices in auto-enrolment that was published in 2021 (AE study hereafter) ( 111 ).

Since starting positions of Member States differ strongly with respect to the organisation of public pensions, labour market institutions and existing provisions in the field of occupational pensions and saving behaviour in the population including their responsiveness to fiscal incentives and risk attitudes, a harmonised introduction of auto-enrolment through EU legislation would not be effective. The need to tailor the introduction to specific conditions in the Member States is so important that even a Directive could not be sufficiently granular to cover general principles while leaving sufficient discretion for Member States to adapt in their transition in national law. A recommendation to Member States therefore appears to be the best suitable instrument. The following options differ in the scope of such a recommendation.

Option 1: baseline: no action on auto-enrolment

Under this option, a Member State would be able to initiate a political discussion about the introduction of auto-enrolment in their jurisdiction, making use of the AE study. They could also apply for technical support from the Commission under the Technical Support Instrument. This possibility exists for some time but has been used only by few so far and only as option in the broader context of strengthening the position of beneficiaries ( 112 ).

Option 2: Enabling auto-enrolment

Legal barriers may exist in national legal systems that discourage the use of auto-enrolment or even prevent it. These obstacles may stem from the fact that auto-enrolment is so far not envisaged in legal systems. Instead, often the law requires the savers to declare their willingness to enrol in a pension scheme. Undoing such barriers would enable the use of auto-enrolment. An explicit legal basis for auto-enrolment would be a strong enabling signal ( 113 ). It would also be an essential precondition to define the main features of auto-enrolment, e.g. which entities are entitled to enrol, which savers are eligible and which products are available to them. Auto-enrolment has been practiced in employment contexts enabling employers or social partners in many Member States, which fostered occupational pension schemes. When used for occupational pensions, the regulatory framework may need to define what will happen to the accumulated pension savings if the employee changes job, industry or starts working abroad.

To protect savers that are auto-enrolled, policy makers would need to ensure appropriate safeguards so that retirement savings are invested in a transparent and sufficiently safe way. Providers of retirement savings would need to be properly supervised and investment policies would need to reflect the specificities of long term retirement savings. This could mean appropriate regulation of providers and supervisors would need to be well equipped to enforce prudential safeguards and to intervene effectively if schemes do not follow appropriate investment and risk management strategies.

To build and maintain the credibility of the system, this enabling approach should also include commitments that private retirement savings will not be merged with public pensions in the future and that the government will not direct private retirement savings into investment for public needs. Governments would also need to ensure that the character of auto-enrolled pension savings as supplementary pensions are ensured. Auto-enrolment systems are intended to complement, not compete with, existing occupational pension schemes, particularly those established through social partnership or collective bargaining arrangements.

This requires that parameters are set in a way that the new schemes do not crowd out well-functioning existing schemes. This would also mean a clear separation from public pensions so that the contributions to the supplementary scheme are understood to be individual savings and not as an additional tax. Auto-enrolment schemes should also not replace already existing private schemes with a mandatory character. Nor should the enabling of auto-enrolment disadvantage members in already existing schemes. 

Option 3: promoting auto-enrolment through the application of good practices

In addition to enabling the use of auto-enrolment, Member State governments have various means, at different intervention intensity, at their disposal to promote the use of auto-enrolment. Annex 2.1 analysis the evidence on critical design elements and Table 28 in that Annex provides a repository of good practices. The importance of a good design of the schemes in ensuring the success of auto-enrolment has been evidenced by the case of two Member States that have abandoned it. Estonia found that supplementary pensions produced low returns for savers 114 , Lithuania was inter alia concerned about distributive effects from tax incentives to the benefit of only part of the population.

An essential ingredient of any promotional efforts would be information to pension savers, employers and other stakeholders. Accompanying information campaigns would need to create awareness, understanding and buy-in. Although auto-enrolment circumvents the passiveness of parts of the population in taking active saving and investment decisions, opt-out rates might be higher if there is no minimum awareness, understanding and trust. A public debate early in the design phase would be a good practice to define objectives, clarify warranted effects, create trust and address concerns. Historically, the introduction of auto-enrolment was accompanied by information campaigns on pensions in the UK, Italy and New Zealand. The Irish statistical office has monitored the population’s awareness on auto-enrolment via annual surveys since 2022 ( 115 ). A survey by a private consultancy firm among businesses in Ireland

revealed that two thirds or respondents had concerns a year prior to the introduction of auto-enrolment across the areas shown in the table below.

Table 11: Employers concerns about the introduction of auto-enrolment in Ireland

Source: AON (2024) ( 116 ).

Ideally, measures to improve financial literacy should run in parallel to information campaigns on auto-enrolment to create synergies. Targeted financial education programmes, aimed at the population eligible for auto-enrolment, might be good practice to exploit synergies resulting from providing information that is specific to individual needs and meeting them at a time when they are open to learning.

A sufficiently prolonged and gradual implementation can be considered a good practice. The AE study describes a short implementation period as the reason behind the lack of success of auto-enrolment in Italy, Poland and Turkey. This short implementation gave stakeholders little time to prepare for their obligations or for the competent authority to establish licensing arrangements, and reduced the effectiveness of the information campaign to raise awareness amongst savers before the regime started to become operational. Ireland consulted and committed to introduce auto-enrolment in 2018, with the system becoming operational at the beginning of 2026. It organised various information sessions for different types of stakeholders, as well as an advertising campaign which began several months before the launch date. Despite the long preparation time, implementation was postponed several times, and it will ultimately take a decade for the scheme to be fully bedded in.

Since auto-enrolment primarily targets savers that do not take active investment decisions, there should not be a complex decision environment regarding their contribution rates, how they invest and how they can use their accumulated pension claims later on. A single or a limited number of default options is considered good practice. In occupational pension schemes, members contribute to the scheme set up by the employer or agreed by social partners as is the case in social partner agreements in Germany. An alternative approach is used in the UK, Poland and Canada, namely that employees can select funds pre-selected by their employers ( 117 ). The UK offered small employers to automatically enrol their employees into a standard scheme called ‘Nest’ (see below). A related successful example is the Swedish prime pension, which allocates contributions to a default fund unless members decide otherwise. Ireland will offer members a limited set of investment options along risk categories, with those who do not choose an option being enrolled in a default “life-cycle” option. Overall, most countries that use auto-enrolment also favour the use of pension funds with life-cycle strategies – including Poland, Canada and the UK’s Nest (see below). Outside auto-enrolment, the Swedish default AP7 Såfa plan is another example for applying a life cycle strategy.

How eligibility and default options are designed would depend on Member States’ specific policy objectives, legal and social conditions. Their choices would ideally be as inclusive as possible. 

1.Since pension saving becomes more effective the longer the accrual period, it is advisable that eligibility criteria foster an earlier start of the pension savings process, while not disadvantaging older workers enrolling in supplementary pension schemes. When used in an employment context, eligibility for employees with atypical work contracts and opt-in possibilities for self-employed could be a good practice.

2.The magnitude of the contribution rate cannot disregard the contributions to other parts of the pension system, other social security contributions and what savers can afford. A low contribution rate is likely to reduce the number of opt-outs, especially for low-income earners. At the same time, it reduces the cumulative retirement income of those most in need.

3.Beyond the level of the contribution rate, flexibility to adjust it, opt-out and re-enrolment conditions will incentivise people with career interruptions or atypical work patterns to enrol and keep enrolled. These parameters are particularly important for women since they are more often involved in caregiving activities than men, therefore interrupt their career. Re-enrolment occurs every three years in the UK and every two years in Ireland, unless the employees are in another pension scheme. A different approach is used in New Zealand where new employers must enrol new employees.

4.Conditions on early withdrawal are likely to influence opt-out rates of different social groups and the cumulation of their retirement income, too. Early withdrawal is possible in several countries such as New Zealand, the UK, Ireland and France under different conditions, predominantly related to circumstances of individual hardship or inability to work due to disability or illness.

It would also be warranted to set rules for the financial intermediaries that manage the investment of the pension savings. A good practice would be to ensure that savers have the choice between receiving a lump sum or an annuity ( 118 ). Governments may also want to require that financial intermediaries involved in the investment policy do not charge excess fees to the detriment of savers’ retirement income. This would also need to cover fees in the decumulation phase to avoid payments in form of an annuity becoming too expensive.

Fiscal incentives and public subsidies are suitable means to promote the use of auto-enrolment in supplementary pensions. Experience shows that pension schemes are most successful where there is a fiscal incentive for individuals to remain a part of such schemes ( 119 ). Poland pays lump sum amounts to members from the public budget to reduce the opt-out of persons with low income. Ireland will provide a State “top up” of €1 for every €3 saved by an individual, while the UK provides tax relief on contributions. The academic literature describes tax deductibility as effective tool in making auto-enrolment attractive for middle-income earners. Since auto-enrolment is usually used in the employment context, tax incentives can also be crucial to entice employers and to compensate them for costs they incur. It is common that employers are obliged to contribute to their employers’ occupational pensions ( 120 ) and can deduct their contributions from their tax duty. Many Member States set quantitative ceilings on the amounts deductible whereas Spain and Germany offer employers additional tax incentives.

Different policy instruments have been used to support smaller employers in offering occupational pensions and could also be used to promote the use of auto-enrolment. These measures were applied in view of the supply of occupational pensions being often limited to either large firms or to employers participating in collective agreements with social partners. A successful practice in Poland and the UK was that obligations for smaller employers were phased in with a delay to give them more time to prepare.

EIOPA identified the following principles as important for the success of auto-enrolment: (1) Raising public awareness and pension literacy for building trust, (2) Deciding on the timing of its introduction in a contextual setting, (3) The actual design of the scheme. The key features should include (a) start early, i.e. for young age cohorts, (b) keep it simple, i.e. avoid complexity and offer clear default options, (c) make it the norm, (d) build trust through stability, clarity, transparency and strong supervision ( 121 ).

The replies to the targeted consultation revealed that stakeholders consider public incentives as the most important design feature, followed by the avoidance of adverse effects on statutory pensions and their sustainability. The other design elements asked for are in a broad range and only early withdrawal possibilities stand out as considered of lesser importance.

Table 12: Hierarchy of best practices along their importance, stakeholder replies to the targeted consultation

Note: Average measured as priority on a scale 1 to 8 with one the most and 8 the least important, frequency measured as how often referred to in 112 responses to the consultation.

Option 4: Active government support: Obligations for employers and public administration

The government can intensify its support for auto-enrolment by addressing some key obstacles. When applied in a work-place context, options 2 and 3 leave the decision on the actual use of auto-enrolment to the discretion of employers. Possible employers’ resistance to offer auto-enrolment to their employees can be overcome by obliging employers to auto-enrol their employees. The cost of administrating members’ pension savings and allocating the contributions into investments can be reduced by establishing a body that performs these tasks.

An obligation for employers to contribute to their employees’ occupational pensions is standard in most systems. The fiscal burden is reduced by employers benefitting from possibilities to deduct their contributions from tax, see above. Auto-enrolment yielded a very strong increase in the participation in occupational pensions when governments obliged employers to enrol their employees, i.e. it became mandatory for the employer to offer occupational pensions, though not mandatory for employers as they have the possibility to opt out. The AE study described the obligation for employers to offer auto-enrolment as a best practice, referring to the results in the Canadian province of Quebec and some US States. The UK, Poland and Ireland apply it too ( 122 ).

Some countries use a public body for the administration of supplementary pensions. Such a body can also act as investment vehicle or as master trust that outsources investment management to private pension funds. Lithuania used the state social insurance body ‘Sodra’ to auto-enrol employees and self-employed and allocate them to pension providers. The UK created a new public body called ‘Nest’ and financed it for the first years. It serves as a master trust and was established to ensure that all employers are able to fulfil the obligation to offer auto-enrolment. Therefore, Nest is required to accept any employer, even if very small, and it also allows individuals to consolidate their multiple pension pots into one. Ireland has created a one-stop shop ( 123 ) to determine employee eligibility, collect all contributions (employee, employer, State), administer these contributions and procure the service of investment firms to invest the proceeds. Although used in a mandatory and not in the auto-enrolment context, the Swedish AP7 Såfa functions as default investment fund; the administration is processed by the Swedish Pensions Agency ( 124 ).

The use of a public body would allow synergies between public and supplementary pensions concerning the identification, calculation and payment of contributions. Such a centralised model, with state agencies acting as contribution collectors and providers of investment schemes might further offer efficiencies of scale and simplified management. It would have more bargaining power with asset managers than employers, and could obtain lower fees. Delegating the administration to a public body would further imply that existing payroll systems in firms would require little adjustment, which reduces administrative burden for firms and over-proportionately benefits smaller enterprises. Since the use of such a body in the UK and Ireland is complemented with an obligation for employers to enrol their employees it passes part of the administrative burden from employers to the public thereby evading other fiscal incentives for employers to make use of auto-enrolment. The organisation of occupational pensions through social partners provides a comparable function in some Member States, leaving the actual use of auto-enrolment to the discretion of business associations and trade unions. The creation of a public body would therefore be more suited for Member States without such a role for social partners.

5.2.2 Fostering access to attractive pension products

After discussing different options for increasing participation in pension savings, this section focuses on returns and better risk management of the provided pension savings. More specifically, the first sub-section discusses how to advance better investor returns, ensure more transparency on cost and returns and increased risk management capacity across workplace schemes. The second sub-section looks into possibilities to allow the Pan-European pension product PEPP to gain market traction and to develop into a product that can compete with national pension products and intensifies competition towards lower fees and better terms for pension savers.

5.2.2.1 Fostering scale and performance of the IORP sector through the review of the IORP II Directive

As stated previously in this staff working document, economies of scale are essential to lowering administrative and investment costs, enhancing risk diversification, and expanding access to professional portfolio management. Without sufficient scale, pension providers may struggle to offer cost-efficient, resilient, and transparent products.

The Commission intends to review the IORP II Directive with a view to strengthening the framework for asset-backed pension provision and fostering better long-term investment outcomes. The review also seeks to make the IORP model more relevant for Member States where such institutions remain underdeveloped.

The primary objectives of the IORP reform are to promote stronger and more sustainable investment outcomes for beneficiaries, enhance transparency around both costs and returns, and strengthen risk management practices across supplementary pension schemes. In addition, it aims to remove barriers that hinder efficient investment and to support the scaling up of pension providers, enabling them to benefit from economies of scale, improve asset allocation, and deliver greater value. In doing so, the reform seeks to raise standards across the supplementary pensions landscape, fostering greater value, increased resilience and long-term security for members.

A well-calibrated reform can help address persistent fragmentation, unlock more efficient and productive long-term investment, and ensure that a wider range of pension arrangements operate under sound governance and transparency standards.

Two basic option to achieve scale and performance are identified for IORPs – an option based on “moral suasion” and another based on “mandated scaling” approach. The two options are presented, in addition to the baseline scenario.

Option 1 – No policy change

This option would mean maintaining the current framework unchanged and relying on slow, organic growth to deliver scale and improvements over time. The underlying assumption is that time, market dynamics and national implementation will gradually lead to improvements in scale, governance and cross-border activity. However, this evolution is currently unfolding at a very slow pace, and not uniformly across Member States. In several countries, IORPs remain marginal or entirely absent, and the existing framework does not appear well adapted to support the development of occupational pensions in these contexts. While preserving regulatory stability, this option risks reinforcing fragmentation and limiting the role IORPs can play in addressing pension adequacy and long-term investment objectives at EU level.

Option 2: strengthening the prudent person principle

Option 2 would entail providing guidance to support the effective and consistent implementation of the prudent person principle referred to in Article 19 “Investment rules” of the IORP II Directive. The aim would be to promote a shift in the focus of IORPs investment policies from simplistic conservative asset allocations neglecting the return-leg of pension funds objectives to more balanced risk-return investment strategies that pursue long-term value for pension scheme members by diversifying across risks, including across asset classes, and that are guided by sound governance, and the knowledge and ability to manage the risks undertaken.

Scale is essential for achieving genuine diversification. Larger schemes are structurally better positioned to spread risks across a wide range of asset classes, including alternative asset classes, that require sufficient resources and expertise. The clarification of the prudent person principle should therefore stress that achieving adequate scale enhances the ability of IORPs to diversify effectively, manage risks holistically, and access a broader set of investment opportunities, while maintaining sound governance and risk management standards.

As the prudent person principle is a concept applicable to the entire supplementary pension sector and other retirement saving vehicles, a generalised guidance would have a wider outreach and benefit than limiting it solely to IORPs.

The prudent person principle governs how IORPs 125 should invest and manage their asset portfolios. This principle has been interpreted differently across Member States, which often limits pension schemes in diversifying investments and allocating to equities. According to Article 19 of the IORP II Directive, the prudent person principle dictates that assets must serve the long-term interests of savers, balancing return with security, quality, and liquidity. Portfolios should be adequately diversified to avoid excessive reliance on any single asset, issuer, or sector. In practice, sufficient scale greatly facilitates this diversification, enabling schemes to gain more balanced exposure across asset classes and to manage illiquid investments prudently. Investments in assets outside regulated markets are permitted but must remain at prudent levels. Environmental, social, and governance factors must also be integrated into the investment decisions. Those managing the schemes must possess the necessary expertise and maintain robust risk management systems.

The principle is a flexible yet demanding fiduciary standard, establishing a decision-making framework based on diversification, risk management, and suitability rather than a fixed asset mix. Scale is a key enabler of prudence, as it strengthens the capacity to diversify, manage complex risks, and pursue long-term investment strategies in line with the scheme’s objectives. Equity investments are not inherently imprudent; they can be crucial for a diversified, long-term pension portfolio as long as their risks are well understood and aligned with the scheme’s objectives. It would be good practice for managers to document compliance with the principle in the statement of investment-policy principles for schemes under the IORP II Directive. This document should outline investment objectives, diversification targets, exposure limits, and establish oversight, monitoring, and reporting mechanisms. It should also include rules for managing liquidity, quality, and conflicts of interest. The principle applies to both defined benefit (DB) and defined contribution (DC) schemes, focusing on matching assets and liabilities for DB schemes and prudent selection and monitoring of investment options for DC plans.

Option 3 – Enhanced supervision and moral suasion

Option 3 would entail amending the IORP Directive to reinforce supervisory oversight and governance standards while relying primarily on moral suasion, that is, the use of supervisory pressure and transparency, to encourage IORPs to improve in scale, efficiency and scheme quality, including through thematic reviews by supervisors focusing on several key aspects.

To complete the protection of members and beneficiaries of all pension institutions operating at their own risk, the Member State option to allow the extension of the application of the Directive is reinforced to explicitly allow extension to funded institutions that are excluded from the scope of Directive (EU) 2016/2341, in particular those that constitute an important component of the overall employment-related retirement landscape alongside IORPs, as well as institutions providing only personal pensions and not subject to other Union prudential legislation. For funded institutions operating pension schemes under Regulations (EC) No 883/2004 and (EC) No 987/2009, this would only be possible insofar as it is compatible with the obligations arising from those Regulations.

These regular reviews by supervisors would cover investment performance, the cost efficiency per participant, the fulfilment of indexation promises, and any curtailment of benefits due to performance or cost-related issues, while experiences and best practices will be shared within EIOPA. Furthermore, they would assess whether the investment strategy is properly aligned with obligations and evaluate the competence of trustees or boards overseeing these investments. The strengthened PPP would be part of the supervisory dialogue. In case of shortcomings remedial plans have to be agreed and acted upon, that could – where needed – consider potential means to scale up. Also, transfer rules would be simplified to make the IORPs framework more accessible, especially in Member States where it is currently underdeveloped.

Beyond moral suasion, this option would implement most of the recommendations from EIOPA’s advice and technical input.

In consequence, the core design features of Option 3 would be:

·A regular supervisory dialogue touching upon a series of early warning parameters, such as: insufficient levels of funding ratio, high costs per participant, backlog in indexation (i.e. shortfall in promised payouts), moderate returns, number of benefit cuts implemented, low number of scheme participants, closed funds, a fund with an aged population (so-called grey fund), declining number of (active) contributors, frequent or long-term vacancies at fund management or supervisory board, limited investment experience of directors. The frequency and type of supervisory review would be based on a risk-based approach, proportional to the risks and deficiencies identified.

·Notification or disclosure obligations in cases of sustained underperformance. This would ensure members are informed of risks to expected benefits and create pressure on providers to take corrective action (“accountability and disclosure”).

·National Competent Authorities (NCAs) to require a plan for remedial action, where fund boards need to indicate how they would manage identified vulnerabilities, e.g., where building scale is a solution, by recourse to asset pooling, appointment of multi-scheme asset managers or scheme mergers. Grant NCAs to set deadlines for remedial action – e.g., with respect to increasing low technical provisions or take measures to halt a drop in the number of participants.

·Failure to implement the action plan and address deficiencies could be followed by appropriate supervisory measures..

This option would also involve amending the IORP II Directive to implement most of EIOPA’s recommendations ( 126 ), with the objective of gradually improving investment practices, transparency and governance across supplementary pension schemes, as well as the optional application of the Directive to other funded pension schemes. The focus would be on strengthening risk management frameworks, improving disclosure on costs and returns, and supporting more sustainable and balanced investment strategies. At the same time, the reform would seek to address structural barriers that limit efficiency and scale, such as fragmentation, limited cooperation between schemes, and uneven supervisory capacity. While these measures would not transform the landscape immediately, they could over time enable IORPs to operate more efficiently, manage risks more effectively, and offer more stable outcomes for members. The amendments would also aim to adjust certain design features of the current framework to make it more adaptable and functional in Member States where IORPs remain underdeveloped.

Option 4 – Minimum scale and efficiency requirements

This option would retain all the elements of Option 3, including thematic reviews, soft nudges, improved supervisory practices, and the implementation of most of EIOPA’s recommendations. However, it would go further by introducing binding regulatory requirements and specific supervisory interventions to address persistent structural inefficiencies.

A key measure would be the establishment of a mandatory minimum efficient scale for IORPs. This could include providing templates for winding up or merging schemes in cases where the minimum efficient scale and performance are not achieved. Furthermore, it could involve the mandatory appointment of scheme trustees if existing vacancies are not promptly filled or if current officeholders lack the necessary skills required for effectively managing an investment portfolio strategy.

In consequence, the most salient features of Option 4 would be the following:

·Mandatory asset pooling requirement for IORPs below a certain threshold of assets under management, unless they can clearly demonstrate strong governance and value for members.

·NCAs to require all small schemes (defined as below a minimum efficient threshold of around 5 billion euro) with persistently low returns, high cost per member ratios, limited diversification, or governance failures to pool assets with other schemes, to join larger consolidated vehicles or transfer their portfolios to such larger vehicles.

·NCAs to draw up rules and guidance on “collective value transfer” (e.g., make liquidation easier for funds that cannot comply with governance, scale and/or performance requirements). NCAs would also establish simplified procedures, e.g. a digital notification form with explanatory notes as in the case of the Dutch National Bank, so that pension funds could report the collective value transfer and the liquidation of the pension fund easily and send the required documents to beneficiaries and regulators.

Ultimately, this approach would provide for strong supervisory powers to address underperformance and other inefficiencies and thus foster trust in occupational pension schemes by promoting stronger governance, higher transparency and more sustainable outcomes.

5.2.2.2 Fostering access to attractive pension products through the review of the Pan-European Pension Product (PEPP)

People that are well informed and overcome procrastination bias may still be discouraged from investing parts of their savings in pension products when they encounter high fees and low expected returns. Making pension products more attractive is therefore an important complement to incite engagement with the pension issue and foster a higher participation in pension savings. To ensure citizens’ trust in long-term investment products such as pension products, it is therefore central to both address the net risk/return ratio and to ensure the product provides a diversified asset allocation suitable for a multi-decade investment with an organised decumulation phase.

At the initiative of the Commission, the Union legislator established a legal framework for a Pan-European Pension Product (PEPP) with Regulation 2019/1238. This voluntary pension product could be offered across the EU Member States, complementing national pension products. The PEPP regulation ensured the standardisation of core features such as transparency requirements, investment rules, switching right and type of investment options. PEPP intended to deepen the single market for pension products by offering savers a wider choice and facilitating the cross-border distribution of pension products. While the European nature of PEPP should support the needs of people that change residence between Member States, it was also expected that the offer of the PEPP as alternative to national supplementary pension products would stir competition towards more attractive national pension products at lower fees. The PEPP regulation entered into force in March 2022.

As it became evident from the Commission’s public consultations, the current PEPP has not yet lived up to its potential. Uptake remains anaemic, with only two providers in the whole EU offering the PEPP, which is, clearly, far below expectations and therewith PEPP has also failed in triggering more competition on national pension markets towards more attractive terms for savers. Most participants of the public consultation mentioned that this outcome is due to several factors of the PEPP design. These include the high cost and complexity for providers to offer the product due to the embedded features such as a mandatory advice component, providing an advised offer across multiple countries, a fee cap of 1% and unfavourable and varied national tax treatments for different personal pension products in Member States despite a recommendation 127 to offer a similar tax treatment in 2017 (see Annex 5 on the national tax treatment for PEPP), as well as delays in the implementation of the PEPP Regulation by Member States. Additionally, some consultation reactions indicated that the PEPP is deemed as unwelcome competition to incumbent providers of private pensions, that are offered at fees that are often very high on account of their use of traditional (physical) distribution channels. In other words, offering PEPP with its low profit margin would impinge on the high profit margins generated by other products.

The Commission’s 2025 targeted consultation on pensions reform as well as EIOPA’s 2023 Technical Advice on PEPP Market Development indicate that the main obstacles to PEPP uptake are: (i) lack of cross-border scalability; (ii) lack of immediately recognisable product features; (iii) and limited retail investor familiarity with the concept of a PEPP. Stakeholders agreed that PEPP Regulation (EU) 2019/1238 pursued the goal of promoting cross-border personal pension saving through a harmonised product. However, they observed that its implementation has been hindered by:  low provider uptake largely due to the complex design features prescribed in the PEPP Regulation and its delegated rules, combined with the fee cap for the basic PEPP; as well as the lack of interoperability with existing investment product and fund structures, while also pointing to fragmented national tax treatment.

Policy Objectives of the PEPP review

The aim of the review is to create a framework for a personal pension product that offers opportunities for solid retirement savings and responds to diversified users’ expectations. The revision of the framework should allow for increased uptake, choice and competition. To achieve this, in addition to the baseline scenario, three options are presented.

Option 1: Baseline scenario

Under Option 1, the Commission would not take any immediate action to amend or accelerate the review of the PEPP Regulation. Instead, it would follow the original timeline established by Article 73 of the PEPP Regulation. The rationale would be that PEPP is still in its early stages.

Option 2: PEPP with flexible design features while keeping the fee cap

Option 2 is to leave the entire design features and distribution modalities for the Basic PEPP to the discretion of the PEPP provider. The dual structure, which allows for offering both a basic version and tailored versions of the PEPP, would be removed. The only regulatory constraint is that the emerging product can be sold within a modified (ex VAT) fee cap of 1% ( 128 ). It is in line with the observation that the current 100 basis point fee cap is not impossible to comply with and had no hindered the emergence of two cost efficient PEPP providers.

Option 3: Basic PEPP without fee cap

Option 3 is based on the premise that the fee cap is removed as it was suggested by most participants in the targeted consultation to be the principal feature that inhibits the more wide-spread take-up of the basic PEPP (“lack of cross-border scalability”). Unlike for Option 2, the dual structure of the PEPP would be maintained with the possibility to offer variants only if the basic PEPP is offered. This Option 3 entails a basic PEPP that is designed to remain affordable, accessible, and simple for savers across the EU. To ensure cost efficiency, the Basic PEPP may not provide any capital guarantees and must be suitable for sales without additional investment advice because it is suitable by design, despite the lifting of the fee cap. In the absence of the (cruder) tool of a fee cap, Option 3 will have to be more specific in the design features, so that the basic PEPP remains affordable accessible and simple for pension savers. Therefore, Option 3 will introduce additional design features: a basic PEPP needs to contain a simple lifecycle glidepath to ensure that asset allocation evolves towards more liquid and safer asset classes as the pension saver approaches retirement. To remain cost-efficient, the Basic PEPP needs to accomplish its lifecycle glidepath by investing in cost-efficient underlying, such as listed shares, bonds, non-complex UCITS funds or money market instruments. Tailored PEPPs would present additional features and more sophisticated investment strategies, but like the Basic PEPP, they would also be subject to the value-for-money framework and included in EIOPA’s register. 

In the initial phases of the glidepath, a higher allocation could be made to growth shares, gradually evolving towards more value (dividend) shares and increased allocation to fixed income assets as retirement and decumulation approach. A small pocket of unlisted assets is possible and would suit particularly the initial stages of the lifecycle but must not exceed 5% of the asset allocation.. The main features of the basic PEPP in Option 3 would be: no embedded capital guarantees, no obligation to provide the basic PEPP with advice due to its inbuilt features, no fee cap and instead a value for money assessment. The PEPP would be eligible for auto-enrolment schemes that benefit from employer contribution and suppliers have the choice to define the type of pay-out phase in agreement with the client under national law where applicable

Option 4: Basic PEPP without fee cap complemented by an autonomous tailored PEPP

This option is also based on the assumption that the fee cap for the Basic PEPP is removed. But, contrary to Option 3, this option would allow providers to offer PEPP with different options, without an obligation to also offer the basic PEPP. Option 4 acknowledges that people have different preferences and that a more sophisticated asset allocation (beyond mere life cycling) can provide additional return for the pension portfolio. This additional tailored PEPP option is likely to come at a higher cost and, based on the particular structure chosen, at either higher risk due to the asset allocation chosen or a lower risk if guarantees are added.

Option 4 rests on the premise that more variations in the design of a PEPP would on average result in better gross returns. It also acknowledges that people value security and are often willing to pay for guarantees ( 129 ), which this option allows the provider to offer as part of a tailored PEPP. More complex variations offered through the tailored PEPP would also require at least some basic form of investment advice to ensure that the potential subscribers understand the more complex product features and options and, for example, the implications of capital guarantees on the expected return for a multi decade product. Option 4 would equally require a value-for-money to ensure supervisors can at least with hindsight check that costs stay under a reasonable control. Apart from the features embedded in option 3, the main “value-added” of the PEPP in Option 4 would be that savers would have a choice that would correspond to their preferences – either a simple cost-efficient personal pension products (basic PEPP) or a more sophisticated one reflecting their potential preferences (tailored PEPP).

5.3 Options discarded at an early stage

Legislation on the use of and good practices in auto-enrolment, the establishment of an EU wide pension tracking systems and private pension providers delivery of data for a pension dashboard were dismissed early on. Harmonised rules on auto-enrolment and pension tracking systems look inconsistent with different starting positions of Member States’ pension systems. Since participation in occupational pensions is strongly driven by domestic factors, harmonised rules would undermine their effectiveness in fostering supplementary pension savings. Most information from pension tracking systems would be for domestic use, the ETS is the instrument to ensure inter-operability of national tracking systems across Member States. The imposition of harmonised reporting obligations to feed pension dashboards be less effective and probably burdensome in the current context of widely diversified pension systems. Already the report of the CMU High Level Forum advocated that the EU encourages and supports developments in the Member States and discussions with the Member States and consultation with stakeholders revealed that binding legal instruments are no feasible policy instruments in this area.

The option of dismissing the legislative framework for PEPP in view of its low take up was also discarded. Discussions with existing and potential providers helped identify parameters that could make PEPP an attractive pension product.

6.    What are the impacts of the policy options?

6.1 Recommendation on pension tracking, pension dashboards and auto-enrolment

The impact of a recommendation will depend on how Member States make use of them. The effect will be different across Member States depending on whether they take an enabling or a promoting approach, how they use the elements incorporated in the promotion approach. Furthermore, differences in starting positions and need to adjust to country-specific circumstances will make the impact different across Member States. In this context, a recommendation will not have an effect if the aspects in scope are already covered and well developed in a Member State.

While the primary objective of the measures discussed below is an increase in the rates of participation and saving in supplementary pensions, the social effect depends on whether vulnerable population groups can also benefit from its introduction. Vulnerable groups would benefit at least indirectly since a stronger supplementary sector creates room for manoeuvre for public pensions to fulfil its broad obligations, including its redistributive role and providing adequate support for those requiring non-contributory benefits. The direct effect depends primarily on the design of the eligibility conditions and incentives provided, which will be at the discretion of the Member States rather than on the different options presented in the previous chapter. Despite this limitation to quantify the social effects, the more people, including lower-income groups, participate in supplementary pensions, the more favourable the effects. Such effects can however not be taken as given but will depend on the design of the measures (see Annex 2 for a more detailed discussion).

6.1.1 Pension tracking

Since information and transparency are a precondition for saving decisions, pension tracking systems have a strong enabling character to support retail savers in identifying their pension claims and possible gaps and act upon them. The analysis in Annex 3.2 summarises the main transmission chains and their empirical support that underpins the comparison of the different option’s impact in this section.

With regard to the first, the status quo option without further supportive EU action, it can be expected that the developments of national pension tracking services will remain relatively slow and sporadic, with a high probability of non-compatible solutions, hindering access to pension information, pension transparency and financial planning both at national and cross-border level, and thus undermining pension adequacy for individuals .

The second option consists of a recommendation addressed to Member States to develop national pension tracking services, also outlining recommended design features of such services, building on the technical advice prepared by EIOPA. Such a recommendation would provide a clear political signal at EU level on the importance of pension tracking, reinforcing the demand for tracking services. At the same time, it will provide guidance with regard to the design and set-up of tracking services, thus contributing to the quality and compatibility of the services put in place, including compatibility with the ETS to foster the development of cross-border pension tracking in the EU. Beyond inviting Member States to develop new tracking services, it will also guide them to increase the coverage and improve the design of existing services, while leaving room for necessary flexibility in the view of diversity of national pension landscapes.

Somewhat encouragingly, at least in the area of social protection, there is a strong track record of recommendations backed up by support and monitoring mechanisms shaping policies at national level. E.g., within three years of the adoption of the Recommendation on access to social protection, 15 Member States had taken steps to improve participation to social protection schemes for specific categories of workers and the self-employed ( 130 ); following the adoption of the Recommendation on long-term care in 2022, all Member States have appointed national coordinators and submitted implementation plans that show ambitious policy measures and investments ( 131 ).

In consequence, a recommendation has the potential to boost the development of high-quality and comprehensive national pension tracking services, in particular if social partners and stakeholders are duly involved, and financial and technical support is available from EU programmes (e.g. the German pension tracking service was developed with RRF support ( 132 )).

The third option, imposing a legal obligation on pension institutions to offer pension tracking services, would allow pension savers receiving information on their supplementary pensions in aggregated format and in a more harmonised format than in the currently varying formats of pension benefit statements. Some standardisation of the information may even foster IORPs and PEPPs delivering pension benefit statements to their members more efficiently. It, however, appears problematic both from legal and policy perspective to apply it to other types of pension providers. A comprehensive pension tracking system, which provides individuals with a complete overview of their pension rights, by definition covers both statutory and supplementary (occupational and personal) pension schemes, which are subject to different EU legal frameworks. Statutory pension schemes are considered part of statutory social security in the understanding of the Regulation 883/2004 ( 133 ), whereas some occupational pension schemes are covered by the Directive (EU) 2016/234 ( 134 ) and some personal schemes by Regulation (EU) 2019/1238 ( 135 ).

6.1.2 Pension dashboards

The creation of a pension dashboard as monitoring tool would not have a direct impact on the performance of national pension systems. The intended direct effect would be to incentivise Member States to undertake pension reforms, which will ultimately lead to a better performance in terms of fulfilling the adequacy and sustainability objectives.

In Option 1, Member States would continue to base policy decisions on pension reform on the currently available statistical information and ad hoc analysis. They can use the data that is already compiled for all EU Member States for the Ageing Report and the Pension Adequacy report and the various OECD pension monitors to assess the performance of their pension system relative to peers. These data sets are very detailed and granular, therefore more useful for policy analysis than as a tool to inform about aggregate pension gaps and to create peer pressure on Member States to act.

Option 2 could entice them to cater for a regular and forward-looking approach that gives due prominence to supplementary pensions. It could also initiate the development of evaluation tools at national level ( 136 ). The national dashboard would support the design of appropriate policy measures. The enhanced transparency could also lead to greater awareness in the population about the need to cater for their retirement income, i.e. a tool reinforcing the effectiveness of pension tracking.

A recommendation to create a national pension dashboard that includes supplementary pensions could particularly entice those Member States that do not provide projections about supplementary pensions to the Ageing Report. Eleven Member States already delivered such data for the Ageing Report ( 137 ), which suggests that political recognition of the importance of supplementary pensions is an essential driver for their statistical coverage. It also suggests that it is not impossible to do so. Among those that have not projected supplementary pension expenditure, seven linked it to data and six to resource issues in their input to the EIOPA advice. Further four referred to complexity and three that it was not necessary to do projections on supplementary pensions since these do not endanger the sustainability and adequacy of public pensions.

National pension dashboards under Option 2 would be tailored to Member States’ starting positions in terms of data availability and policy objectives. They could stimulate public debate about pension reform and in conjunction with pension tracking systems incentivise individuals to care about their retirement savings. However, they would not allow comparisons of pension systems’ performance across Member States.

Since this option gives discretion to Member States in which and how they compile data and undertake the projections, the data will not be comparable across Member States. The non-comparability effect could be reduced if Member States were transparent about which pension schemes are included and excluded and if they agreed for the purpose of the projections on a set of common definitions and macroeconomic assumptions.

Option 3 would accomplish comparability of pension indicators across Member States and a dashboard could become a tool for discussion among Member States about good practices and relevant experiences. The pension index compiled by Mercer, comes close to a tool that can deliver pointed messages via a consistent ranking across the three dimensions of adequacy, sustainability and integrity covered in the index (see Table 9 in Section 3.2). However, this index is not entirely data driven as it combines statistical data with the assessment of national pension experts ( 138 ).

This option would impose on the Commission to establish a tool for pension monitoring. The recommendation would invite Member States to submit data for that purpose ( 139 ). Involvement of Member States appears crucial as they are best placed to identify relevant data sources. They could thus be recommended to identify suitable statistics and indicators and to share them. Mandating pension supervisors and EIOPA to collect and centralise the data would imply economies of scope as they already possess a lot of the data on supplementary pensions. They may, however, face barriers to receive data ( 140 ), which would need to be addressed.

However, already the first step of establishing a joint classification will be difficult given the stark differences in Member States pension systems ( 141 ). It will delay the production of such a European pension dashboard. This is a major disadvantage of Option 3. In consequence, it will also be challenging to define and impose a harmonised approach on pension products and assets for which data will need to be collected.

Respondents to the targeted consultation considered the inclusion of supplementary pensions and a forward-looking approach the most important qualities for a dashboard. They attributed less priority to other qualities such as a granular breakdown or cross-country comparability. In relation to the importance of dimensions to be covered, pension adequacy, i.e. income replacement was quoted as the most important dimension that dashboards should cover, followed by sustainability. Other dimensions were reported of less importance (see Table below). Submissions by public institutions attributed a high priority to cross-country consistency, unlike other stakeholder groups.

Table 131: Consultation replies about pension dashboards (ranking with 1 the most important dimension and 5 the least important one)

Notes (1) Detailed data about occupational and personal pensions, in addition to statutory pension, (2) pension data by different cohorts of the population (e.g. by gender, age, type of employment, economic sector, income, etc.), (3) projection of pension adequacy and sustainability, based on transparent and robust assumptions,(4) consistent data and methodology across Member States to allow for comparisons the level of retirement income relative to work income now or in the future (5) other included avoidance of reporting burden, (6) the level of retirement income relative to work income now or in the future, (7) measured by its capacity to deliver a decent level of retirement income in the next decades in face of a declining working age population, ( 8) others mentioned were tax treatment, contributions and that the ranking would depend on the target audience.

6.1.3 The possible impact of auto-enrolment

The concept of auto-enrolment in pension schemes is designed to help people save for retirement by automatically signing them up unless they choose to opt out. This approach addresses a widespread tendency to delay saving, which often leads to inadequate retirement funds. Countries that introduced auto-enrolment have seen increases in participation rates and the more detailed analysis in Annex 3.1 details good practices and suggests the magnitude of the increase in the participation rate depends importantly on the design features. The use of good practices would therewith be a key parameter of the impact.

The replies to the targeted consultation revealed that stakeholders consider the creation of a legal framework the most important role for the public sector (see table below). Tax incentives or public subsidies for the population follow at second place. Respondents ranked the other asked public tasks at a comparable level, with somewhat less priority allocated to tax incentives for employers and with administrative support (Option 4) ranking at lowest priority.

The status quo - Option 1 implies that Member States that consider auto-enrolment as part of pension reform will do so. They can make use of the good practices identified in the AE study and complement it with experiences made in the meantime in other Member States, for example Ireland. The observation that the release of the AE study did not trigger debates on auto-enrolment in most Member States suggests that not many other Member States will use this tool to increase the participation in supplementary pensions. The Commission’s use of soft tools such as speeches, participation in conferences, interaction with civil society is unlikely to create sufficient political momentum.

The recommendation on auto-enrolment aims to create political momentum to support its introduction and create the appropriate preconditions. The enabling approach in Option 2 can foster public debate and facilitate the introduction of an auto-enrolment system. In particular, it will raise awareness on the need to promote supplementary pensions and create the relevant preconditions for that purpose. Sufficiently strong supervision as economic preconditions should already exist in most Member States and the other parts of the pension package analysed in this document aim to ensure that auto-enrolled savers earn a decent return on their contributions. Access to IORPs, as well-regulated financial vehicles for pension saving, is accomplished through the IORP II Directive. Its review and the clarification of the prudent person principle will improve IORPs’ capacity to generate decent returns for retirement savers. The PEPP review has also the objective of allowing access to profitable investment possibilities, either via the use of PEPPs themselves or via the competitive pressure they will exert on underperforming national pension products.

The comparison of experiences with auto-enrolment from other countries suggest that the impact on participation rates is limited if Member States do not accompany them with a good design of the measure. A further disadvantage is that even in the simplest approach employers are concerned about the possible administrative burden and costs, as well as risks to established occupational pension practices – especially those organised by social partners. These concerns were frequently raised in the responses by business associations to the Commission’s consultation activities in preparing this pensions package. The replies to the consultation also suggest that many other stakeholders value and advocate the use of existing good practices. Such stakeholders may be disappointed if these good practices do not feature prominently in either the Commission recommendation or in public debates in the Member States.

The advantage of a recommendation that includes good practices, which is Option 3, would be a higher participation rate compared to the enabling approach. Apart from best practices improving the effectiveness of auto-enrolment and reducing the opt-out rate, the inclusion of good practices in a recommendation would be a richer public debate that might lead to a higher likelihood that Member States use those practices that make auto-enrolment most effective.

The specific use of good practices would however still depend on the conditions in the Member States, which are very different. The recommendation could therefore provide principles for the good practices to guide Member States to adjust them to national needs. Thus, the nature of a recommendation ensures that Member States would still have a choice whether and which good practices they want to apply. The table below summarises the responses of stakeholders of what they consider the prioritisation of good practices. To the extent the inclusion of good practices translates into a more granular discussion, awareness in the population about the importance of supplementary pensions will increase and individuals will be more likely to understand and take action.

Table 142: Hierarchy of best practices along their importance, stakeholder replies to the targeted consultation

Note: Average measured as priority on a scale 1 to 7 with one the most and 8 the least important, frequency measured as how often referred to in the 112 responses to the consultation. * Under “others”, some stakeholders referred to the design of default options.

As possible disadvantage of Option 3, the inclusion of good practices, is likely to lead to public debates in Member States about the introduction of auto-enrolment becoming complex. With a view of the tendency of political debates increasingly carried out via social media, the complexity of good practices implies a risk of overburdening stakeholders and the use of simplified messaging. This can, for example, be the case around the question of employers’ contributions, their administrative burden and the magnitude of countervailing tax deductions. Employers co-financing is a good practice that translates into tax shortfalls for the public budget ( 142 ). It can be anticipated that the role of tax incentives, highlighted by many stakeholders in the consultation as key to the take up of supplementary pensions, will dominate the public debate and that fiscal costs will have a key influence on Member States’ willingness to follow the recommendation.

The use of a public body in Option 4 to administer pension savings and organise the investment, either directly or in form of a master trust, would reduce administrative costs and enable greater control over the fees charged by financial intermediaries. A recent survey documents that reducing administrative costs in the management of occupational pensions is a key concern of employers ( 143 ). It would also provide a solution to the portability issue that arises if individuals change their employer. These features would be most beneficial in Member States with many small firms and without a role of social partners in organising occupational pension schemes. Examples of asset-backed (defined benefits) public pension funds indicate that state-sponsored pensions asset management may not only be efficient but also effective in providing long-term value and defuse sustainability pressures (see box below).

Making it mandatory for employers to auto-enrol their employees would boost the participation rate. It could also reduce the need for the public sector to create complementing fiscal incentives. Depending on the role of employers’ co-financing that is part of the good practices in Option 3 in a mandatory regime, tax losses for the public might be larger or lower. An obligation to offer occupational pensions may however undermine the interest of employers to make this offer voluntarily as a means to attract, maintain and engage their workforce. In Member States where terms on occupational pensions are part of collective wage agreements between social partners, an obligation to offer occupational pensions could interfere with the institutional set up for wage bargaining.

A possible downside of Option 4 could be that the establishment of a public body competes and potentially crowds out the activity of private entities. Incumbent IORPs could see their market share challenged. Moreover, the costs for the public sector in organising supplementary pensions will be an additional factor in the public discussion, discouraging Member States with limited budgetary room for manoeuvre (Annex 2.1). It could also interfere in Member States where social partners or other private entities already organise supplementary pensions, depending on design. A key disadvantage is the risk that many households will understand the introduction of auto-enrolment as an enlargement of the scope of the public pension sector. If they do so, they may not discriminate between their mandatory contributions to public pensions as “forced savings” and their contributions to supplementary pensions as “optimising their returns on savings”. This may have an adverse impact on their perception of their own tax burden, their expectation that their supplementary pensions will be invested in their interest and on their supply of labour. This risk may be highest in Member States where trust in public institutions of providing a clear separation between public and private activity is low.

6.2 IORP Review

The IORP options ultimately aim to increase the benefits savers can draw from occupational pension schemes in terms of enabling IORPs to generate higher returns. The options differ between an indirect approach of fostering scale economies through a supervisory environment that is more conducive to consolidation and asset pooling (options 2 and3) or through enabling IORPs to implement more return-oriented investment strategies.

The main advantage of Option 1, which involves no policy change, is that it avoids regulatory disruption and preserves the current balance between EU and national competences. It imposes no new compliance burdens on IORPs or supervisors. However, leaving the current framework unchanged would likely perpetuate inefficiencies, and uneven levels of member protection across Member States. This inaction would limit the sector’s contribution to financing investment needs and hamper meaningful progress toward a more integrated and efficient European pension landscape.

The guidance on the prudent person principle of Option 2 should inspire pension assets investment management, contributing to foster virtuous practices that pursue long-term value and return for schemes participants while effectively managing the risks according to the stated risk appetite. Since the prudent person principle significantly influences asset management and investment strategies for IORPs, encouraging IORPs to avoid excessive reliance on any single asset, issuer, or sector would impact investment diversification. Additional guidance on the principle would mitigate risk and enhance portfolio stability, thereby allowing pension schemes to allocate resources more effectively across various asset classes, including equities and alternative investments. Consequently, there is potential for higher returns within robust risk parameters ( 156 ). Thus, the risk of significant financial losses is reduced, while the long-term sustainability of pension schemes is enhanced.

Reiterating that the duty of care of the IORPs’ boards towards members and participants envisaged by the prudent person principle is not limited to the preservation of the capital but also to providing adequate net returns on pension savings would nudge small and inefficient IORPs to pool assets or consolidate in larger funds that would be able to access to and diversify across a wider range of asset classes, including those with lower liquidity and longer investment horizons that can enhance return, spread costs more efficiently, and leverage on more robust risk management. The guidance on the principle would thus foster greater accountability and transparency, oversight, monitoring, and conflict management, ensuring alignment with stated objectives and regulatory standards. It provides stakeholders with clearer insights into investment strategies, thereby fostering trust and confidence in the management of pension funds. Overall, strengthening the use of the prudent person principle as fiduciary standard would enhance diversification, improve risk management, and promote accountability within IORPs. It would ensure pension schemes are better equipped to serve the long-term interests of savers by balancing the need for security with opportunities for growth.

Option 3 would enhance the IORP framework through supervisory improvements, soft regulatory nudges and targeted amendments. The coverage of the prudent person principle in supervisory dialogues would increase its impact compared to an unsupervised guidance. Linking the discussion of the prudent person principle with scale and other possible causes of underperformance could make the principle more meaningful. Managers of pension funds might be more inclined to seek for asset pooling or consolidation if they understand that scale is an important determinant of their capacity to generate higher returns for their members and to manage the associated risks.

The introduction of thematic reviews would encourage better risk management, investment alignment, and governance practices without imposing rigid new obligations. It would align with the “social partnership” approach that underpins the successful occupational pension schemes in many Member States. This option would also make the framework more accessible, particularly in Member States where IORPs are underdeveloped, and would also allow Member States to extend the application of the Directive, on a voluntary basis, to other funded institutions operating pension schemes at their own risk. By implementing most of EIOPA’s recommendations, this option would promote greater consistency across national systems. Regulatory dialogue would further enhance governance, ensuring that investment decisions are made by qualified professionals in line with the interests of members and beneficiaries.

However, because it relies primarily on voluntary improvements and supervisory pressure, the impact on achieving economies of scale might take more time and persuasion when compared to the more incisive tools deployed in Option 4. Fragmentation and the persistence of small, subscale schemes may remain a structural issue for a somewhat longer period, especially in markets lacking strong political or economic incentives to expand occupational pensions.

Option 4 would force a more decisive structural reform, albeit at the price of a more “top down” approach involving binding regulatory tools and supervisory powers aimed at achieving critical scale and improving governance. Requiring a minimum efficient scale could promote consolidation where needed, helping to reduce inefficiencies, lower costs per members, and allow greater diversification across risks and asset classes, thereby improving the return-risk profile of occupational pension schemes.

On the other hand, binding regulatory requirements and specific supervisory interventions to address persistent structural inefficiencies necessitates that these inefficiencies can be identified upfront and translated in clear legal requirements. It also requires a sound justification for the events and threshold that initiate supervisory action. Given the variety of possible situations and country-specific environments, this is a challenging task. The specific dimension of consolidation as means to address structural inefficiencies is also likely to create tensions with the objective of securing competition, which often means the avoidance of market concentration. Option 4 could face strong stakeholder resistance in its implementation and there is a non-negligeable risk that the immediate benefits of a more incisive option would be more than neutralized by resistance.

6.3 PEPP Regulation review

Since consumers have only been offered PEPP for a short time and in very few Member States, there is little information about their interest in the product. Consumer association Better Finance highlighted the relatively high returns the first PEPP on the market has generated, which suggests that PEPP has a potential to become a pension product that savers find attractive.

The discussion of the impact of the options below is therefore based on discussions and consultations with all stakeholders and Commission staff analysis of possible effects.

Option 1 would be to maintain the PEPP Regulation unchanged. This scenario would only be viable if early signs had shown clear, organic growth. However, the current 1% fee cap, when combined with the other mandatory features, has rendered the product unattractive for most providers. While the Commission could still modify the amount of the cap or adjust the risk-mitigation techniques through technical standards (level 2 measures), other fundamental issues would remain and would require amendments of the PEPP regulation (level 1). Keeping the current framework unchanged would likely mean that the PEPP remains a niche, symbolic product, rather than a meaningful contribution to closing the EU’s pension savings gap.

Option 2 would grant PEPP providers greater freedom in their design and distribution strategies for the basic PEPP. The choice between life-cycling or other risk-mitigation strategies, such as guarantees will appeal to different types of providers. Providers could differentiate their offerings, potentially resulting in a wider array of innovative and cost-effective products. Such diversity in product offerings enhances consumer choice, catering to the varied retirement needs and preferences of EU citizens. As the policy retains a fee cap for the basic PEPP, it offers a level of consumer protection against excessive charges, reinforcing trust within the PEPP market. By maintaining the 100-basis point fee cap and removing mandatory product features, such as the advice requirement for the Basic PEPP, Option 2 would most likely address the above issue of avoiding complex design features.

However, as the consultation has shown, the 100-basis point fee cap is seen as a major blocking factor to many stakeholders. Many stakeholders perceive the fee cap as excluding many additional potential PEPP providers, such as the insurance and asset management sectors. In practice, the fee-capped Basic PEPP would become an online product offered by neo-brokers that assemble a simple, off-the-shelf lifecycle strategy at very low cost. Option 2 might therefore not sufficiently address the identified problem of low provider uptake. At the same time, for such a simple product the current 1% fee cap, in particular if its composition would be further adjusted , would also not necessarily lead to better outcomes for savers, nor would it ensure a proportional relationship between benefits and returns to savers and expenses borne.

As a result, stakeholders advocate for the removal of the fee cap for the Basic PEPP as an impediment to the emergence of more scalable pathways to a personal pension product. The rationale behind this preference lies in its potential to be more inclusive, accommodating a broader range of prospective suppliers and investment strategies that underpin the PEPP. Option 2 would therefore also not address the problem of limited consumer recognition.

The removal of the fee cap is also portrayed as enhancing consumer choice and accessibility to personal pension products across Europe. More inclusivity in terms of providers and investment strategies to underpin a personal pension product are perceived as a vital component in achieving the overarching goals of the PEPP initiative, which include promoting retirement savings and ensuring financial security for individuals throughout the European Union. By designing a basic PEPP that is essentially cost-constrained to be distributed online-only, Option 2 would also not address the issue of lack of interoperability with existing product and fund infrastructures. Overall, therefore, it is not likely that this option would significantly address the pension gap.  

Option 3 therefore proposes the removal of the fee cap for the basic PEPP, which has been a request by market participants and is also acceptable to consumer organisations if it is accompanied with sufficient other cost controlling features. It therefore has the potential to alter the competitive dynamics in the market. Without the constraint of a fee cap, providers may innovate more freely and optimise their cost structures from various angles. This is expected to result in increased competition as providers strive to deliver quality, potentially resulting in improved features for consumers. The cost of physical distribution is known to be a major cost driver for any retail product. To keep costs under control, the basic PEPP would embrace digital innovation by favouring distribution via online channels under IDD or as an execution-only (retirement-suitable) product sold online under MiFID. The tailored PEPP also under option 3 keeps the advice requirement.

Furthermore, PEPP savers should receive clear information on the pay-out options prior to retirement (lump sum, organised withdrawals or annuities or a combination).

Option 3 for the basic PEPP would leverage life-cycling investment strategies targeting mainly “execution-only investment products, as described in the MiFID framework. The basic PEPP would be also allowed to invest a limited amount (maximum 5% of its portfolio) into unlisted assets to take advantage of the possibility of better returns. By mimicking the “execution-only” investment propositions, Option 3 introduces an execution-only basic PEPP which would have instant brand recognition. The basic PEPP aims to merge the “pension suitability” features of the PEPP Regulation with the operational success of non-complex products that are currently sold on an execution-only basis under MiFID, Article 24. These design features would avoid the above issue of overly complex design features.

Option 3 would therefore be designed to cover a broad scale of investable instruments, efficient portfolio management, depositary oversight, and cross-border distribution while maintaining the pension purpose and consumer protections of the existing PEPP Regulation. By doing so, Option 3 seeks to encourage additional potential providers of a PEPP to enter the PEPP market using existing investment products or investment fund ranges. They can embrace existing and successful life cycle strategies as investments underpinning the PEPP. This would address the above issues of low provider uptake and lack of interoperability with existing product and fund infrastructures.

The basic PEPP is designed as an execution-only product. While it is frequently argued that the elderly need in-person advice, the PEPP is a product catering to the population under ca 45 years. This group of investors is generally digitally savvy and able to navigate e.g. comparison website or to purchase products online. The life cycling glide path towards the decumulation phase allows the consumer to first profit from multi-decade gains in the capital markets with broad diversification, before reducing capital market exposure in a controlled manner.

On the other hand, the basic PEPP under Option 3 cannot include guarantees, as well as deviations from the most basic form of lifecycle glidepath or invest no more than 5% of unlisted assets in its portfolio. However, its features that may appeal to a large part of savers.

The alternative Option 4 maintains the duality of the PEPP but disconnects the provision of tailored PEPP with options from an obligation to offer the basic PEPP. This option comes with the advantage that a certain share of private market exposure can be added to the portfolio, even beyond the early “growth face” of the product’s lifecycle. It also allows for other risk-management strategies beyond life-cycling, such as guarantees or other strategies adequate for retirement savings. This fosters innovation and broadens diversification but also introduces less familiar asset classes (and less predictable risk portfolios) into the accumulation portfolio.

The choice of the preferred option would therefore hinge on whether more consumer acceptance would be achieved by relying on a trusted retail brand or by pointing out the diversification effects that can be achieved for instance (i) with a portfolio that contains less liquid asset classes in the initial stages of the” lifecycle” or (ii) a more complex investment strategy that uses derivatives to stabilise or even guarantee investment outcomes.

The tailored alternative PEPP allowed along the basic PEPP under Option 4 would be characterised by more sophisticated options (investment and/or risk management) and likely by higher costs. In other words, while the design features of the Basic PEPP are driven by the aim to reconcile per se suitability for retirement savings with choosing the lowest cost alternative on all the salient design features, the tailored PEPP is more permissive in terms of product features that engender higher cost (in the hope that more complex investment strategies achieve either higher return or stable returns with lower volatility). Value for money assessments would need to be ensured for both types of PEPP, which are expected to compete.

In consequence, Option 4 would allow competition between a basic PEPP with a simple lifecycle glide path and a tailored PEPP with the option for more exposure to illiquid (riskier) assets for a longer period in a lifecycle or stabilized return profiles by means of guarantees or derivatives used as tools to manage the portfolio more efficiently. While these latter investment strategies might enhance return or reduce risk (volatility of performance), they invariably come at a cost.

7. How do the options compare?

This section compares the options analysed above against three key criteria: (1) effectiveness in accomplishing the objective, (2) efficiency in accomplishing them at low costs and (3) cohesion defined as covering spill over to other policy objectives.

7.1 Pension Tracking Systems (PTS)

With regards to pension tracking, an assessment of PTS Option 1 – status quo/baseline scenario- relying strictly on uncoordinated national initiatives may be insufficient to achieve the desired policy objective. The current fragmented landscape and the limited and sporadic progress to date in terms of pension tracking underscores the need for a clear EU-level impetus to close the existing pension information gaps for individuals. Conversely, while PTS Option 3 - a binding legal instrument for pension institutions - would provide the highest guarantee of achieving legal certainty and a harmonised EU-level action, it does not appear feasible to apply it to all types of pension institutions, given their different legal frameworks in the EU.

Since the status quo/baseline scenario (PTS Option 1) relies on uncoordinated national initiatives, it is not very effective in achieving the desired policy objectives of ensuring adequate retirement income through higher participation and savings in supplementary pension schemes or products. Member States would neither face an increased incentive nor guidance through good practices to prepare pension tracking systems. The current fragmented landscape and the limited and sporadic progress to date in terms of pension tracking suggests that EU-level impetus (PTS Option 2) would be more effective to close the existing pension information gaps for individuals. A binding legal instrument (PTS Option 3) is the most effective option since it provides the highest guarantee of achieving legal certainty and a harmonised EU-level action.

The nature of a recommendation leaves Member States discretion on how they follow up. A recommendation would not increase the costs Member States that would create a PTS without a recommendation (PTS Option 1), nor would be expected that the use of good practices reduces the efficiency of the measure. Those Member States that do not consider creating a PTS would face higher costs if they follow a recommendation compared to keeping the status quo. However, inaction will not yield benefits, so that the option of a recommendation (PTS Option 2) is more efficient than the status quo (PTS Option 1). The legal instrument (PTS Option 3) implies reporting obligations for pension providers, which would need to match the existing obligations for IORPs and PEPPs to avoid creating an additional burden. The reporting costs would be hard to justify if pension funds are under the legal obligation but there is no pension tracking system in a Member State that will aggregate the information and make it available to pension savers.

Keeping the status quo on pension information (PTS Option 1) will not help strengthen the pension sectors’ contribution to the EU objectives, especially the objective to make the EU economy more productive and competitive. A recommendation that Member States introduce PTS using the good practices (PTS Option 2) would be therefore more coherent with the EU objectives. While the legal instrument (PTS Option 3) might also be more effective in supporting other EU objectives, it could not be applied to all pension providers. Member States’ pension systems, often comprising statutory, occupational and personal pillars are deeply embedded in national social, labour and fiscal regulatory traditions. PTS Option 3 may raise significant concerns regarding proportionality and subsidiarity, if applied to all pension providers.

Table 153: PTS policy options

Against these considerations of effectiveness efficiency and coherence, combining PTS Option 2, a Commission Recommendation, and PTS Option 3, legal obligation to some categories of pension providers, emerges as the most proportionate course of action. Building on the technical advice prepared by EIOPA, this approach would provide the necessary political momentum, by outlining a common set of guiding principles and best practices in terms of design features to foster the development of national tracking systems, combined with the obligation for some categories of pension providers, namely IORPs and PEPPs, to report to such systems. It would also ensure greater compatibility of these national tracking systems with the European Tracking Service (ETS), enhancing the development of cross-border pension tracking.

A Recommendation would be more appropriate than a Commission Communication as it strikes the right balance between effectiveness and proportionality, by catalysing action for developing new tracking services, while also guiding Member States to improve the design of existing services. In the area of social protection and within the overall framework of the social Open Method of Coordination (OMC), there is a strong track record for tangible implementation of recommendations backed up by mutual learning, exchange of best practices, peer reviews, as well as support and monitoring mechanisms shaping policies at national level, especially if social partners and other stakeholders are duly involved and if financial and technical support is available from EU programmes (see section 6.1.1.2). Accordingly, this option would stimulate progress towards the specific objective of increasing information for individuals about their future pension levels and pension products available to them, without imposing a one-size-fits-all solution in an already fragmented situation.

7.2 Pension Dashboards

The status quo (PM Option 1) would not entice Member States to create pension dashboards and is therefore the least effective in terms of facilitating pension reforms leading to adequate retirement income through higher savings in supplementary pension schemes or products. The cross-country comparability accomplished through a European dashboard (PM Option 3) will make it more effective than the creation of national pension dashboards (PM Option 2) albeit the advantage of comparability would need to be based against the national dashboards better tailored to Member States needs. Since the existence of a European dashboard does not prevent Member States from creating own national dashboards, this disadvantage should not be significant. By contrast, there is a certain likelihood that the creation of national dashboards would benefit from the peer pressure of a European counterpart with data from other Member States with wider coverage, more granularity or higher accuracy.

The efficiency dimension considers the costs of the various options against their benefits. Costs of collecting data, making projections and creating dashboards increase with the sophistication and granularity of the approach. Costs would be highest for PM Option 3 and lowest for the status quo (PM Option 1). The additional costs of moving from national pension dashboards to a European pension monitor depend on data becoming sufficiently harmonised. Creating a classification and mapping all existing pension products already looks challenging against the diversity of schemes used in Member States (see Table 50 in Annex 4). These extra costs need to be balanced against the benefits of cross-country comparability. Both are difficult to quantify. The net benefits attributed to peer pressure at the EU level in fostering domestic pension debates and in the preparedness to learn from experiences in other Member States seems impossible to quantify. On balance and since the costs are more certain than the benefits, the cost element is considered more important for the net benefits than the potential benefits. Therefore, the table below allocates a superior ranking for the criterion efficiency to PD Option 2 over PD Option 3.

Regarding coherence, the value of better information on smarter policy decision making seems undisputed, though it is also evident that good information is not the only critical determinant of policy decisions. The coherence of the policy options would still be correlated with their efficiency. Since PD Option 1 cements the status quo it would be incoherent with the objective to foster transformational change in the EU. It is therefore assessed as the least coherent option.

Table 164: PD policy options

7.3 Auto-Enrolment (AE)

The options on auto-enrolment vary in their effectiveness in accomplishing the objective of higher participation in supplementary pensions. This economic effect can be measured through the expected increase in the participation rates if Member States follow the recommendation. Social impacts will depend in practice on how Member States design their auto-enrolment system with respect to participation of vulnerable parts of the population and allowing sponsors to invest retirement savings in projects delivering on political priorities, such as innovation, competitiveness, the digital and green transition, defence.

An enabling recommendation (AE Option 2) is likely to lead to a higher economic impact defined as higher participation in supplementary pensions than the status quo (AE Option 1). The higher increase in participation rates accomplished in countries that applied good practices suggests that the recommended use of good practices (AE Option 3) will lead to an even higher participation in supplementary pensions and therewith amplify the effectiveness of introducing auto-enrolment. The active support through creation of a public body (AE Option 4) would reduce the administrative burden to employers of offering occupational pensions; an obligation for employers to offer auto-enrolment to their employees would foster participation. Both effects would boost the impact of auto-enrolment on the participation rate, making this option the most effective.

However, as all impacts depend on how Member States will make use of the recommendation, the diversity of Member States starting positions and pension systems prevents a one-size fits all calculation of benefits and costs. To judge efficiency, the criteria suggested consists in anticipating to what extent Member States do not make use of the recommendation or even choose a more ambitious design than recommended. The latter would be welcome and the recommendation should be formulated in a way that does not discourage them from doing so. Nevertheless, this uncertainty, which is in the nature of a recommendation, makes is impossible to undertake a cost-benefit analysis.

Since few Member States have considered auto-enrolment since the release of the EA study that showcased good practices, the status quo (AE Option 1) is likely to lead to no significant number of additional Member States following them. The enabling approach (AE Option 2) bears a likelihood that some Member States will keep the status quo. This is balanced by a significant likelihood that when a Member State designs enabling measures, it also makes use of some good practices. The enabling approach would therefore be more efficient than the status quo. If the recommendation suggests the use of good practices (AE Option 3), it appears unlikely that Member States do not make use of them. The downside would not be larger than under the enabling approach. It is also possible that some Member States create a public body (akin Option 4), implying a superior efficiency of AE Option 3 relative to AE Option 2. AE Option 4 does not allow for overshooting but carries a possibility that some Member States consider the costs of setting up a public body as too expensive for their budget.

Coherence is assessed as the fulfilment of other policy objectives. AE options 2 to 4 are consistent with the objectives of the Savings and Investment Union and the competitiveness compass. They support the European Pillar of Social Rights. Differences in the coherence of the different policy options will ultimately depend on how Member States implement them. As general direction, their degree of coherence is likely positively correlated with their effectiveness. Given that the status quo option implies very little change, it appears as the least coherent. The creation of a public body (AE Option 4) implies that public money is not available for other purposes all other things equal. Conflicting priorities on how to spend public money means that the coherence of this option would be judged as under-proportional to its effectiveness. The use of a public body in the design of investment policies could disturb competition among private IORPs ( 157 ). An obligation for employers could interfere in wage bargaining processes, further reducing the coherence of this option. The use of good practices (AE Option 3) would allow designing the measure in a way that fulfils other objectives, most trivially the good practice of accompanying financial literacy initiatives, making it the most coherent.

Table 175: AE policy options

7.4 IORP review

IORPs are important vehicles to enhance pension adequacy by complementing public statutory pensions in workplace environments. Their development and growth can also contribute to channel more long-term capital in the EU capital markets, and via these to the general economy, at a juncture where the EU economy urgently needs to deploy investments to shore up its competitiveness. A greater role of funded occupational pension funds in the capital markets can benefit market liquidity and contribute to the overall financial stability.

Option 1 envisages no EU action compared to the status quo. In several countries, IORPs remain marginal or entirely absent, and the existing framework does not appear well adapted to support the development of occupational pensions in these contexts. Scale is currently building at a very slow pace, mainly driven by supervisory pressure, which is not uniformly applied across Member States. While preserving regulatory stability and social partners’ prerogatives on how to organise their occupational pension schemes, Option 1 risks reinforcing fragmentation and limiting the role IORPs can play in addressing pension adequacy and long-term investment objectives at EU level.

Option 2 could be effective in enabling pension funds to generate higher return for members and since larger pension funds are likely to benefit more from this potential, it could equally foster market-driven consolidation.

Option 3 offers the most credible and immediately operational pathway for strengthening the EU’s occupational pensions landscape. By tackling structural inefficiencies directly and enabling the emergence of scale and professional governance, Option 3 supports long-term value creation, while respecting the prerogative of social partners in organising occupational pension schemes in a bottom-up approach, therefore ensuring coherence with general policy objectives. It fosters more accountability of IORPs boards towards their members as regards not only capital preservation but also long-term returns. Eventually, this can contribute to more diversified and resilient asset allocations while channelling a larger share of retirement savings towards productive investments and infrastructures, including in the areas identified as EU priorities. This is the most coherent and efficient option, although impact on scale and related advantages would take longer to materialise than with Option 4. Like Option 4, it can entail higher costs of supervision compared to the status quo.

Option 4 is the most effective one in achieving more scale on a shorter-term basis. By introducing harmonised metrics and triggers across the EU and effective supervisory powers in case of non-compliance, this option would allow more rapid convergence towards scale, performance, and governance policy goals. It would also foster more equal treatment and protection of IORPs members’ rights across the EU. However, this approach is demanding on supervisors’ capabilities to processes and subjects them to liability risks if concerned pension providers ask for a judicial review. In addition, the establishment of a mandatory minimum efficient scale for IORPs would require significant time and resources and could still vary from one IORP to another. This option is therefore less efficient than Option 2. Moreover, it would contravene a market-driven approach by mandating consolidation in a “top-down” manner that is not compatible with the current governance of occupational pension schemes. Establishing legal pathways to consolidate to address structural inefficiencies also risks incoherence with the objective of competition policy to keep market concentration at bay. Even if such a top-down approach were possible, the mandatory consolidation could still be subject to the approval of the national competition authority whose decision would not be bound by the rules of this proposal. It is therefore not coherent with the paradigm of consolidation being market-driven.

Table 18: IORP policy options

7.5 PEPP review

Keeping the current framework unchanged in Option 1 would likely mean that the PEPP remains a niche, symbolic product, rather than a meaningful contribution to closing the EU’s pension savings gap. This option would hardly be effective with respect to fostering a larger uptake and thus would not address the problem of the pension gap.

Option 2 avoids complex design features. Ensuring cost-efficiency through a fee cap could make it attractive for online providers to offer PEPP as complement to other execution-only products. This increase in the supplier basis would need to be balanced against a high likelihood that the fee cap will continue to discourage the insurance sector, and potentially even the asset management industry, from offering the basic PEPP. Consumer recognition is expected to remain limited.

Option 3 would appear to make the basic PEPP more attractive to a wider range of suppliers that are currently dissuaded from entering the private pensions market due to the fee cap. In terms of immediate consumer and provider uptake, there is the expectation of increased uptake with Option 3, as it manages to merge the pension suitability features of the PEPP Regulation with the operational features of the execution-only rules in existing EU bodies of law in this field.

Option 4 caters to more specific client needs regarding both the portfolio design and choices or risk management. This option will also appeal to traditional suppliers of savings and investment products who supply more diversified portfolios, while applying a broad range of risk management tools. Option 4 has its own merits in terms of product design and diversification into a broader set of asset classes.

The criterion for the effectiveness of the PEPP options in the table below is their potential to improve the attractiveness of pension products- the objective of this measure. This relies on the extent the different options can boost consumer recognition and ultimately savers’ interest in making use of PEPP. This is related to the argument that a larger take up of PEPP will increase competition and lead providers of national pension products to offer better terms to retail savers. Efficiency is judged against the likelihood that the option accomplishes sufficient supply of PEPPs. This depends on incentivises for providers to enter the market and offer PEPP. Cohesion is assessed as synergies with the savings and investment account, which is an alternative measure to foster retail investment, though not with the purpose of retirement savings.

The prospects of the PEPP to become an effective retirement saving product depends on the potential of the different options to incentivise providers entering the market and offering PEPP. The removal of the fee cap in PEPP Option 3 is likely to generate the largest interest by providers as many flagged it as discouraging element. The possibility to offer either the basic PEPP, the tailored PEPP or both (Option 4) will be more inviting than the possibility to offer a basic PEPP with a fee cap (Option 3) and is therefore more effective than the latter. Efficiency is judged against consumers recognition, which determines savers’ interest in making use of such offer. The ranking of the options therefore depends on whether more consumer acceptance would be achieved by relying on a trusted retail brand or by pointing out the benefits of diversification effects. Maintaining the fee cap in Option 2 is likely to generate the highest consumer acceptance, the combination of two types of PEPPs in Option 4 gives consumers a broader choice, which is appropriate for a long-term pension product serving PEPP savers with different retirement needs. Both demand and supply for PEPP determine the criterion for coherence, for which a relevant dimension is in this case the potential of PEPP to challenge incumbent providers on national supplementary pension markets. In addition, PEPP can be seen as a complement to savings and investment accounts ( 158 ), which are meant as vehicles for investing with a medium-term horizon as opposed to the long-term nature of the PEPP (see Box in chapter 5). Over all three criteria, Option 4 is the superior solution.

Table 19: PEPP policy options

8. Preferred option

The above comparison of policy options revealed that a recommendation for good practices on auto-enrolment (AE Option 3) and pension tracking systems (PTS Option 2) and for national pension dashboards to be developed (PD Option 2) are the best options based on their effectiveness, efficiency and coherence for increasing the likelihood of achieving higher supplementary pension savings and thereby also higher retirement incomes for future pensioners. Following an analysis of the options for revising the EU legislative framework yielded a preference for an approach that scales up pension funds by intensifying supervisory reviews in support of more efficiency (IORP option 3) together with an approach that removes the fee cap and allows more choice between a basic simple and tailored PEPP, allowing also for either more exposure to illiquid assets or guarantees or potentially a combination of both (PEPP Option 4) to foster competition for attractive personal pension products. This section details how the policy options could be operationalised. It continues with a discussion on their interactions and closes with an analysis of the quantitative impact of such a supplementary pension package.

8.1 The preferred policy measures

A Commission recommendation can promote Pensions tracking systems to enable individuals to make informed choices about their pension savings. Member States will be recommended to develop pension tracking systems and will be provided guidance with regard to the design and set-up of tracking services, building on the EIOPA advice and good practices identified in Member States (PTS Option 2). The Recommendation would promote the establishment of new pension tracking services as well as expansion and upgrade of existing ones, to maximise their coverage, quality and compatibility, including compatibility with the ETS to foster the development of cross-border pension tracking in the EU. The recommendation could be backed by financial and technical support available from EU programmes, such as the Technical Support Instrument.

The Commission’s recommendation is that Member States establish national pension dashboards with aggregated forward-looking data that covers public and supplementary pensions in order to (PD Option 2) help them assess their multi-pillar pension systems, creating transparency on the effectiveness of their pension policy. The use of common definitions when creating national pension dashboards would allow aggregating them at a later stage into a European Pension dashboard that establishes benchmarks for comparison with the pension performance in other Member States and consequently allowing for an informed public discussions and learn from good practices in other Member States.

The design of the pension dashboard should follow up on the EIOPA advice, and employ data and statistics already used by the Commission or available to Member States. From the outset, it should use data on public pensions compiled in the ageing report and the pension adequacy report. This option would involve the Commission recommending to Member States that they engage in identifying data that is relevant for their pension system and developing methodologies that allow transforming statistics into forward-looking projections. The agreed definitions, assumptions and practices for the Ageing report and Pension Adequacy Report provide a framework to start such an exercise. Accuracy can be improved through coverage of new data sources and learning from practices used in other Member States over time.

A Commission recommendation could encourage Member States to adopt auto-enrolment in supplementary pension schemes to enhance retirement savings while preserving flexibility in national implementations and the role and autonomy of social partners when applicable. Auto-enrolment requires a legal framework defining eligibility of persons, default options, and investment vehicles, but these details cannot be standardised and must be kept at Member States discretion. To avoid prescriptive designs, the recommendation lists principles for good practices to prevent features that undermine effectiveness (AE Option 3). This would allow Member States to align auto-enrolment with their public pension systems’ redistributive and poverty reduction goals. Although existing investment regulations (e.g., IORP II, Solvency II and PEPP) already address oversight, the need for effective supervision appears so important that it would be part of the recommendation.

To make the use of auto-enrolment effective, the recommendation could ask Member States to apply good practices. These include careful preparation with stakeholder engagement, information campaigns to build public trust, and phased implementation starting with small contributions and limited participants. For example, the targeted consultation revealed a split picture of whether the legislator or employers and social partners should establish the pension scheme available and if no choice is taken – the default option ( 159 ). It also emerged as important to establish a limited set of default parameters, while still allowing participants to customise choices. The specific parameters for contributions, investment vehicles, and payout modalities would need to depend on political choices and preferences in the Member States, Synergies with financial education initiatives and gradual expansion of participation would help ensure understanding of long-term investment risks and benefits.

Most design features must be country-specific, balancing objectives like early savings accrual and support for older workers. Member States may vary in approaches to eligibility for low-wage earners, contribution rates (e.g., low rates to maximize participation vs. higher rates for significant savings), and fiscal incentives. Inclusive access is prioritised, with suggestions such as public subsidies for low-income earners, flexible participation for caregivers, and voluntary top-up options for irregular income groups. These measures aim to reduce inequalities and address the needs of disadvantaged populations.

Employers and social partners play a central role in employment-based auto-enrolment, though making the offer of occupational pension schemes mandatory for employers seems not a feasible policy option in many Member States. It may, however, play a role in Member States where country specific conditions allow, for example where they do not interfere with collective wage bargaining processes. The use of auto-enrolment in collective bargaining agreements or the public provision of administrative support for small employers could facilitate adoption. Public bodies or reformed products like IORP and PEPP might serve as alternatives in countries with fragmented employer participation. Coverage gaps for self-employed workers and those changing jobs require country-specific solutions, such as transferable schemes or hardship criteria for early pension access.

The preferred policy choice for the review of the IORP II Directive is the support of greater scale and cross-border efficiencies in the supplementary pensions sector through enhanced supervisory scrutiny and targeted adjustments to the current framework (IORP Option 3). The provision of clearer guidance on the Prudent Person Principle to the entire supplementary pension sector will nudge IORPs, PEPP providers, as well as providers of other retirement saving products nationally regulated towards investment policies that efficiently and effectively combine risk mitigation with long-term return objectives. The combination of tackling structural inefficiencies, enhancing professional governance and conducting thematic supervisory reviews should result in larger scale and providers more focused on the long-term value creation for schemes participants. The guidance on the prudent person principle will not only help improve pension adequacy but also foster diversified asset allocation. The latter one would eventually enhance pension funds’ resilience and support capital markets and long-term investment. While some stakeholders may face short-term adjustment costs to align with clarified or revised requirements, these transitional costs are expected to be limited and outweighed over time by the anticipated efficiency gains.

The preferred policy choice for the review of the IORP II Directive is the support of greater scale and cross-border efficiencies in the supplementary pensions sector through enhanced supervisory scrutiny and targeted adjustments to the current framework (IORP option 3). The provision of clearer guidance on the Prudent Person Principle to the entire supplementary pension sector will nudge IORPs, as well as providers of other retirement saving products nationally regulated towards investment policies that efficiently and effectively combine risk mitigation with long-term return objectives. The combination of tackling structural inefficiencies, enhancing professional governance and conducting thematic supervisory reviews should result in larger scale and providers more focused on the long-term value creation for schemes participants. The guidance on the prudent person principle will not only help improve pension adequacy but also foster diversified asset allocation. The latter one would eventually enhance pension funds’ resilience and support capital markets and long-term investment. While some stakeholders may face short-term adjustment costs to align with clarified or revised requirements, these transitional costs are expected to be limited and outweighed over time by the anticipated efficiency gains.

The preferred policy choice is to differentiate between a framed basic PEPP and a PEPP with options that would allow for an unlimited number of more sophisticated tailored PEPPs (PEPP Option 4). Targeted changes to the PEPP regulation aim at creating a framework in which providers find it attractive to offer “Basic PEPP” that builds on other design features than a fee cap to ensure a cost-efficient offer. Next to the basic PEPP there will be the tailored PEPP with options that offset potentially higher costs or risk by potentially higher returns from a broader set of investment opportunities or guarantees. This will require that providers of the alternative PEPPs offer at least basic investment advice to customers to understand the complexity of the product features.

8.2 Interactions between the different policy measures

A mix of policy measures is proposed comprising recommendations calling on Member States to take action in their area of competence and improvements to EU legislation where necessary. Recommendations would specify the general policy direction, while allowing Member States to apply discretion in adjusting the general direction to national circumstances. The policy measures at EU level aim to make national policy measures compatible with the functioning of the single market. This compatibility is important because a stronger use of supplementary pensions in individual Member States may create spill-over to capital market conditions in other Member States, for example via investment of pension providers in other Member States or the common goods of financial stability and market integrity.

Since the preferred policy options in each area would address different root causes, they would be largely complementary. For example, it seems highly unlikely that the creation of pension dashboards would lead to Member States deeming pension adequacy to be sufficiently high that governments backtrack on reforming the supplementary pension sector, i.e. they do not create pension tracking systems or no longer consider increasing participation through, for example, auto-enrolment.

Although the aim of both auto-enrolment and pension tracking is to achieve higher participation rates, auto-enrolment targets procrastination, while pension tracking information problems. Moreover, the former will primarily cover employees, whereas participation of self-employed depends on specific design elements such as voluntary participation that Member States may want to allow for. Access to pension tracking would be beneficial also for workers already auto-enrolled. Apart from reassuring them about the benefits of their occupational pension contributions, it could entice them to save more in personal retirement products, or – if this is a feature that Member States allow for- to top up their contributions.

Higher transparency on pension entitlements, availability of PEPP and effective supervision of providers of pension saving vehicles could increase confidence and initiate an increase in voluntary saving for retirement purposes in the population. Higher competition for attractive pension products via PEPP could reduce fees, enticing private households to save more in retirement products. These incentives would be increased if pension funds gain scale and access to investment opportunities that allow further diversification and higher returns. The suitability of PEPP for workplace situations can be an asset in Member States or industries where social partners have no significant role as simple alternative product to life-insurance products. They may also play a role as default investment option for smaller firms to auto-enrol their employees or for self-employed to build up personal pension wealth.

Greater competition in supplementary pension products and more intense supervisory scrutiny towards the profitability of occupational pension schemes will likely improve assets risk management and align asset allocation with long-term objectives of pension schemes’ members. Faced with more intense competition, smaller schemes may find compelling reasons to pursue mergers or otherwise pool assets together to take advantage of cost synergies and greater diversification opportunities professionally managed, which could ultimately benefit their members via higher returns on their pension savings. Consolidation in the pension fund sector could lead to lower fees. The prospects of both lower fees and higher returns could boost households’ appetite in shifting parts of their financial assets from liquid instruments into long-term retirement savings.

A shift of part of pension funds’ asset allocation towards long-term investments will help increase the supply of long-term patient capital, including equity funding, the depth of both public and private EU capital markets, as well as the capital available for infrastructure projects. The combination of higher take up, more transparency on pension entitlements and saving needs, higher returns after fees for savers and a sound supervisory framework can instil the confidence required to allow for material effects.

The intended higher contribution of supplementary pensions to retirement income aims to reduce the pension gap as important social challenge. The benefits would unlikely spread evenly over the population. Since wealthy people have a higher potential to save, they could benefit much more than less well-off people. The creation of pension tracking systems could reinforce this effect. Auto-enrolment will benefit primarily wage earners, the extent low-wage earners, employees with atypical contracts or with care obligations receive fair benefits will depend on the design and specifically the design of fiscal incentives. Supplementary pensions do typically not incorporate a redistributive component and could cement the income inequality during work life into retirement. If retirement savings yield a good return, wealthier people would benefit over-proportionally from this effect. The pension dashboard can help detecting trends in pension inequality and enable corrective measures, for example in the statutory pension system.

8.3 Expected effects of the pension package

Since the pension package analysed in this document is of enabling character, its effect will strongly depend on how Member States, households, employers and financial intermediaries make use of them. The potential benefits of a larger and more efficient supplementary pension sector for individuals, as well as for the EU capital markets and economy will depend on Member States’ uptake and implementation, and on market integration and development. The following estimates should therefore be interpreted as illustrative. They present possible outcomes under specific assumptions; the underlying calculations and assumptions are detailed in Annex 2.

The benchmark for the following analysis is the quantification of the long-term savings need to offset the decline in the public benefit ratio presented in chapter 2.1 it shows that this objective can be accomplished through a combination of more savings in long-term retirement schemes or products and a higher return on these savings. A higher return will allow for lower contributions to the pension scheme and vice versa.

The variation in countries’ experiences with auto-enrolment can be used to set assumption about its impact on the participation rate. The increase was relatively small at about 10% in IT and PL, which implemented auto-enrolment rather quickly. The increase was above 20%pts or is expected to be that high in the UK and IE. The different scenarios presented in Annex 2.4 yield an increase in the participation rate for the EU from 21 to a range of 33 to 38%, which seems to be a realistic value for the use of good practices.

Amounts spent to support wage earners differ substantially between Poland and Ireland. Being enrolled in a pension scheme yielded higher returns than savings compared to bank deposits. This spread has historically been higher than 1 percentage point. This additional capital income would need to be accumulated over the length of the saving period. Assuming a regular saving over 20 years, a 1% return advantage would increase the accumulated saved amounts by 22% after 20 years.

The return could be even higher if the other measures in the pension package led to lower fees and more profitable investments. The EIOPA cost comparison shows that IORPs costs are at about 0.5% in some Member States compared to 1% or higher in others, suggesting that a 25 basis points reduction in fees is feasible outcome of higher competition and more consolidation.

Pension funds’ returns on investment are volatile over time. For example, using the average returns over the last 20 years of sovereign bonds, private bonds and public equity investments and assuming that the investment share of equity will increase from 30% to 35% of the investment portfolio would boost returns by 20 basis points. A shift within the equity portfolio from listed to private equity might further boost the return on pension savings by another 5 basis points. Adding these two elements to the 1% return advantage over bank deposit gives 1.5% additional returns, which would increase savings by 15% after 30 years ( 160 ).

The more detailed scenarios and calculations in Annex 2 show that the combination of higher participation in supplementary pensions and higher returns on these pension savings would allow young individuals to close their pension gap. However, the assumptions spelled out in Annex 2 are numerous and cannot be linked to specific policy measures.

The assumptions of auto-enrolment into occupational pensions with contributions paid from employees’ current income means a gradual increase in the savings that pension funds can invest in the economy. The assumed contribution rate of 4% of wages and salaries implies limited scope to boost the size of the pension sector, at least in the short-term. The impact could be much larger if employees top up their contribution to occupational pensions if they saved more, which, however, seems not the most promising avenue in view of continuing contributions to statutory pensions and other social security. The Irish analysis of the macroeconomic effects of the introduction of auto-enrolment quoted in Annex 2.1 assumes a small temporary decline in GDP growth if the saving rate increases. If the households’ saving rate is assumed to remain constant, there is scope for higher pension savings if households redirect their accumulated financial wealth into pension products. For example, EU households kept about EUR 10 trillion in bank deposits in 2024 of which almost 5 trillion in term deposits. The pension sector would also grow if self-employed increased their voluntary pension savings or participated in occupational pensions. For this to happen will require that individuals recognise that they can earn higher returns on pension products than on more liquid financial assets and have confidence that their investment in pension products is sufficiently safe and higher returns will materialise.

The enhanced supervisory mandates in the IORP II Directive aim to foster consolidation in the IORP sector and to incentivise smaller funds to pool assets with larger asset managers or funds. The clarification of the prudent person principle aims to enable a larger share of pension fund investments in less liquid but potentially more profitable asset classes such as private equity. The scenarios in Annex 2.4 quantify these measures came up with additional private equity investment of EUR 25 billion, corresponding to a fourth of IORPs EUR 100.4 billion holdings in private equity in 2024 (3.7% of total assets). The largest impact would be from the largest pension funds increasing their investment, contributing half of the estimated increase. Market consolidation could contribute a sixth to this increase, asset pooling of smaller IORPs could mobilise an additional sixth. Convergence of the still nascent French IORPs to the average PE share could add another sixth under the assumption that 50% of these IORPs delegate the management of their assets in a way that they accomplish the same ratio of private equity as large IORPs.

The potential fiscal impact on Member States’s budget will depend on the design of the specific policies. Among the possible measures that will be proposed, auto-enrolment can be expected to have the largest consequences for public budgets, as Member States may find it useful to introduce incentives and subsidies to set up the schemes and incentivise uptake ( 161 ). However, in the medium-to-long term those costs or loss of revenues will possibly be mitigated by higher tax revenues on pension income and general economic activity, both expected to be higher than the as-is situation thanks to the positive effect of larger and more efficient pension sector to enhance pension income and finance productive investments.

It should be noted that the increase in participation rates is much higher if these are defined relative to the eligible population. Setting the eligibility criteria is therefore a central element. Making auto-enrolment affordable for low-wage earners to expand the eligible population may necessitate public subsidies. Poland spends less than 0.05% of GDP in contributions to members. Irelands’ estimate of contributing 0.4% of GDP as public contributions may present a maximum for many Member States. The impact of fiscal support will depend on whether they are paid as a subsidy or a tax deduction. A subsidy leads to public expenses, i.e. a 1% contribution on the wage bill would cost a Member State about 0.4% of GDP. If the contribution is instead tax deductible, the government can recoup the lower tax revenues on the contribution from the taxation of the pay outs over the medium to longer term. The burden for the government would consist in the need to refinance the tax shortfall until the pay-out can be taxed. This can be approximated by the interest rate on sovereign debt. In addition to the level of the interest rate, factors that impact on the costs for the public sector are the length of time until the payout will be taxed and possible shortfalls from individuals falling into lower tax brackets when in retirement. The section on the fiscal impact of auto-enrolment in Annex 2.1 provides a more detailed analysis.

9. How will actual impacts be monitored and evaluated?

The proposed policy option of developing pension dashboards would offer a tool to track the impact of the measures. Other than this tool, the pension statistics compiled by EIOPA, Eurostat, the ECB and the OECD would also allow the effect of the measures to be assessed. Changes to the number, size and asset exposures of IORPs would become visible in the existing EIOPA data bases. EIOPA data would also allow monitoring the uptake of PEPP and its regular reports on the consumer perspective can inform about changes to market conditions for personal pension products. As part of the European Semester exercise, the Commission has already reinforced its monitoring of Member State performance including in the area of supplementary pensions. This exercise will therefore also offer a way of tracking Member States policy progress.

An acceleration of reform progress should also be visible in the national measures on supplementary pensions reported to the Eurogroup as part of its annual monitoring of progress with national measures towards the Capital Markets Union (Savings and Investments Union) ( 162 ). This monitoring process includes supplementary pensions and is carried out by the Financial Service Committee with support by the European Commission. Member States have reported on their position with respect to the items covered by the recommendation on autoenrollment, pension tracking and dashboards in summer 2024. In the follow-up exercises they informed about policy measures at national level with respect to the reforms of supplementary pensions and the creation of pension tracking.

Outside the EU reporting framework, the Mercer pension index would likely record faster convergence of laggards to best performers. Progress should ultimately be reflected in upward revisions of detailed performance indicators such as replacement rates, risks of old age poverty, public age-related expenditures.

Member States can also be requested to report as part of their contributions to the ageing and pension adequacy reports about key parameters as for example:

1.Number of Member States with auto-enrolment schemes; participation rates versus opt-out rates; participation rates among different income groups/male versus female/education level; value of assets under auto-enrolment schemes.

2.Implementation of national pension tracking systems; coordination of national systems into an EU system.

3.Projections on expenditure and contribution to replacement rates of supplementary pensions for the ageing report and the pension dashboards.

Given the political nature of the recommendations and the lead time required for pension reforms, an evaluation of the policy measures undertaken could be envisaged after 5 years and a comprehensive evaluation of their impact after 10 years.

Annex 1: Stakeholder consultation

The Commission has formally consulted stakeholders to collect their views on the upcoming legislative and non-legislative initiatives detailed in this Staff Working Document. In particular, it has launched a public targeted consultation ( 163 ) running from 13 June 2025 until 29 August 2025, seeking feedback on the use of automatic enrolment to increase participation in supplementary pensions, on the development of pension tracking systems, allowing individuals to get an overview of their future retirement income, on dashboards, providing public authorities with a better view on their pension system adequacy for clearer personal information and stronger policy data, and on policy options for possible reviews to the PEPP Regulation and the IORP II Directive. It has also launched a Call for Evidence ( 164 ) on 23 June 2025, with deadline on 21 July 2025, with the aim of collecting general comments, opinions and views on how to improve the current framework for supplementary pensions, as well as sharing evidence on best practices. It has also organised a dedicated hearing with social partners at European and national level.

Regarding the public targeted consultation, the Commission has received a total of 112 submissions They include 1 academic/research institution, 42 business associations, 23 company/businesses, 4 consumer organisations, 4 NGOs, 3 EU citizens, 15 public authorities, 6 Trade Unions and 14 contributions qualified as others. Responses came from Austria (3), Belgium (24), Bulgaria (2), Croatia (1), Cyprus (2), Finland (4), France (6), Germany (24), Greece (1), Ireland (6), Italy (9), Latvia (1), Lithuania (1), Luxembourg (3), Malta (1), the Netherlands (8), Poland (3), Portugal (3), Slovakia (1), Spain (2), Sweden (5) the United Kingdom (1), the United States of America (1).

Regarding the Call for Evidence, the Commission has received 47 total submissions which amount to 37 unique responses ( 165 ). They include 1 academic/research institution, 18 business associations, 2 employers’ associations, 5 companies/businesses, 2 consumer organisations, 1 NGO, 7 EU citizens and 1 non-EU citizen. Responses came from Italy (4), Luxembourg (1), Belgium (8), the United Kingdom (2), Czechia (1), Poland (1), Sweden (1), Spain (2), Germany (9), France (1), the United States (2), Romania (1), Portugal (1), Mexico (1), Switzerland (1) and Bulgaria (1). Responses covered topics spanning from general considerations and the role that supplementary pensions can play in supporting the EU economy, to specific policy issues such as autoenrollment, pension tracking systems and dashboards, PEPP, IORPs, the prudent person principle and the role of tax incentives.

General comments

The great majority of respondents was in favour of the Commission’s efforts on supplementary pensions, while underscoring that the latter should complement and not compromise first pillar pensions. There was broad-based recognition of the role that pension funds can play in supporting the economy as well as the depth of capital markets, and of the double objectives (social and economic) that characterises pension policies. Overall, the vast majority of respondents supported the overarching SIU goals under which pension policy reforms would be designed, and agreed in this light that improvements to supplementary pensions are warranted to unlock their full potential for Europe’s citizens and economy.

Among business associations there was strong consensus on the urgent need to expand and strengthen second pillar pensions across the EU, framed as a response to the pressures on the first pillar, demographic ageing, and the low adequacy of current retirement systems. Respondents generally recognised the complexity and diversity of pension systems across Member States. On this, there were diverging opinions on whether reforms should be led at EU versus national level. Nevertheless, there was consensus that any EU-level action should respect national differences and avoid a one-size-fits-all approach. Respondents also emphasised the need to enhance supplementary pension schemes, particularly through auto-enrolment and financial incentives, and to increase participation in funded pension schemes as well as shift investments towards long-term, productive assets (like public and private equities) to improve returns and their sustainability over time. Another diverging point was the role of financial intermediaries, with some respondents affirming the key role of financial advisors and others underlining that despite available tools, citizens still lack motivation to invest in their pension and need nudging tools.

One employer association strongly emphasised the need to respect subsidiarity in pension system design, advocating for EU-level involvement to be limited to broad frameworks rather than prescriptive rules, as well as the need to reduce regulatory burden. Another employer organisation and two public authorities stressed the importance of principle 18 of the European Charter of Social Rights which states that workers and the self-employed in retirement have the right to a pension ensuring an adequate income commensurate with their contributions and guaranteeing a dignified existence, with equal opportunities for men and women to acquire pension rights.

Companies/businesses supported the overall objective of developing the supplementary pensions sector to boost EU capital markets. They vouched for well-structured supplementary pension plans as well as engagement and choice management to make pension systems work well, with emphasis on default options, plan consolidation, and digital tools to support savers.

The input of consumer organisations focused on the poor performance and often high costs of pension products, which puts citizens’ financial security at risk, as well as on the need to tackle issues with financial advice, inducements as a cost driver as well as due to their inherent conflict of interest, and lack of transparency.

The NGO questioned the use of pension funds to promote stock market investments given the low-risk investment nature of these funds and the vulnerability of elderly people.

Citizens’ opinions on the need for EU action were mixed yet underlined that better market structure and broader reforms than just financial ones are needed to improve the EU’s pensions landscape. These include tax incentives, transparency obligations, and one response suggested that citizens could be able to set their own retirement date to motivate saving and investing in higher-growth options.

Finally, although outside of the scope of this exercise – concerning first pillar pensions, the one business association that shared feedback on the issue recognised that the role of public pensions remains foundational. It underlined that greater transparency in first pillar would be useful, such as through mandatory disclosure of estimated benefits (even based on EU-wide guidance on forecast methodology) to all citizens. The response emphasised the importance of financial education and informed decision-making and did not advocate for a structural overhaul of the first pillar itself, but for making it more predictable, understandable, and realistic, as that can be expected to prompt more individuals to save through second and third pillar schemes.

Auto-enrolment

While the Call for Evidence asked for general feedback on auto-enrolment (AE), the targeted consultation zoomed in on specific design elements and good practices. The replies to the specific questions in the targeted consultation revealed some differences in views across stakeholder types beyond the traditional breakdown of stakeholder types. Responses from financial sector representatives turned out to be relatively uniform compared to stark differences from non-financial stakeholders. Therefore, the replies on auto-enrolment were re-filtered with a more granular breakdown across non-financial stakeholder types.

Among business associations, there was broad support for AE as an effective policy tool to increase occupational pension coverage across the EU. However, there was strong recognition that its design and implementation need to be adapted to national contexts, respecting existing pension systems, social partners, and labour laws. The prevailing view was that AE should be part of a coordinated, multi-measure approach to address retirement savings challenges rather than be seen as the silver bullet, or a standalone solution. Diverging opinions emerged on: mandatory AE backed by legislation and social partner involvement for legitimacy versus voluntary and flexible AE tailored by Member States, allowing employees to opt in/out; the role of social partners as key to success versus a possible barrier; and the role of governments in providing administrative support and establishing default plans versus leaving the initiative to market competition, limiting state control. Business associations generally supported the role of public authorities as regards the provision of incentives and subsidies to both employers and employees, and communication campaigns. However, several business associations from the financial sector warned that auto-enrolment and the establishment of default funds by public authorities could undermine competition, and introduce entry barriers to the market. Moreover, they warned against any dirigiste intervention by the State in the pension funds’ asset allocation. On the preferred allocation strategy for the default pension plan, there was general consensus on life-cycle strategies, particularly for default contribution plans.

Employers’ associations’ feedback was more uniform as both responses voiced their opposition to mandatory or EU-imposed AE, taking the example of Member States like Germany where occupational pensions are deeply rooted in voluntary systems and social partner agreements. Support came instead for national autonomy, subsidiarity, and the preservation of existing effective models.

Among companies/businesses, there was strong support for broad-based AE as a means to maximize participation in second-pillar pension schemes, with a focus on learning from international success stories (e.g., US, UK, Australia), and the goal of expanding coverage efficiently and cost-effectively. Some employers noted that employers shall be given the flexibility to set up their own pension schemes, without the necessary intermediation by social partners.

For consumer organisations, there was conditional support for AE into occupational pension schemes, based on ensuring that pension plans provide real value to their members, as well as an emphasis on consumer protection, performance, and flexibility within occupational schemes. There was also opposition for AE in third-pillar pensions, which should remain an individual choice.

Among citizens, the main view was that nudging mechanisms and default settings should be retained to improve participation, but at the same time there should be an introduction of greater individual control, modern investment design, and decoupling from rigid employment-based structures.

Concerning the relevant design features that determine the success of auto-enrolment, most responses attribute the highest priority to state incentives. Possibilities to withdraw early emerged as the least relevant practice. The ranking of the other elements is shown in the table below. Differences in the average prioritisation of these other elements turned out comparatively small.

Table 620: Hierarchy of best practices along their importance, stakeholder replies to the targeted consultation

Note: Average measured as priority on a scale 1 to 7 with one the most and 8 the least important, frequency measured as how often referred to in the 112 responses to the consultation.

The breakdown by stakeholder group revealed that the high prioritisation attached to state incentives is not shared by consumer organisations and employers respectively their associations. Consumer organisations and trade unions consider other dimensions than those pre-selected as most meaningful. Those that detailed their reply highlighted the presence of well-performing default pension schemes. Non-financial corporations and consumer organisations attributed a high priority to administration facilities offered by the State. Employers also attach a much higher importance to the design of the opt-out window than other stakeholders. Public authorities, together with consumer organisations and trade unions, consider the preservation of statutory pensions of very high importance. Supervisors attributed the lowest priority to this objective

Table 721: Differences in the ranking of good practices across non-financial stakeholders.

*) ranking attributed by financial and non-financial stakeholders.

The importance that consumer associations attribute to the features of the default investment plans was subject to a further question in the targeted consultation. Stakeholder allocated the highest importance to flexibility to adjust at later stages and capital safeguard. They attached the lowest importance to life-cycle investment strategies and minimum contributions.

Table 822: Hierarchy of priorities for the design of the default investment scheme, stakeholder replies to the targeted consultation

Note: Average measured as priority on a scale 1 to 6 with one the most and 6 the least important, frequency measured as how often referred to in the 112 responses to the consultation.

The ranking in the previous table is determined by responses from financial stakeholders. Non-financial stakeholders had a different perspective. Employers, public bodies and consumer associations ranked life-cycle investment strategies as top priority. Consumer organisations, supervisors and other public bodies consider capital guarantees as low priority, unlike trade unions. These four stakeholder groups also rank minimum contributions that can be increased later as much more important than the total sample

Table 23: Differences in the ranking of good practices across non-financial stakeholders.

*) ranking attributed by financial and non-financial stakeholders.

Stakeholders are split on the question who should be responsible to establish the default pension plan that eligible participants should enrol in. About half of the reply refer to employers and social partners, the other half to the legislator. Financial sector responses show the same split. Non-financial private stakeholders predominantly see a role for employers and social partners, public bodies for the legislator.

When asked about what public actions are required to make auto-enrolment work, stakeholders allocated the highest priority to the creation of a legal framework, followed by tax incentives. The provision of administrative support emerged as the task that stakeholders considered the least important one. This contrasts to the relatively high rank attributed to administration facilities by the State in the previous question.

Regarding provisions that support equal opportunities for self‑employed and employees not covered by auto‑enrolment, views are split across stakeholders. 40% of them did not reply or indicated they did not know. Among those that provided a reply, tax incentives or subsidies to save in personal pension plans was the most frequent option (25%). Comparable incentives to participate in default occupational pension schemes (14%) or other options (18%) were less frequent.

Table9 24: Hierarchy of enabling actions along their importance, stakeholder replies to the targeted consultation

Note: Average measured as priority on a scale 1 to 7 with one the most and 8 the least important, frequency measured as how often referred to in the 112 responses to the c0nsultation.

Non-financial stakeholders do not differ in the importance they attribute to the legal framework. Supervisors give a higher rank to other dimensions than those pre-selected. This high rank is due to two respondents, which refer to the importance to take heterogeneity across Member States into account. The low priority attached to public administrative support is shared by all non-financial stakeholders except consumer organisations.

Table 25: Differences in the ranking of good practices across non-financial stakeholders.

*) ranking attributed by financial and non-financial stakeholders

Pension Tracking Systems

Regarding the current state of pension tracking, there are divided views on the effectiveness of current pension tracking systems (PTS). Out of 96 respondents to the question, 42.7% (41) consider their national system to function well. However, a combined 49% see a need for significant change, with 31.3% (30) stating their system should be improved and 17.7% (17) reporting their country has no tracking system at all. The replies came from the above-mentioned Member States and third countries with very diverse national situations, ranging from countries with a comprehensive PTS covering all pillars (for instance FR, SE) to those that do not have a tracking system in place at the moment (for instance PL, BG, IE, MT).

A primary limitation identified is the narrow scope of existing schemes. Most participants reported that pension tracking is very limited, often providing data on only the public pensions and lacking information on occupational and personal pensions. Despite this, in member States where systems exist, a majority of respondents mentioned that their functioning is regularly evaluated to identify potential improvements.

Respondents highlighted several core functionalities that an effective PTS must include to empower citizens with their retirement planning:

1.Comprehensive coverage of pillars: the most critical feature is complete coverage of all pension entitlements. A majority of respondents emphasized the importance of a system that includes all three pillars (statutory, occupational and personal) to give citizens a full overview.

2.Core information: the system’s main purpose should be to provide clear and essential data

1.Accrued pension rights: The majority of respondents emphasized that information on past contributions and accrued entitlements is the essential foundation on which a PTS should be built.

2.Future projections: Approximately half of the respondents reported that projections of future pension benefits are crucial for enabling individuals to assess the retirement income and make informed decisions.

3.Simulations: A majority of respondents stated that it is important for any simulation of pension benefits to provide both the gross amount and the amount net of taxes.

3.Usability and accessibility: a successful system must be user-friendly and accessible to everyone:

1.Simple access: Most respondents reported that a PTS should have a simple and secure access, with a focus on ensuring accessibility for less-digitally-proficient users. Some mentioned the need for both digital and non-digital access options to avoid excluding some vulnerable groups.

2.Avoiding information overload: The majority of respondents indicated a preference for the system to provide only basic information, with links to external websites, in order to maintain clear focus.

4.Phased rollout of advanced features: More complex tools are considered a secondary priority. Some respondents mentioned that simulation tools for exploring different scenarios should be implemented at a later stage. Similarly, a few respondents suggested that information on early withdrawal options should also be implemented later and be presented with appropriate warnings.

In terms of stakeholders breakdown, business associations conveyed their strong overall support for PTS across the EU to improve transparency, raise awareness of retirement income, and empower individuals for better retirement planning. They also believed that PTS should be voluntary in terms of design at EU level, with national flexibility prioritised. A majority agreed that PTS should be user-friendly and accessible and cover all three pillars of pensions, while a few stressed the importance of excluding non-pension financial products to avoid confusion. The one company/business that shared its position highlighted the value of existing national PTS infrastructure and stresses the need for flexible, cost-efficient interoperability between systems to support mobile workers, while warning against regulatory overreach that might complicate or hinder effective implementation. One consumer organisation voiced strong support for developing comprehensive national PTS interconnected at the EU level, with an emphasis on transparency, user empowerment, and standardisation across all pension-related long-term investment products. Among citizens, suggestions included expanding PTS beyond traditional pension information to reflects a holistic approach linking pension planning with work-life factors, and filling transparency gaps in current pension statements, especially around investment, leading to difficulty for retail investors to access detailed information – an issue which should be addressed through technological and governmental solutions.

Regarding challenges perceived in setting up a PTS, respondents state that setting up a comprehensive PTS involves political, technical and logistical challenges:

A fundamental aspect is the lack of political drive. A majority of respondents reported that there are challenges in achieving the political consensus needed for the establishment of a comprehensive PTS. Data management and security seem to be one of the most difficult technical hurdles. Ensuring the accuracy and comparability of data from various providers is a challenge, with the majority of respondents emphasising the need for transparency, standardisation, and quality control. Almost all respondents acknowledged data protection as a priority and a legal requirement, particularly in relation to compliance with the GDPR.

The long-term management of a PTS has been highlighted. The majority of respondents underscored the complexity and long-term nature of maintaining and governing the system, which requires strong institutional support, stable funding, and effective public-private partnership. Approximately half of the respondents mentioned the importance of ensuring broad and easy access to the platform and delivering information in a clear, understandable and user-friendly way. Regarding cross-border interoperability, some respondents considered it to be a challenging task that should be addressed at a later stage, as the immediate priority is for each Member States to first establish its own well-functioning national system.

Table 26: Consultation replies about pension tracking systems (ranking with 1 the most important dimension and 5 the least important one)

Pension Dashboard

On pension dashboards, mainly business associations conveyed an opinion in the Call for Evidence and strongly supported the development of pension dashboards at EU level as a key tool to monitor pension adequacy, sustainability, and coverage across Member States. The consensus was that dashboards could significantly improve transparency, policy analysis, and benchmarking to close pension savings gaps. At the same time, respondents emphasised the importance of avoiding additional reporting burdens on pension providers (especially IORPs) and the need to build on existing data and systems.

The targeted consultation accomplished more responses and a wider coverage across stakeholder groups. The table below summarises the average responses on the qualities a dashboard should provide and the dimensions it should cover. Respondents gave the highest priority to the coverage and the forward-looking character of a dashboard. They assessed cross-country consistency as the least important quality. Regarding the dimensions to be covered, pension adequacy and sustainability emerged as the most important ones.

Table 27: Consultation replies about pension dashboards (ranking with 1 the most important dimension and 5 the least important one)

Notes (1) Detailed data about occupational and personal pensions, in addition to statutory pension, (2) pension data by different cohorts of the population (e.g. by gender, age, type of employment, economic sector, income, etc.), (3) projection of pension adequacy and sustainability, based on transparent and robust assumptions,(4) consistent data and methodology across Member States to allow for comparisons the level of retirement income relative to work income now or in the future (5) other included avoidance of reporting burden, (6) the level of retirement income relative to work income now or in the future, (7) measured by its capacity to deliver a decent level of retirement income in the next decades in face of a declining working age population, ( 8) others mentioned were tax treatment, contributions and that the ranking would depend on the target audience.

Responses by financial sector respondents are in line with the total shares. Notable differences have become evident for some non-financial stakeholders. Trade unions assessed the detailed coverage as among the least important qualities, and supervisors that of a forward-looking approach. Trade unions, supervisors and other public bodies highlighted cross-country consistency as important quality. Regarding the importance of dimensions, the only notable difference between non-financial respondents and the total sample is that responses from trade unions assessed fiscal costs as the most important dimension to be covered.

IORPs

Among business associations there was general consensus that the IORP II Directive’s minimum harmonisation approach remains appropriate and should generally be preserved. Many emphasised that NCAs should retain flexibility to supervise based on national legal, labour, and tax frameworks, and that future amendments should respect subsidiarity and proportionality principles. Several respondents also highlighted the importance of recognising national diversity (e.g. in countries with few defined contribution plans). The prudent person principle was generally considered to provide an adequate framework for guiding investments. Some, however, saw room for stronger incentives to diversify portfolios, including through access to alternative assets, while others cautioned against prescribing investment strategies at EU level. Concerns were also raised that restrictive national investment limits (e.g. on private equity) may reduce returns and diversification. On consolidation, most respondents did not identify an immediate need for further action, though some acknowledged that greater scale could be beneficial over time. Finally, reducing administrative burden was also seen as a practical way to encourage participation.

Employers’ associations supported maintaining the IORP II Directive as a “minimum harmonisation” framework, noting that occupational pension systems are deeply rooted in national traditions and labour laws, and that excessive EU-level standardisation would be inappropriate and potentially harmful, as it could compromise flexibility or efficiency, and one respondent explicitly warned against applying Solvency II-style capital requirements to IORPs.

The two business/companies’ responses that were submitted showed diverging views, with one advising against introducing excessive obligations, and the other backing efforts to expand scale and cross-border activities including facilitating cross-border occupational pension operations.

The consumer organisations that provided input argued that the current IORP practices underperform and that a smarter, more participant-focused and flexible regulatory environment is needed to safeguard retirement outcomes across the EU. It also advocated for reforming IORP investment governance and strategies to improve long-term returns and protect real value, better align portfolios with demographics via life-cycle approaches, enhance transparency, oversight, and consumer protection, and enable cross-border market development under a balanced regulatory framework.

The single citizen response, finally, observed that in Germany IORPs are small and controlled by entities subject to Solvency II requirements, which may limit equity investment.

Prudent person principle

Business associations were generally supportive of maintaining and reinforcing the prudent person principle as a core framework for pension fund investment, particularly under the IORP II Directive. They stressed the importance of flexibility, diversification, and the need for local adaptation, rather than rigid investment rules or quantitative limits imposed at the EU level. There was also a consistent recognition of the principle’s value in long-term investment strategies, especially when equity and alternative assets are involved. Some respondents criticised restrictive national rules that limit diversification, undermine performance and pension adequacy, and called for recalibrating such restrictions to align with international standards and EU goals. One respondent also pointed to a plan to further develop discussions on incorporating alternative assets (besides equity) under the PPP while still observing prudence, risk management, and suitability.

Employers’ associations advocated for balancing risk management with return opportunities, encouraging investments in equity and alternative assets without excessive capital requirements, and for allowing a global reach of investment that is not limited to the EU, as geographical restrictions could negatively impact returns.

The only company/business that expressed an opinion on the matter had a similar position, that excessive obligations and restrictions should not be imposed. Finally, the one citizen that had feedback on the topic stated that prudential rules should not hamper pension funds’ ability to build their own diversification.

PEPP

Business associations widely stated that the current PEPP framework has failed to meet its goals due to structural flaws in the regulatory design which makes it not possible to commercialise the product compared to other products in the market. Most of them supported a thorough reform, simplification and a more pragmatic approach to meet the broader goal of improving access to adequate and sustainable pension provision in Europe. There were diverging views on using the PEPP as a second pillar product, as some expressed caution against expanding the scope before core issues are fixed, while others supported its use into occupational pensions to increase coverage. There was near-universal criticism of the 1% cost cap, viewed as unrealistic, stifling innovation and incompatible with investing in alternative assets (e.g. private equity). Some defended the concept of a cap to ensure affordability, but not in its current form. Suggestions included removing or revising the cap, or rather focusing on value for money. Responses also expressed concerns on the prescriptiveness of investment strategies and risk-mitigation rules. On distribution, there were mixed views as many opposed mandatory financial advice for the basic PEPP, whilst others highlighted the value of financial intermediaries, especially given the complexity of pensions. In any case there was broad agreement that distribution should reflect market reality (digital + human factor). On tax incentives, there was very strong consensus that varied tax treatment across Member States is a critical issue and that without national tax incentives, PEPPs cannot compete with existing products. However, harmonizing tax treatment was seen as politically and technically difficult. There were also repeated concerns on the possibility that the PEPP could disrupt national pension systems, and that it should be kept as a third-pillar product. Finally, several responses supported the idea of transforming the PEPP into a label or voluntary standard, rather than a rigid product, which could also allow existing products to transition into PEPP.

The one employers’ association that responded on the issue strongly supported the core aim of the PEPP, but expressed significant reservations about extending PEPP into occupational pensions, particularly in the German context, stating that the use of PEPP should remain complementary to national pension products.

The one company/business that responded voiced strong support for regulatory simplification and flexibility in order to broaden the reach of the PEPP, including removing or easing cost caps to enable innovative pension solutions.

According to one consumer association, which expressed strong support for the PEPP, the product can be seen as an opportunity to improve competition in personal pensions, which currently suffer from high fees and poor performance. Despite the performance issues of personal pensions, PEPP in its current state is not competitive. However, the organisation emphasised that substantial simplification and consumer focus are essential in the review, that the PEPP should focus on delivering value for money, transparency, and accessibility, it should maintain the cross-border aspect and its tax treatment should be aligned with national regimes to level the playing field. PEPPs could also be seen as having an occupational pension role. According to both consumer organisations moreover there should be greater flexibility in fee caps.

Finally, citizens’ responses highlighted two major challenges limiting the PEPP’s success: inconsistent national tax treatments that undermine competitiveness and uptake, and a structural design flaw in the PEPP model, particularly the inflation capital guarantee, which makes the product financially unfeasible.

Tax incentives

Business associations showed strong and widespread support for attractive and well-structured tax incentives to encourage retirement savings, particularly in supplementary pensions (both occupational and individual). Tax policy was recognized as a critical lever for increasing participation, especially in auto-enrolment and life-cycle pension models. At the same time, there was an acknowledgment of the complexity of harmonising tax treatment across Member States, due to national competence and differing economic and social goals. Some differences in opinions existed insofar some respondents supported more consistency across Member States (e.g., equal treatment for PEPPs), while others emphasised the limits of EU-level harmonisation due to national control over taxation. One response also highlighted that national reforms that favour occupational schemes at the expense of individual ones are seen as problematic, especially without adequate transitional support.

A similar trend could be observed among companies/businesses, which also showed overall consensus is in favour of expanding public incentives including tax-based, especially for currently under-covered groups such as the self-employed and those working in small companies, and particularly to increase participation in retirement plans.

Among other stakeholders, the only one that shared feedback on the issue, a citizen, was in favour of supporting EU mutual recognition of pension tax relief, promoting tax best practices among Member States, and granting equal tax treatment for PEPPs and national pension products.

Annex 2: Who is affected and how?

The following table attributes the main benefits and costs of the preferred options to four categories of stakeholders: citizens and employers, non-financial corporations and financial firms as well as the financial sector and the public sector. The subsequent sections zoom in on the evidence of both qualitative and quantitative effects. Since the quantitative impact will depend strongly on how Member States and the financial industry make use of the new opportunities offered by the policy measures, it is not possible to quantify the effects with confidence. Despite these limitations, a further section presents scenarios for such quantification. A final table assesses the main costs and benefits across the different options discussed in the main text.

Table 28: Main costs and benefits across different options

1.Evidence on the impact of introducing Pension Tracking Systems

Financial literacy in the EU is not high, on average. A Eurobarometer survey from July 2023 showed that 82% of EU citizens lack a high level of financial literacy. This varies greatly across Member States and demographic groups, with women, young people, and older adults generally having lower literacy levels. In particular, some individuals across the EU have low engagement with financial planning for retirement, which can be explained by cognitive and behavioural biases. These include a focus on the present (preference for consuming today over saving), inertia (people know they should save for old age but are reluctant to engage with a future and complex subject), projection bias (people overestimate the degree to which their future preferences remain the same as in the present and make short-sighted decisions), reference dependence (people make decisions involving loss and risk in relation to particular reference point) and the complexity and constraints of pension products.

Member States committed to undertake measures to improve financial education and developed financial literacy strategies, also under the impulse and coordination of the Commission ( 166 ). To help financially literate savers, these measures would need to be complemented by effective informative tools and facilities to support them. 

The actual use of pension tracking systems, where they exist, suggests that there is demand for them. Three Member States that already have comprehensive platforms reported about their use in a survey to the Commission in 2024. One of them reported that one adult out of three visited the website of the pension tracking system in 2023. The second Member State reported that it has been visited by 60% of the population aged between 16 and 80 and by 90% regarding the population aged 65 to 69 in 2024. A third Member State reported an equivalent to 6% of the active and retired population in the first nine months of 2023.

Ideally¸ pension tracking systems (PTS) should aim to transform a complex, fragmented pension landscape into a single, understandable overview of pension entitlements. Although not quantifiable, more transparency leads to increased trust in pensions and welfare systems, more accountability of pension providers and altogether better service for individuals. Several studies found that trust is an important determinant of participation in pension schemes and a recent research paper was able to establish that there is a causal link from trust to participation ( 167 ).

The empirical analysis of the introduction of a digital pension app in Germany found evidence of beneficiary effects: the compilation of data from all pension pillars and their aggregation into a digital tool significantly reduced individuals’ self-reported uncertainty about their future retirement income ( 168 ). Users gain pension knowledge and self-efficacy, which are essential foundations for making sound financial decisions, even if they do not immediately alter the allocation of savings. In addition, various business stakeholders stress pension tracking as an important awareness tool for improved financial planning ( 169 ). In terms of practical gains, PTS could help individuals locate and recover “lost” pension pots from previous schemes, which might otherwise go unclaimed.

Access to pension tracking systems is expected to yield a positive behavioural change. The provision of information on pension claims at a time when people are interested in their consequences can help them overcome the behavioural biases of procrastination that lead to suboptimal retirement savings. The large-scale experiment in Germany quoted above found that access to a digital tracker led to a measurable increase in savings balance for participants ( 170 ). Conversely, a randomized survey experiment with Dutch data tested the effectiveness of an online interactive pension tracker but was not able to detect significant impact on the immediate willingness to act or change savings behaviour for savers with a pension gap ( 171 ). Although the current academic research led to mixed outcomes, it could be argued that by fostering greater pension awareness and encouraging more active financial planning, the creation of pension tracking systems may contribute to an aggregate increase in individual retirement savings, which in turn may boost the savings pool available for long-term investment projects.

The positive impact on financial engagement and planning and, in some cases, eventual savings behaviour seems to be most pronounced among financially less literate individuals. In the previously mentioned large-scale experiment in Germany, the increased savings effect was almost entirely driven by participants with low financial literacy. This study concluded that simplifying pension information disproportionately benefits the most vulnerable groups, thereby potentially reducing disparities in pension planning and savings behaviour. By nudging individuals to provide more for their own retirement (notably by saving more on occupational and personal pensions), pension tracking systems could reduce future dependency on social assistance and state-funded minimum pension provisions over the long-term. Furthermore, by highlighting projected pension gaps, they may encourage delayed retirement or continued labour force participation, affecting old-age dependency ratios.

A pan-European network of interconnected national pension tracking systems, as envisioned by the European Tracking Service (ETS) project ( 172 ), could be an enabler of EU cross-border mobility. A research project warned that lack of cross-border pension overviews “could possibly impede the mobility of labour” ( 173 ). When pensions are portable and transparent, workers have less an incentive to stay idle to preserve pension rights and can optimize career choices. By undoing an important obstacle for employees’ move into better-fitting jobs, better portability of pension rights can make the labour market more efficient, potentially raising productivity.

The entity operating the PTS, whether a public agency or a public-private partnership, carries responsibility. It must ensure the technical functioning of the system, coordinate data standards across providers, and guarantee both independence and user trust. EIOPA considers transparency, accountability and non-profit operation to be central conditions for legitimacy. Annual reports and clear service standards are recommended as safeguards.

Building and running a pension tracking system involves multi-faceted costs. The most significant cost drivers are start-up costs. They concern the initial IT development of the infrastructure, system testing, and the complex technical integration of data from different pension providers, each with their own legacy systems. After the launch, operational costs pertain to hosting, maintenance, cybersecurity, data management, user support as well as governance expenses. Administrative costs include coordination and compliance costs. Systems need strong data protection (GDPR compliance), user authentication, and interfaces in multiple languages.

In particular, an operational impact falls on pension providers and data holders. These include statutory pension institutions, occupational funds, insurers and other personal pension providers. They are responsible for transmitting accurate and standardised data to the PTS. This obligation can be demanding for small or non-digital providers, who may need to develop entirely new IT capabilities to comply. Ultimately, these costs would affect the returns on retirement savings. To limit the burden, EIOPA ( 174 ) advises that the system should initially rely on a very small core dataset, including basic identifiers of the citizen and provider, accrued entitlements, projected retirement income and expected retirement age, before expanding to more detailed filed at later stage.

A design choice in this context is the way in which data is exchanged between providers and the PTS. Two main models exist. In the so-called live access model, the PTS does not store data centrally. Instead, when a citizen logs in, the system connects in real time to the provider’s database, retrieves the relevant information and displays it on screen. The advantage of this model is that providers remain in full control of their data and citizens see the most up-to-date information. The drawback is that it requires providers to guarantee constant connectivity, rapid response times and robust IT systems, as users will not tolerate long delays. By contrast, in the central database model, providers transmit their data periodically to a repository managed by the PTS. The citizen then accesses this central system directly, without the need to query each provider in real time. This model provides faster service for users and places less demand on providers during each login, but it introduces some challenges. The operator becomes at least partly responsible for the quality of the consolidated information. The centralised storage increases the need for strong security and governance arrangements.

Supervisory authorities must ensure that the PTS is embedded within a robust legal framework. They must clarify the allocation of responsibilities between providers and operators, particularly in relation to liability for projections. Finally, the establishment of a national PTS could have a European dimension. Mobile workers often accumulate pension entitlements in more than one Member States. Without cross-border compatibility, information would remain fragmented. For this reason, national PTS should be designed with interoperability in mind, ensuring eventual connection to the ETS. This may require additional legal and technical adjustments.

In summary, the limited evidence on costs shows that these vary widely depending on the scale of the national project, national economic circumstances, labour costs and so on. Since access must be free for citizens, some degree of public funding is necessary. To prepare for this, Member States are encouraged to conduct feasibility studies to estimate the cost per user under different design scenarios. EU co-funding via the Recovery and Resilience Facility (RRF), the Digital Europe Programme (DIGITAL), the Technical Support Instrument (TSI), the European Social Fund Plus (ESF+) or the exchange of best practices via the ETS could contribute to the design and roll out of a pension tracking system.

2.The impact of a pension dashboard on different stakeholders

Regarding the impact of a pension dashboard on different stakeholders, the EIOPA advice compares benefits and costs for EU authorities, national authorities, private pension providers, citizens. Regarding the preferred option for additional data collections, it lists additional costs for national authorities for the collection and validation of additional data. Particularly, those authorities that have not been able to make pension projections would need to invest in doing so.

Private pension providers would incur one-off costs for reporting basic pension data. This would be limited for IORPs to the breakdown by age and gender since other data is already collected. Non-IORPs would face higher reporting costs. Citizens would not incur additional costs but could benefit from a transparent view of the contribution of supplementary pensions to pension adequacy. This transparency could also incentivise additional private pension savings, which would directly benefit pension providers in terms of higher demand for their products, thereby compensating them for the higher reporting costs.

Ultimately, citizens would benefit from higher growth and employment as the result of a larger supply of capital and its better allocation in the economy.

3.Evidence on the impact of auto-enrolment

Auto-enrolment intends to overcome people’s behavioural bias of procrastination that causes low and late voluntary private saving for retirement purposes. Actual experiences from countries that made use of it demonstrates that it can lead to a substantial increase in the participation in supplementary pensions. Academic research on auto-enrolment in the US, UK and New Zealand suggests that the impact can be in the order of 30 percentage points or higher increase ( 175 ). The review article by Besheab et al (2024) reports that auto-enrolment accounts in the US for 40% private sector workers 401(k) or comparable plans; almost 60% of pension plans administered by a major provider were with auto-enrolment. Berk et al. (2025) analyse an experiment where British employees where either automatically enrolled or could enrol, finding the participation was 48 %-pt higher for those that were auto-enrolled 9 months later and their balances were more than GBP 100 higher. Similar results were obtained in a second experiment with new employees ( 176 ). The participation rate in New Zealand increased from about 15% in 2006 to above 80% in 2022, that in the UK from 17% to 75% ( 177 ).

Actual increases were much smaller for EU Member States that had introduced auto-enrolment. They were at about 5 percentage points in Italy since 2006 and 10 percentage points in Poland and Lithuania ( 178 ). Economic papers identified some reasons for the different turnouts. Employers’ obligation to offer pensions fostered the take up in the UK; other pension measures impaired trust in PL ( 179 ) and participation in occupational pensions; competition with an established regime that offers severance pay held back participation in Italy ( 180 ).

The empirical estimates undertaken in the AE study with a panel approach that covers experiences in 18 countries and controls for a host of macroeconomic and structural factors, yielded a positive impact of auto-enrolment of 16 percentage points on the participation rate. The latter was defined as active members relative to working age population. The Irish impact assessment on auto-enrolment expects that the coverage could increase by 750,000, which is equivalent to 21% of working age population.

Designing Auto-Enrolment Schemes

Member States can influence the impact of auto-enrolment through the design of its elements, such as eligibility criteria, the contribution rate, opt-out procedures or fiscal incentives. The review of the US experience by a group of academics revealed low opt -out and withdrawal rates; it also suggests that the magnitude of the default contribution rate had no impact on the participation decision ( 181 ). The salary of employees, however, influences whether people opt out or not. Low-wage earners tend to opt-out more often, suggesting that affordability is an issue for them ( 182 ). 75% of low wage earners are enrolled in the UK compared to 88% of all eligible workers ( 183 ). An analysis of the participation of employees with very low wages in Germany, however, found that about 23% of those that earned so little that they genuinely could not afford to save for retirement enrolled and 15% remained in the scheme ( 184 ). This result is consistent with research in the UK that a significant number of low-wage earners did not opt out, although they cannot afford and would be better placed if they opted out ( 185 ). This evidence also suggests that incentives could be differentiated for high and low-wage earners, for example envisaging state participation to contributions (i.e. subsidies) for the latter ones, to ensure greater participation and lower opt-out rates.

Auto-Enrolment Triggers and Contribution Rates

Countries using auto-enrolment employed eligibility criteria to exclude people considered not able to afford the contribution to supplementary pensions. The UK made eligibility conditional on a minimum annual income of GBP 10,000; Ireland has set the earnings threshold at EUR 20,000. There is trade-off in keeping low wage earners outside auto-enrolment. On the one hand, it avoids reducing their disposable income; on the other hand, auto-enrolment will contribute little to the reduction of their old-age poverty. Research in Ireland showed that the planned design of auto-enrolment will likely cover too few households in the lower income bracket to have an impact on the risk-of-old-age-poverty rate ( 186 ). Despite considered successful in the UK, auto-enrolment did not prevent low participation of people that are self-employed, low-wage earners, those with a migration background ( 187 ).

Good practices to circumvent unwarranted effects could consist in starting with a low contribution rate, while announcing its increase over time, like Ireland is planning, or to offer the possibility for people below the income threshold to opt in (which is possible in the UK and will be in Ireland). This might be beneficial for example for young people with low income and support from parents, for women with support from partners, or people with disabilities ( 188 ). The creation of opt-in windows can also be beneficial for other participants that wish to participate but are not eligible for specific regimes, for example self-employed excluded from occupational pension schemes. Wealthier participants may also appreciate possibilities to voluntarily top up their contribution rate. Another good practice would be flexibility in opt-out, re-enrol possibilities and flexibility in the contribution rate. This could cater for the needs of people with atypical work contracts or career breaks ( 189 ), for example because they are caregiver to family members. The impact of auto-enrolment on the gender pension gap seems to be ambiguous ( 190 ).

The inactivity bias leads to some stickiness in the initial decision and analysis with US data suggests that this is a favourable effect of auto-enrolment, especially for people with low financial literacy ( 191 ). A survey by the Irish Central Statistics Office, ahead of the introduction of auto-enrolment, found that 72% of employees in scope of the scheme (i.e. those without an existing occupational pension) intended to remain part of it once enrolled. This is consistent with findings from other countries that people stick to the default option. Around 80% of employees of large companies auto-enrolled into 401(k) plans in the USA chose both the default contribution rate and the default investment fund and 50% still kept them after 3 years ( 192 ). Currently, 57% of the savers use the default option in the Swedish premium pension, corresponding to 75% of all premium pension savings ( 193 ).

The research on the US and UK experiences shows that the impact of auto-enrolment on the saving rate of those enrolled is not easy to cover. The UK expected that auto-enrolment would lead to GBP 17 billion in additional savings per annum ( 194 ), which corresponds to 13% of households’ savings in 2024. Research for the Irish proposal estimates the additional savings could boost the replacement rate by 11 %-pts if enrolled for 20 years and 17%-pts for 30 years ( 195 ). The empirical study in the AE study did not find a positive effect on the contribution per member. Analysis with British data suggests that auto-enrolment reduced private savings. The effect was, however not 1 to 1, but 1 to 0.2, i.e. about 20% of savings are crowded out. Analysis with US data show that the effect on saving behaviour changes over time. The possibility to withdraw pension savings before retirement significantly reduces the magnitude of ultimate retirement income. The effect was up to 60% compared to the accumulated contributions. A different position is taken by Briere ( 196 ), who argued with a view to the experiences in France that the absence of early withdrawal possibilities disincentivises younger people to stay enrolled if their savings are locked in until retirement ( 197 ). In the same vein, analysis produced by a French market participant argues that withdrawal possibilities have an important role in explaining the take up of the PER as voluntary pension product in France. 198

These arguments suggest that design features, and among them early-withdrawal rights, are an important determinant of the impact of auto-enrolment on participation decisions and the level of retirement savings. Some researchers as well as responses to the consultation furthermore highlight access to profitable investment opportunities net of fees as precondition for the impact of auto-enrolment on the amount of savings to materialise.

While the US research suggests that the size of the contribution rate does not have an impact on the decision to participate, it is still important for the accumulation of savings. In Member States with an already relatively high participation in voluntary occupational pensions like Belgium or Germany, occupational pensions contribute modestly to retirement income because the contribution rate is low. Despite the success of auto-enrolment in the UK in terms of participation, a large share of the British population is still perceived as under-saving for retirement because savers do not voluntarily top up their pension savings ( 199 ). Older employers do not benefit enough from auto-enrolment because their contribution period is too low. Low wage earners and those with career interruptions do not save enough via auto-enrolment to accomplish a significant contribution to their retirement income. Design elements such as top-up lump-sum public payments for low wage earners instead of tax incentives or special rules for care givers could mitigate these effects.

Overall, supplementary pensions replicate the income differences during work life in retirement age. To the extent that higher income classes benefit from tax incentives and better opportunities to earn return on their savings, it would even amplify income inequality. Empirical analysis of the Irish proposal, however, shows a more nuanced picture. Higher income classes can expect a larger relative reduction in their disposable income than less wealthier ones because of the minimum income eligibility condition ( 200 ).

Auto-Enrolment and the Role of Employers

The take up of auto-enrolment will also depend on interests of employers offering it. In a survey by the Irish statistical office, more than half of the respondents indicated “that their employer did not provide an occupational scheme” compared to about 25% that got an offer from their employer, but did not to join. Overall, people have a lot of trust in employers in providing an adequate retirement income. In a Eurobarometer, 55% said they trust, thereof 11% that they trust a lot, which is higher than the trust they have in pension offers from financial intermediaries ( 201 ). Research on participation in the US 401(k) regime suggests that the effect of auto-enrolment was larger than that of employers’ co-financing ( 202 ).

Employers’ traditional motivation in providing occupational pensions was to attract talent, strengthen employees’ loyalty and to gain a reputation as responsible entity. This motivation has lost in importance in an increasingly mobile work environment whereas the reputational gains are not quantifiable. Employers’ interest seems to be higher in social partner models because involvement of many employers and a large pool of employees ensure scale effects and allows for portability within the industrial sector. When used in collective agreements, the offer of occupational pensions also helps attract and keep employees in the sector. There is also an indirect beneficial effect of occupational pensions for employers since parts of the contributions to occupational funds flow back to the corporate sector via pension funds’ investments 203 . The choice between higher salaries or higher contributions to occupational pensions also offers an extra degree of freedom in collective bargaining. This could theoretically lead to the mitigation of price-wage spirals in case of external price shocks or the moderation of a sudden increase in capital costs if wage earners agree in times of these shocks that a part of their salary is not paid out but be allocated to their pension rights.

The offer of occupational pensions as a tool to attract talent is largely limited to large firms. SMEs have made much less use of this instrument. Since the costs of offering occupational pension has a large fix component, SMEs have a natural disadvantage. The organisation of such schemes through social partners supports SMEs, the possibility to use basic Pan-European-Pension-Products (PEPPs), could further reduce this disadvantage. Auto-enrolment is likely to increase the number of employees that agree to such an offer. The proportionally higher costs for SMEs of setting up auto-enrolment has been recognised for example in the UK and IE and led them to phase in auto-enrolment gradually. Smaller employers were given more time to adjust in the UK, while Ireland will gradually increase contribution rates over 10 years.

Auto-enrolment may have a positive effect on employers’ incentives to offer occupational pensions because they reduce the likelihood that too few employers voluntarily enrol in them to make the offer economical. The expectation of too little take-up may discourage in particular smaller employers from offering occupational pensions. From the perspective of employers, mandatory enrolment would therefore be superior to automatic enrolment with an opt out clause. Smaller employers still would have to shoulder larger administrative costs. This is why tax incentives or other compensation schemes or public support on administration look justifiable, especially against the background information from the UK that employers pay higher attention to the costs of such systems than to the returns for their employees. Few Member States provide fiscal incentives for employers beyond the deductibility of employers’ contributions from their tax obligations ( 204 ).

An alternative approach can be found in the New Zealand KiwiSaver scheme ( 205 ). In that scheme, employers pay employee deductions and provide relevant information to the tax authority (Inland Revenue), but the main relationship is between the employee and a KiwiSaver provider (e.g. banks, fund managers). If an employee does not choose a specific scheme or provider, they are automatically enrolled into a default scheme by the Inland Revenue (or the employer, if they have a preferred provider).

When set in an employment context, the experience in the US and the UK of declining enrolment rates when people separate from employers are noteworthy. Beshears et al. (2024) found that US employees withdraw part of their 401(k) savings when leaving the employer. In the UK, employees were enrolled in a new scheme when changing the employer, which led to a plethora of small pension pots and small funds investing them, with adverse effects on the possibility to use scale effects in the investment process ( 206 ). In Lithuania, members have their individual accounts and their access is independent from the employer ( 207 ). Denmark and Sweden support flexibilities by applying multi-employer schemes. In German occupational pensions, members that move their job can decide whether they keep their pension account with the previous employer or transfer the rights to a new employer. A centralised administrative body may circumvent portability issues.

Auto-Enrolment and Fiscal Incentives

Tax incentives may boost the interest in occupational pensions and OECD analysis suggests that tax incentives can be a very effective instrument to foster voluntary pension savings ( 208 ). The effects differ, however, between income cohorts. People with high income optimise their after-tax returns of different investment opportunities without a large impact on their saving rate. Middle-income earners individuals tend to increase their savings in reaction to tax incentives, while the impact on low-income earners depends on whether those incentives make participation more affordable. Very low-income earners may not benefit at all if their income is below tax thresholds so that they cannot deduct tax advantages.

Moreover, fiscal incentives for employees to participate in occupational pensions or to save voluntarily or for employers to compensate for their administrative burden result in a loss in tax revenue, which, depending on the chosen relief, can be significant in the short to medium term. According to an OECD estimate, tax-favoured retirement savings can cost governments between 0.5% and 2% of GDP annually, depending on the country’s structure and take-up rates ( 209 ). Data on the magnitude of tax breaks on private schemes collected by the OECD shows that in countries with high occupational pensions such as the Netherlands, Canada, Australia, the government spends around 2% of GDP, in the US and UK about 1% of GDP. This data, however, shall not be taken in isolation, but included in the overall government expenditures for pensions, as in some cases incentives for supplementary schemes are accompanied by reductions in expenditure for statutory pensions.

Tax incentives often disproportionately benefit higher-income individuals who are more likely to participate in private pensions and gain from tax deductions. This can lead to regressive fiscal outcomes, where wealthier individuals receive greater public transfers relative to their income, reducing the overall progressivity of the Member States’ tax system ( 210 ). An analysis of the impact of taxes on supplementary pensions by means of a microsimulation model on income distribution in the EU Member States yielded that tax treatment has an overall redistributive effect. It favours low-wage retirees and put a burden on high wage earners ( 211 ).

Hence, a cost-effective, targeted, and judicious use of tax incentives in combination with other forms of incentives, e.g. subsidies for various income cohorts is a key consideration in the policy design to boost the uptake of supplementary pensions.

Governments may forgo income tax revenue by exempting pension contributions from taxation. Many Member States actually do this, albeit most with quantitative limits ( 212 ). In addition, by exempting investment gains from capital gains or dividend taxes (Exempt-Exempt-Taxed or EET model) and thereby not taxing the investment returns on pension savings, the government foregoes another source of potential tax revenues. Most Member States do not totally exclude retirement savings from taxation. Usually, either the contributions to retirement savings are exempted from tax and the final payout in retirement is taxed or contributions are paid after tax is paid on income and the payout from the pension scheme is tax free. Exemption of contributions provides incentives to save, which is important in systems that allocate a role to voluntary savings for retirement, and to take risks ( 213 ). Deferring the taxation to the payout phase implies a lower present value of the tax revenues compared to the taxation of contributions if the tax rate were the same, with the magnitude of the tax loss depending on the discount rate, the length of the period, and the returns on pension savings.

Taxation of investment income also differs across Member States and tax codes are complicated inter alia because they attempt to avoid double taxation. The impact of tax breaks therefore depends on when the tax is paid and the tax rate applicable to the individual in that moment. If retirement income is lower than earnings during the work career and tax rates are progressive, the government will ultimately forgo tax revenues.

The detailed analysis of the impact of pension taxation by means of a microsimulation model that covers the details of Member States tax systems by the Commission’s Joint Research Centre also draws attention to methodological challenges such as the choice of the appropriate benchmark ( 214 ). Figure 22 compares the impact of tax on households’ disposable income as proxy for foregone revenues with the taxation that would occur in a pure E(E)T system, including thresholds and limits to the deductibility of pension contributions. Although the costs of the actual tax treatment are higher than that of the pure EET system in most Member States, the analysis showed a tax gain for several Member States (see Figure 22) ( 215 ). High foregone tax revenues in some CEEC are caused by tax reliefs on the income of retired persons.

Figure 22: Costs and revenues of pension-related tax expenditures relative to the benchmark of an EET system without exception and limitations, data for 2017.

Source: Barrios et al. (2018).

Governments may also offer fixed subsidies, matching contributions, or kick-start incentives. Poland uses lump sum payments to incentivise lower income earners to participate. The AE study suggested to use such kind of financial incentives in auto-enrolment schemes to close the gender pension gap. Such lump sum payments should favour women because they are overrepresented among the low-wage earners. The AE study also quoted two other precedents: the fixed contribution Germany pays to the parent, which is usually the mother, as part of its promotion of the Riester rent, and the contribution in Chile paid when a child is born. Analysis of the distributional effect of the Irish scheme, however, yields a marginal effect on the gender pension gap because of the minimum income eligibility condition. The lump sum payment will also not undo the lower contribution period of women as cause of the gender wage gap. A study identified care credits and other means to address this cause ( 216 ).

Government contributions constitute direct fiscal outlays. For example, New Zealand’s KiwiSaver benefits from an annual government matching contribution, which cost the government around NZD 900 million in 2021–22, corresponding to 0.4% of its GDP per year. Ireland’s counts with also 0.4% of GDP for public top-up contributions to retirement savings. In Poland, the government pays a lump-sum amount of 250 PLN, equivalent 60 EUR, when a member joins an employee capital plan (PKK) and contributes a broadly similar amount each year. When multiplied by the number of users, this corresponds to about 0.03% of GDP. The costs of such subsidies are more transparent than tax incentives, which accrue and need to be budgeted each year.

If employer contributions are tax-deductible, which is the case in most Member States, albeit with quantitative limits in many, these deductions reduce corporate tax revenues. Tax losses would be higher the larger the take up of occupational pensions. The same effect on tax revenues would materialise if employers pay a higher wage instead of a contribution to their employees’ retirement savings. This potential tax loss might however be offset by other effects of higher wages such as higher social security contributions, higher taxes on income and consumption. To the extent there is a limit on the deductibility of employers’ contribution and none on the deductibility of higher salaries, it may disincentive employers to offer occupational pensions. The forgone revenue for the government would need to be considered against additional future tax revenues in the event higher pension savings improve the capital allocation and lead to higher economic growth. This macroeconomic effect appears however impossible to quantify. The possibility for employers to deduct their contributions for supplementary pensions from other social security contributions implies a revenue shortfall of other social security institutions.

Cost of Auto-Enrolment Schemes for Governments

The fiscal cost of introducing an auto-enrolment pension scheme in a country will vary depending on the scheme’s design. Fiscal costs are mainly driven by additional incentives that a government chooses to provide, such as tax relief or matching contributions, but also the cost of setting up the necessary administrative infrastructure. Building an auto-enrolment system from scratch entails establishment costs for various items such as IT infrastructure, regulatory framework, setting up a central clearing house. Ongoing costs comprise compliance, supervision, and public education campaigns. Ireland’s auto-enrolment system was projected to cost €2.8 billion over 10 years, including €1.3 billion in state contributions and €1.5 billion in administrative and system costs ( 217 ). The long-term annual cost of the government top-up contributions would amount to 0.4% of Irish GDP.

Costs for information campaigns that accompanied the introduction of auto-enrolment were EUR 19 million in New Zealand, around GBP 25 million in the UK, EUR 17 million in Italy ( 218 ). Much higher are the costs for the public sector if it centralises the administrative organisation of collection, registration and investment although in the case of the UK and Ireland the public funding of a body doing this was designed as initial investment and the public bodies should self-finance them from fees after the initiation phase. The UK ‘Nest’ was financed through public loans initially of about GBP 170 million, growing to GBP 1,200 million through follow up loans and interest, which should be paid back until 2038 ( 219 ). The Irish Central Processing Agency is estimated to cost EUR 6.5 million in the first year and accumulated EUR 250 million over 30 years ( 220 ). The Swedish default premium pension A7 Såfa has financed itself from pension contributions without reliance on public funding since its inception in 2010, despite being owned by the state.

Costs of Auto-Enrolment Schemes for Employers

For employers, auto-enrolment can imply administrative costs, in addition to the contributions to the employees’ pension schemes (which may be compensated by partial or full tax relief). In Ireland, the employer contributions are set to rise from 1.5% of the annual salary in the first year of adoption, to 6% in the tenth year following the introduction of the auto-enrolment. Most consultancy reports on the introduction of auto-enrolment focus on the costs of the employers’ contribution to the pension plan, which can be derived from the employers’ contribution rate and the number of staff covered.

Other costs that employers face relate to upgrades of the payroll administration systems, staff training to manage enrolment, opt outs, track contribution or report compliance, or the costs of outsourcing these tasks to external providers. Adjustments to payroll IT systems need to ensure that contributions be calculated, withheld and paid, and, if the employees have choices between different funds, that the contributions are paid to the right fund and ascribed to the right individual. When employees change funds, shifts between funds must be tracked.

Estimates for administrative costs rely mainly on the comparison with the other countries, specifically New Zealand and the UK. In the evaluations for New Zealand and United Kingdom, employers reported that costs were lower than anticipated. The survey revealed annual compliance costs of NZD 770 when autoenrollment was introduced in 2009, declining to about NZD 660 in 2013. These numbers concern costs for micro, small and medium-sized firms ( 221 ). A comparable evaluation in the UK in 2017, revealed median costs of GBP 400 ( 222 ). Consultancy and industry reports for the introduction in Ireland range from EUR 500-2,000 set up costs for an SME, and EUR 200-500 per annum for ongoing administration. Larger employers are estimated to encounter 30-50% lower costs from scale economies.

A 2025 survey of almost 500 employers about retirement plans for their staff suggests that their priorities differ across selected Member States. While the survey did not ask about views on auto-enrolment, the replies are indicative of the elements companies consider most relevant. Overall, helping employers understand their retirement plans (financial wellbeing) and make good decisions (employee experience) through communication and online tools comes high, followed by efforts to improve their retirement outcomes and containing management costs.

Table 29: Employers’ priorities for retirement plans, % of replies in 2025 survey

Share of firms indicating top or high priority, *only asked in DK and DE. Not shows are results for Switzerland and Italy, the latter due to too few observations.

Source: WTW Towers Watson, Rethinking retirement plans in Europe, 2025 Western Europe Retirement Plan Survey.