## prehfwcgea

## Source details

**Canonical URL:** [prehfwcgea](https://www.imf.org/-/media/files/publications/dp/2021/english/prehfwcgea.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2021/english/prehfwcgea.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2021/english/prehfwcgea.pdf.json)

---

### Executive Summary — Overview and purpose
- Takes stock of where European pension systems stand and assesses their key characteristics.
- Presents a novel measure—the Proportionality Measure (PM)—that compares lifetime benefits and contributions to examine long-term sustainability, fairness, and intergenerational equity.
- Documents pension reforms over the past three decades and assesses their macro-fiscal implications using projections encapsulated in the 2021 Ageing Report (EC 2021).

### Sustainability and balance (Proportionality Measure findings)
- Before the GFC, lifetime benefits in present value terms doubled lifetime contributions on average.
- Reforms in the decade following the GFC reduced the lifetime benefit-contribution ratio to nearly 1.5 for younger generations (in net present value terms), aimed at cohorts retiring in the 2040s.
- Subsequent reforms and reversals partly eroded gains such that, as of 2021, the lifetime benefit-contribution ratio stands at about 1.7 for younger generations.
- Long-term pension spending projections:
  - Number of countries with pension spending of more than 10 percent of GDP goes from 9 out of 27 in 2000 to 18 out of 28 in 2050.
- Conclusion: Pension systems in most European countries are not sustainable under current policies.

### Proportionality Measure (PM): definition, benchmarks, and role of r–g
- PM operationalization: ratio of (undiscounted) sum of pension benefits to sum of contributions, both expressed in percentage of GDP, computed for cohorts.
- Actuarial fairness: PM = 1 (when r = g in a pure PAYG system) implies cohort-level actuarial fairness.
- Choice of discount rate:
  - For most European countries the paper uses GDP growth as an approximation to the sustainable discount rate in deriving and interpreting the PM.
- Benchmark Table (sustainable PM depending on r–g and debt level; rows r–g = –2, –1, 0, 1, 2; columns Debt level = 0, 25, 50, 75, 100, 150):
  - r–g = –2: 1.00, 1.06, 1.12, 1.18, 1.24, 1.35
  - r–g = –1: 1.00, 1.03, 1.06, 1.09, 1.12, 1.18
  - r–g = 0: 1.00, 1.00, 1.00, 1.00, 1.00, 1.00
  - r–g = 1: 1.00, 0.97, 0.94, 0.91, 0.88, 0.82
  - r–g = 2: 1.00, 0.94, 0.88, 0.82, 0.76, 0.65
- Key qualitative implications:
  - When r > g, actuarially fair PM > 1; sustainability typically requires PM < 1 if initial debt is high.
  - When r < g, actuarially fair PM < 1 and sustainability can allow more generous benefits.
- For cross-Europe assessment the paper maintains a Europe-wide PM benchmark at 1.0 while noting r–g and debt can demand different sustainable PMs.

### Interest rate–growth differentials and empirical patterns
- Several high-debt countries (including Italy, Portugal, and Spain) experienced notably elevated r–g since the GFC; r–g has been negative for Germany and most Nordic countries.
- Pension system deficit in Europe currently stands at about 2.5 percent on average.
- Pension system deficits accumulated since the 1980s are of the same order as general government debt for an average European country.
- Example: Italy has had r–g at about 2 since 1980; for debt about 150 percent of GDP this requires a PM benchmark of about 0.56–0.65.
- Aggregate historical r–g patterns (as presented): selected multi-period means and medians are reported in Table 2 (preserved in source).

### Reform progress, reversals, and implementation risks
- Reform actions since the GFC:
  - Raised retirement ages.
  - Streamlined benefit entitlements.
  - Flattened profile of pension spending projections.
- Reform momentum waned in recent years; several countries reversed earlier measures.
- Phased-in reforms have implementation risks:
  - Can deepen intergenerational divides for decades by protecting current retirees and shifting costs to younger cohorts.
  - Reversals in emerging Europe often address short-term deficits at the expense of future adequacy.
  - In advanced Europe, reversals partly rolled back retirement age extensions and increased pensions to current retirees (examples noted: Greece, Italy, Spain).
- Specific reversal and policy-change examples:
  - Greece reinstated benefit elements and provided extra payments; Hungary reintroduced a 13th pension; Italy introduced a loosely targeted 14th pension payment and temporary cash benefits.
  - Hungary closed mandatory private DC funds and transferred assets to the state; Ireland transferred the National Pension Reserve Fund to the state.
  - Spain postponed or suspended automatic balancing mechanisms (IRP); Slovakia cancelled automatic adjustment mechanisms.

### Intergenerational equity and political economy
- PMs indicate widening generational disparities:
  - Older generations receive lifetime benefits that exceed lifetime contributions more than twofold.
  - Younger cohorts access public pensions later and receive relatively lower benefits.
- Risk dynamics:
  - Disparities may disincentivize younger generations’ participation and heighten reform-reversal risk.
  - Median-voter aging and higher old-age dependency ratios increase pressure to roll back reforms.
- Political-economy enablers for reform durability:
  - Measures appealing to shared principles of equity.
  - Enhanced financial literacy.
  - Enfranchising younger cohorts (evidence: a 1 percent increase in young voter turnout increases allocation to higher education by 0.77 percent).

### Labor market, adequacy, and affordability
- Contribution and financing context:
  - Average statutory pension contribution rates are 22 percent of wages.
  - Average collection is 7 percent of GDP in revenue—about 0.32 percentage points of GDP per point of payroll tax.
  - Contributions cover on average 70 percent of pension spending; remainder largely financed by general revenue.
- Employment and taxes:
  - Employment rates tend to be negatively associated with labor taxes that include pension contribution rates; micro evidence shows stronger responses among low-income workers, women, single parents, second earners, and older workers.
  - A "10 percentage points change in the after-tax wage can raise employment by 2–5 percent (IMF 2012)."
- Incentives and design:
  - Strengthening actuarial contribution–benefit links in earnings-related systems can mitigate discouraging impacts and may encourage participation and employment.
  - Minimum-pension schemes and minimum-contribution requirements can create bunching incentives (examples: Greece before 2016–17 reforms).
  - Policy trade-offs: reducing replacement rates to ensure sustainability while expanding targeted minimum/basic pensions to protect low-income retirees.

### Adequacy, poverty, and distributional outcomes
- Old-age pension accounts for about 72 (85 including survivors’ pensions) percent of single-person retirees’ total income in the EU on average.
- Replacement rates:
  - ERR projections: Compared to 2009 AWG, ERRs in the 2018 vintage start higher around 2010 and decline by about 3 percentage points in the long term.
  - Effective average accrual rates estimated to fall from 1.6 percent in 2008 to 1.3 in 2050.
- Poverty and spending:
  - Average pension system deficit currently about 2½ percent of GDP after peaking near 3 percent post-European sovereign debt crisis.
  - AWG projections: average pension spending around or above 12 percent of GDP in Europe between now and 2070.
  - Average pension spending was 9 percent of GDP in 2000.
  - Annual Europe-wide pension system deficit will stabilize at about 4 percent of GDP by 2050s.
- Empirical associations:
  - Higher pension spending historically associated with lower real growth (regression R^2 = 0.7225 reported).
  - Pension spending associated with higher deficits and debt (regression R^2 = 0.459 reported).
  - Negative association between pension spending and gross fixed capital formation since euro inception (regression R^2 = 0.2361 reported).

### COVID-19 implications
- COVID-19 has contributed to another wave of increase in public debt, though cushioned by lower interest rates.
- Costs of the crisis have been largely shouldered by younger generations, including scarring and human capital losses.
- Governments will need to strengthen resilience of economies and social security systems against future shocks.
- The crisis strengthens calls for fair, equitable, and sustainable pension systems in the post-COVID-19 world.

### Policy implications and recommended directions
- Address existing imbalances through more equitable burden-sharing across generations.
- Frontload reforms to ensure sustainability and fairness across generations; avoid excessive phasing that defers costs and deepens intergenerational divides.
- Design reforms to:
  - Minimize labor market distortions.
  - Strengthen links between contributions and benefits (actuarial neutrality).
  - Appeal to shared principles of equity to improve political acceptability.
- Encourage later retirement and promote complementary instruments (private or occupational) to support adequacy.
- Enhance financial literacy to improve public support for reforms and participation.
- Use targeted basic/minimum pensions or social assistance to protect low-income retirees while preserving contribution–benefit links.

### Data, methods, and robustness
- Key data sources: AWG Ageing Report vintages (2009, 2018, 2021), Eurostat, OECD, UN population projections, Eurosystem Household Finance and Consumption Survey (wave 3).
- Cohorts: birth-year cohorts; working life defined from age 22 to calculated effective retirement age; pensioners from effective retirement age to 100.
- Lifetime cohort PMs computed for cohorts born 1930–2000 (turn age 65 between 1995 and 2065) using 2021 AWG projections.
- Robustness checks: sensitivity tests on retirement ages, age-spending and age-income distributions; results reported as robust (see Annex 1).
- Annex findings:
  - For nearly all European countries the PM is well above unity.
  - For early cohorts who retired around the GFC and the European debt crisis, PM ranges from 1.4 (Norway and Spain) to about 4.5 (Greece), averaging around 2.0 where sufficient history exists.
  - For younger generations PMs hover around 1.7 on average.
  - Only Estonia, Latvia, and Portugal could consider pension systems sustainable and actuarially fair for younger cohorts under current assumptions.

### Selected numerical highlights (preserved exactly from source)
- Average statutory pension contribution rates: 22 percent of wages.
- Average collection: 7 percent of GDP in revenue—about 0.32 percentage points of GDP per point of payroll tax.
- Contributions cover on average 70 percent of pension spending.
- Pension system deficit in Europe: about 2.5 percent on average.
- Pension spending: increased from about 8½ percent in 1980 to 11¼ percent today in countries with more mature pension systems; reached 12.1 percent in the euro area.
- AWG projection comparisons: 2060 pension spending declines by nearly 2 percent of GDP in mature systems between 2009 and 2021 vintages; about ¾ percent of GDP decline in currently maturing systems.
- Effective average accrual rates: 1.6 percent in 2008 to 1.3 in 2050.
- IRR examples: "3.4 percent in Germany to 7 percent in Greece" for cohorts, exceeding average real GDP growth by about 1.8 and 6 percentage points, respectively.
- PM vintage comparisons:
  - Pre-GFC (2009 vintage): PMs ~2 on average.
  - 2018 vintage: mean and median PMs converged toward 1.6 and 1.5, respectively.
  - 2021 vintage: PMs for younger generations average around 1.7.
- Long-term projection markers:
  - Number of countries with pension spending >10 percent of GDP: 9 out of 27 in 2000 → 18 out of 28 in 2050.
  - Annual Europe-wide pension system deficit will stabilize at about 4 percent of GDP by 2050s.
- Table of sustainable PM benchmarks (excerpt):
  - r–g = –2: 1.00, 1.06, 1.12, 1.18, 1.24, 1.35
  - r–g = –1: 1.00, 1.03, 1.06, 1.09, 1.12, 1.18
  - r–g = 0: 1.00, 1.00, 1.00, 1.00, 1.00, 1.00
  - r–g = 1: 1.00, 0.97, 0.94, 0.91, 0.88, 0.82
  - r–g = 2: 1.00, 0.94, 0.88, 0.82, 0.76, 0.65

*Source: Executive Summary, Introduction and Motivation, Box 3, Annex excerpts — Pension Reforms in Europe: How Far Have We Come and Gone? (IMF staff).*

### Executive Summary ������������������������������������������������������������������������������������������������������

### Executive Summary

### Overview and purpose
- Takes stock of where European pension systems stand and assesses their key characteristics.
- Presents a novel measure—the Proportionality Measure (PM)—that compares lifetime benefits and contributions to examine long-term sustainability, fairness, and intergenerational equity.
- Documents pension reforms over the past three decades and assesses their macro-fiscal implications using projections encapsulated in the 2021 Ageing Report (EC 2021).

### Key findings on sustainability and balance
- Before the GFC, pension systems in Europe provided lifetime benefits that, in present value terms, doubled lifetime contributions on average.
- Reforms enacted in the decade following the GFC reduced the lifetime benefit-contribution ratio to nearly 1.5 for younger generations (in net present value terms), aimed at cohorts retiring in the 2040s.
- Subsequent reforms and reversals partly eroded previous gains such that, as of 2021, the lifetime benefit-contribution ratio stands at about 1.7 for younger generations.
- In the long term, pension spending would remain at unprecedentedly high levels:
  - Number of countries with pension spending of more than 10 percent of GDP goes from 9 out of 27 in 2000 to 18 out of 28 in 2050.
- Pension systems in most European countries are not sustainable under current policies.

### Reform progress, reversals, and implementation risks
- Reforms in recent decades:
  - Raised retirement ages.
  - Streamlined benefit entitlements.
  - Flattened the profile of pension spending projections.
- Reform momentum has waned in recent years and several countries have reversed earlier measures.
- Phased-in reforms have implementation risks:
  - Can deepen intergenerational divides for decades.
  - Protect current retirees while shifting political and fiscal costs onto younger cohorts, increasing reform reversal risk.
- Reversals in emerging Europe often address short-term deficits at the expense of future adequacy, sometimes worsening long-term sustainability.
- In advanced Europe, reversals have partly rolled back retirement age extensions and increased pensions to current retirees, shifting implementation farther into the future (examples noted: Greece, Italy, Spain).

### Intergenerational equity and political economy
- Reforms have widened intergenerational divides:
  - Older generations (current retirees) receive lifetime benefits that exceed lifetime contributions more than twofold.
  - Younger cohorts will access public pensions later in life and receive relatively lower benefits compared to current retirees.
- Such disparities may disincentivize younger generations from participating in pension systems and heighten risks of reform reversals, adding pressure on long-term sustainability.
- Reform acceptability is aided by measures that appeal to shared principles of equity and by enhanced financial literacy.

### Labor market, adequacy, and affordability considerations
- High pension spending and persistent deficits can:
  - Constrain room for policy maneuver.
  - Lead to efficiency losses and crowd out other productive spending, including resources needed for greener and more inclusive economic transformation.
- Room to increase pension contribution rates is limited:
  - Average statutory pension contribution rates are 22 percent of wages.
- Employment rates tend to be negatively associated with labor taxes (that include pension contribution rates), but the employment impact depends on whether contributions are perceived as valuable savings.
- Strengthening actuarial contribution-benefit links in earnings-related systems can mitigate discouraging impacts of contributions and may encourage participation and employment.
- Pension design matters for incentives and adequacy:
  - Systems can introduce distortions (for example, minimum pension schemes may encourage contributing only the minimum years required).
  - Policies should minimize labor market distortions insofar as trade-offs with adequacy are less acute.
  - Measures to encourage later retirement and instruments to complement public schemes can support pension adequacy.

### COVID-19 implications
- The COVID-19 crisis may place higher demands on future pension reforms:
  - Has contributed to another wave of increase in public debt, although cushioned by lower interest rates.
  - The costs of the crisis have been largely shouldered by younger generations, including in the form of scarring and human capital losses.
- Governments will need to strengthen the resilience of economies—and social security systems in particular—to future shocks.
- The crisis will strengthen calls for ensuring fair, equitable, and sustainable pension systems in the post-COVID-19 world.

### Policy implications and recommended directions
- Address existing imbalances through more equitable burden-sharing across generations.
- Frontload reforms to ensure sustainability and fairness across generations, avoiding excessive phasing that defers costs and deepens intergenerational divides.
- Design reforms to:
  - Minimize labor market distortions.
  - Strengthen links between contributions and benefits.
  - Appeal to shared principles of equity to improve political acceptability.
- Encourage later retirement and promote complementary instruments (private or occupational) to support adequacy.
- Enhance financial literacy to improve public support for reforms and participation.

*Source: Executive Summary, Pension Reforms in Europe: How Far Have We Come and Gone?*

### Introduction and Motivation

### Introduction and Motivation

### Definitions and Principles
- The impact of pension systems on public finances and labor market incentives depends on the system as a whole and is usefully described by six interrelated principles: (1) long-term sustainability; (2) actuarial fairness; (3) intergenerational equity; (4) incentive compatibility; (5) affordability; and (6) adequacy.
- Sustainability: assessed by the ability to ensure that pension benefits are funded by pension contributions without requiring additional transfers. More formally, long-term sustainability requires that the present value of contributions at least equals the present value of benefits.
- Actuarial fairness: a system that provides benefits that fully compensate individuals for their lifetime contributions is actuarially fair. Actuarial fairness is assessed at the cohort level—whether aggregate contributions paid by a cohort during working life are compensated by aggregate benefits received by the same cohort. A system can be sustainable on aggregate without being actuarially fair for cohorts or individuals.
- Intergenerational equity: achieved if the balance between the present value of contributions and benefits does not change across generations. A system can be equitable across generations without being sustainable or actuarially fair.
- Incentive compatibility: design affects incentives to participate and contribute. Systems where lifetime benefits exceed contributions (more than actuarially fair) strengthen participation incentives; when additional contributions are not compensated by equivalent additional benefits (actuarial neutrality absent), systems discourage contributions. Pension contributions may be perceived as distortive taxes when benefits are poorly linked to contributions or because of behavioral market failures (myopia, imperfect information).
- Affordability: high public pension outlays can lead to unsustainable debt dynamics, elevated sovereign risk premia, crowding out private productive spending (normally investment), and can make fiscal adjustment more procyclical and harder to achieve.
- Adequacy: too-low public pension spending may cause deterioration in living standards after retirement and risk of elderly poverty. Pockets of elderly poverty are often more effectively addressed with tools outside contributory pension systems.

### Assessing Sustainability, Fairness, and Intergenerational Equity
- The paper proposes a novel application of the Proportionality Measure (PM) to assess sustainability, fairness, and intergenerational equity of pension systems.
- Actuarial fairness is implied when the present value ratio (present value of benefits to present value of contributions) equals 1. The PM operationalizes this by computing for each cohort the ratio of (undiscounted) sum of pension benefits to the sum of contributions, both expressed in percentage of GDP (equation (3) in the source).
- PM links fairness and sustainability:
  - Under the Aaron-Samuelson condition for a pure PAYG pension system, the fiscally sustainable rate of return equals the growth rate of the contribution base (GDP growth). With r = g in a pure PAYG system, PM = 1 indicates an actuarially fair system for a cohort.
  - A PM persistently above unity signals pension schemes requiring persistent intergenerational transfers.
- PM as a measure of intergenerational equity: a constant PM across cohorts would ensure intergenerational equity (ratio of inputs to outcomes identical across individuals/cohorts).
- Limitations of PM: must be complemented by other measures to assess incentive compatibility, affordability, and adequacy. Examples:
  - A system can reach PM = 1 by excessively raising contributions (poor incentives) or by excessively lowering benefits (inadequacy).
  - A country could finance PM > 1 with permanent transfers for non-contributory benefits without threatening solvency, so context matters.

### The Choice of a Discount Rate
- Key parameter: discount rate (r) at which contributions and benefits are valued and representing rate of return on alternative investments.
- Possible choices: riskless interest rate, market rate of return, rate of time preference of individuals. A relevant public-pension discount rate is the yield on long-term government bonds (reflecting rate at which governments can borrow and the riskiness of government long-term promises such as pensions).
- For cross-country comparisons, a basket of European government bond yields has been suggested to obtain a single risk-free rate.
- The internal rate of return (IRR) is the discount rate at which the sum of discounted lifetime contributions equals the sum of discounted lifetime benefits. Interpretations:
  - IRR = r (equivalent to PM = 1 when r = g in PAYG) implies actuarial fairness.
  - IRR > r (equivalent to PM > 1) indicates lifetime benefits more than compensate relative contribution burden.
  - IRR < r (equivalent to PM < 1) indicates lifetime contributions exceed lifetime benefits.
- For sustainability in a pure PAYG system, the fiscally sustainable rate of return equals the growth rate of the contribution base (g). Thus, in the framework, a sustainable benchmark for discount rate is GDP growth and PM = 1 indicates a sustainable pension system when r = g.

### What PM Level Is Actuarially Fair and Sustainable?
- Fairness and sustainability depend on interplay of r, g, and IRR; the benchmark PM consistent with actuarial fairness and sustainability depends crucially on r–g.
- Key qualitative results:
  - The actuarially fair PM is proportional to r–g:
    - When r > g, actuarially fair PM > 1 (higher interest rates or lower economic growth lower the present discounted value of future benefits relative to contributions, allowing higher benefit rates or lower contributions while maintaining actuarial balance).
    - When r < g, actuarially fair PM < 1.
  - The sustainable PM—for most pension systems in Europe—is inversely related to r–g:
    - With r = g, PM = 1 is both actuarially fair and sustainable.
    - With r > g, sustainability requires the present discounted value of benefits to be lower than contributions on average across cohorts; thus, sustainable PM is typically lower than 1 (pursuing cohort-level fairness need not ensure sustainability).
    - With r < g, the system can provide more than actuarially fair benefits while stabilizing debt.
- Given historical variation in r–g and uncertainty, the paper maintains an overall Europe-wide PM benchmark at 1.0 to assess fairness and sustainability, while acknowledging long periods of r diverging from g can substantively change assessments.

### Table 1 benchmarks (sustainable PM depending on r–g and debt level)
- Table presents benchmark sustainable proportionality measure depending on interest rate – growth rate differential (r-g) and the level of public debt in steady state, assuming steady state contributions equal to [text truncated in source at this point].
- Values from Table 1 (rows: r-g = –2, –1, 0, 1, 2; columns: Debt level = 0, 25, 50, 75, 100, 150):
  - r-g = –2: 1.00, 1.06, 1.12, 1.18, 1.24, 1.35
  - r-g = –1: 1.00, 1.03, 1.06, 1.09, 1.12, 1.18
  - r-g = 0: 1.00, 1.00, 1.00, 1.00, 1.00, 1.00
  - r-g = 1: 1.00, 0.97, 0.94, 0.91, 0.88, 0.82
  - r-g = 2: 1.00, 0.94, 0.88, 0.82, 0.76, 0.65

- Note in the source: benchmark sustainable PM depends on r–g and initial level of debt; most European pension systems have run primary deficits and accumulated high pension system debt, which affects required PM values to stabilize debt.

*Source: prehfwcgea - Introduction and Motivation*

### 8.5 percent of GDP.

### 8.5 percent of GDP.

### Interest rate–growth differentials and sustainability
- Several high-debt countries (including Italy, Portugal, and Spain) experienced notably elevated interest rate–growth differentials in the last decade since the GFC, while r–g has been negative for Germany and most Nordic countries.
- Pension system deficit in Europe currently stands at about 2.5 percent on average.
- By the authors' calculations, pension system deficits accumulated since 1980s (or earliest available year) are of the same order as general government debt for an average European country.
- Sustainability constraints imply using a sustainable PM benchmark below unity that is more constraining than an actuarially fair PM when initial debt and pension system deficits are high and r–g is persistently positive.
- Example: Italy has had a positive r–g at about 2 since 1980; for a debt at about 150 percent of GDP (and zero primary pension deficits) this requires a PM benchmark of about 0.56–0.65.
- For many northern countries where r–g has come down to the vicinity of zero, sustainable steady-state benchmarks center around 1.

### Empirical patterns of r–g (Table 2)
- Table 2 reports interest rate – growth rate differential in selected advanced economies for periods 1950–1979, 1980–1999, 2000–2016, and 1950–2016 (means and medians by country).
- Aggregate "ALL" row: Mean 1950–1979 = 23.8; Median 1950–1979 = 23.3; Mean 1980–1999 = 1.8; Median 1980–1999 = 1.9; Mean 2000–2016 = 0.3; Median 2000–2016 = 0.2; Mean 1950–2016 = 20.2; Median 1950–2016 = 21.2; Mean (?) = 20.9.  
  (Note: table entries preserved as in source.)

### Data sources and key concepts used in operationalization
- Key data source for long-term pension projections: AWG Ageing Report, augmented with historical data from Eurostat and the OECD.
- Demographic data and projections: UN population projections.
- Vintages: data separated into three vintages defined by the 2009, 2018, and 2021 AWG projection rounds (EC 2009, 2018a, 2021).
- Cohorts: defined by birth year; working and pension life lengths determined by calculated effective retirement age; pension expenditures and earnings/contributions are assigned to cohorts using population, employment, and age-earnings profiles.
- Pension contributions and intergenerational transfers: calculations exclude state transfers and other third-party revenues (denoted “state contributions” in AWG definitions) except contributions paid by employees and employers used to finance the pension system.
- Pension benefits: rely on AWG concept covering all public pension benefits, including some non-contributory benefits to the elderly (non-earnings-related social assistance provided to elderly who do not fulfil minimum contribution requirements).
  - Non-contributory minimum pensions in 2019: France 0.2 percent of GDP; Italy 0.3 percent of GDP; Greece 0.1 percent of GDP.
  - Exceptions with large non-contributory components: Denmark 5.7 percent of GDP; Norway 2.6 percent of GDP.
- Robustness tests: sensitivity tests on retirement ages, age-spending and age-income distributions indicate a high degree of robustness of baseline results (see Annex 1 for technical discussion).

### Appropriate discount rate to assess sustainability (Box 1)
- Sustainability implies the pension system's long-term IRR equals the implicit rate of return on contributions or pension wealth; the benchmark depends on pension design and financing mix.
- Pure fully funded DC system: benchmark = market rate of return on accumulated assets.
- Pure PAYG earnings-related systems: benchmark = growth rate of the wage bill (Aaron-Samuelson condition); this is the notional interest rate typically chosen for NDC systems.
- In practice, most European pension systems are mixed (earnings-related PAYG components, tax-financed safety nets, basic/minimum pensions, earmarked taxes/surcharges).
  - Based on the 2021 Ageing Report, only Latvia would be able to cover its pension spending from contributions throughout 2070; Denmark finances its universal public pensions from general taxes on PAYG basis.
- For most European countries the paper uses GDP growth as an approximation to sustainable discount rate in deriving and interpreting the PM.
  - Evidence suggests over the past 25–60 years growth rates of GDP and compensation of employees have been very close in most European countries.
  - For some countries deviations between GDP and wage growth are sufficiently large to affect the PM (examples cited: Czech Republic, Lithuania, Romania; among more developed systems, Greece and Ireland show noticeable differences).

### Linking r–g to PM benchmarks and actuarial fairness (Box 2 and Box Table 2.1)
- When r = g, the PM metric delivers a value of 1 when pension systems are exactly sustainable from general governments’ perspective.
- The higher r–g, the higher the PM value consistent with actuarial fairness (other things equal).
- Illustrative cohort example: contributory period 30 years and retirement period 15 years.
  - When r–g = 0, PM = 1.
  - For r–g between –2 and +2 percentage points, the corresponding reference value for an actuarially fair PM is between 0.65 and 1.56.
- Box Table 2.1 (selected entries preserved):
  - For B = 2* C, contributory/retirement lengths 30/15:
    - r–g = –2 => 0.65
    - r–g = –1 => 0.80
    - r–g = 0 => 1.00
    - r–g = 1 => 1.25
    - r–g = 2 => 1.56
  - For other B/C and lengths the table lists corresponding actuarially fair PMs (table preserved as in source).

### Sustainability versus actuarial fairness: key cases
- Sustainability condition (aggregate intertemporal budget constraint) requires PDV of contributions and benefits to offset initial debt D0 (all variables as ratios of GDP).
- Three general cases with positive initial debt:
  - When r = g: actuarial fairness stabilizes debt; to make debt converge to zero requires higher contributions relative to benefits in PDV.
  - When r > g: a pension system that provides less than actuarially fair benefits can stabilize debt; contemporaneous balances must offset part of interest on initial debt, requiring PDV(benefits) < PDV(contributions).
  - When r < g: a pension system can offer more than actuarially fair benefits and still stabilize debt; providing actuarially fair benefits will result in declining debt-to-GDP ratio.
- The appropriate PM benchmark for sustainability depends on r–g, the size of initial debt, and the PDV of primary balances.

*Italic: Source: prehfwcgea - 8.5 percent of GDP.*

### Box 3. Sustainability and the Proportionality Measure

### Box 3. Sustainability and the Proportionality Measure

### Key relationships between r, g, and the Proportionality Measure (PM)
- When r = g, a PM = 1 is consistent with intertemporal actuarial fairness and does not by itself create dynamics that require adjustment across cohorts.
- When r > g, a PM < 1 is actuarially fair and required for sustainability because higher r relative to g makes initial debt more burdensome, requiring lower benefits or higher contributions over time and on average across cohorts (compared to a case when r = g).
- When r < g, a PM < 1 is actuarially fair and will also ensure sustainability, while the possibility of a Ponzi game could allow for a more than actuarially fair benefits to stabilize debt.

### Determinants of the sustainable PM benchmark
- Lower initial debt tends to increase the sustainable PM benchmark values.
- Higher primary balances tend to increase the sustainable PM benchmark values.
- Conversely, higher initial debt and/or lower primary balances tend to decrease the sustainable PM benchmark values.

### Implications for contributory public pension systems in Europe
- Near all contributory public pension systems in Europe are likely to have a positive initial debt and run primary deficits, requiring a sustainable PM < 1 when r > g.

*Source: Box 3. Sustainability and the Proportionality Measure (from the provided content).*

### Box 3. Sustainability and the Proportionality Measure (continued)

### Box 3. Sustainability and the Proportionality Measure (continued)

### Historical context and drivers
- Over the past three decades, predominantly pay-as-you-go (PAYG) pension systems reached near full coverage and faced mounting expenditure pressures.
- Estimated public pension spending to GDP ratio (AWG definition) increased from about 8½ percent in 1980 to 11¼ percent today in countries with more mature pension systems and reached 12.1 percent in the euro area.
- Before the GFC, pension spending in Europe was expected to increase on average by 2½ percent of GDP between 2007 and 2060, exceeding 13 percent of GDP in 15 countries and reaching as high as 24 percent in 2 countries (EC 2009).
- The GFC and the European debt crisis accelerated reform activity: the number of main pension measures taken increased threefold compared to earlier two decades (Carone and others 2016). Some measures were short-term; deeper parameter reforms were often implemented gradually.
- Political fragilities emerged: some countries reduced contributions to private pensions and redirected these to finance public pensions; some rolled back past reforms during recovery.

### Key reform directions
Eligibility
- Tightening pension eligibility increases contributory periods and lowers benefit periods and raises potential output by expanding the labor force.
- Common measures: increasing early and normal retirement ages; increasing required contributory years for full benefits; tightening early retirement options for special regimes; increasing compulsory retirement ages.
- Most countries increased retirement ages of women more steeply to harmonize with men’s ages.
- Statutory retirement ages in the EU are set to increase from 64 for male and 62 for women in 2008 to about 66½ in 2050 and about 67 in 2070.
- EC (2021) predicts that, for a one-year increase in the statutory retirement age for men (by 2030), the availability of alternative pathways to retirement would lead to an increase in the effective retirement age by about 0.7 years.
- Caveat: a substantial gap exists between statutory and effective retirement ages and this gap is expected to widen; raising contributory requirements risks exacerbating elderly poverty if individuals cannot extend working lives.

Benefit formula
- Discretionary or ad hoc benefit cuts were used during crises (e.g., elimination of 13th and 14th monthly payments in Greece, Hungary, Portugal; consolidation of parallel funds in Greece).
- Parametric reforms (addressing distortions or systemwide design) are often more successful and durable than ad hoc cuts.
- Extending the period over which pensionable earnings are calculated was a key parametric reform (examples: Austria, Finland, France, the Netherlands, Poland, Portugal, Slovak Republic; Greece and Norway followed since the GFC).
- Restricting pensionable earnings to few recent or best-performing years dilutes link between current earnings and future benefits and can lead to excess pension payments benefiting higher earners.
- Temporary pension freezes occurred in Czech Republic, Finland, Greece, Portugal, Slovenia; many countries moved from wage indexation to full or partial price indexation or mixed wage-price valorization.
- Indexation below nominal wage growth penalizes early career years and can contribute to disincentives and unfairness; indexing below inflation erodes purchasing power and adequacy.
- In defined benefit systems, reducing accrual rates has been used (examples: Austria, Greece, Italy, Norway, Turkey; Italy adjusted annuity factors in NDC system).
- Effective average accrual rates have been estimated to fall from 1.6 percent in 2008 to 1.3 in 2050 (Figure 3).
- To protect adequacy, many countries increased or introduced basic or minimum pension allowances (Germany, Greece, Luxembourg, Poland, Slovak Republic, Slovenia, Spain, Sweden, Turkey, United Kingdom) or expanded social safety nets outside the pension system (Italy); some provided targeted benefits through social assistance (Ireland, Sweden, Turkey); few abolished basic pensions (Poland).

Financing
- Revenue measures were used less frequently given already high income and payroll taxes in many countries.
- Stimulus packages during the GFC included reductions in employer social security contributions (often temporary and/or targeted).
- Some countries increased, sometimes temporarily, social contribution rates or introduced temporary levies (Finland, France, Greece, Iceland, Ireland, Italy, Luxembourg, Portugal, Switzerland, United Kingdom); others reduced pension contributions (Germany, Hungary, Lithuania, Sweden).
- On balance, estimated pension contribution burden in advanced economies remained broadly constant since 1990s.
- Reforms sometimes harmonized contribution rates/bases across worker categories (Greece) or changed financing mix to reallocate burden away from employers (Finland, Lithuania in 2019).
- Other measures: increased ceilings on earnings subject to contributions (Slovak Republic), reduced social security tax exemptions (Belgium), reduced tax burden on pension income (United Kingdom).
- Pension income in several countries enjoys preferential treatment despite limited justification to tax pension income differently from other labor income (IMF 2012).
- Some countries earmark tax revenues or set up reserve funds to finance current or projected future pension liabilities.

Automatic adjustments
- Sustainability factors link retirement ages or minimum contributory periods to life expectancy; introduced legislatively in Italy (dating back to 1995 via notional annuity calculation), Denmark, France, and later in Estonia, Finland, Greece, the Netherlands, Portugal and announced in Belgium.
- NDC systems inherently adjust benefits via annuity factors linked to life expectancy (Italy, Latvia, Poland, Sweden).
- Germany’s sustainability factor allocates a share α in the change in pensioner dependency ratio to lowering relative pension level and a share 1–α to raising the contribution rate.
- More elaborate automatic balancing mechanisms respond to wider shocks by altering indexation or contribution rules to steer the system toward financial balance (example: Sweden’s 1998 mechanism linking indexation and notional pension capital to an NPV ratio of system assets and liabilities).
- Examples: Spain’s Pension Revalorization Index (IRP) established in 2013 and suspended in 2018; Luxembourg’s readjustment mechanism foresees partial or zero indexation to real wages when pension system is in deficit.
- Institutional reviews or conditional rules exist in some countries (example: Luxembourg—sustainability review every 10 years with interim assessments after 5 years and resetting contribution rate to maintain pension reserve above 1.5 times annual expenditures; Estonia—government analyzes financial and social sustainability every five years; Greece—parametric adjustments to ensure long-term increase in pension spending would not exceed 2.5 percentage points of GDP over the level in 2009).

### Reform outcomes and quantitative effects
Pension spending
- Reforms enacted since the GFC shaved off nearly half of the increase in pension spending projected in the 2009 Ageing Report and are expected to broadly stabilize pension spending in the long term.
- Pension spending as a share of GDP is expected to increase until about 2040, when pressures from population aging will start to abate.
- The long-term adjustment has been larger in countries with mature pension systems: pension spending projections for 2060 between the 2021 and 2009 Ageing Report vintages decline by nearly 2 percent of GDP in these countries, compared to about ¾ percent of GDP in countries with currently maturing pension systems.

Pension contributions
- On average, pension contributions projections have changed little since the GFC; adjustments have been less systematic (Figure 5).
- Concentration of changes in contribution revenue around zero in the 2021 Ageing Report vintage indicates measures affecting contributions have been small or largely temporary.
- The balance of adjustments in spending and contributions illustrates limitations countries face in raising revenue from contributions when pension systems and/or labor taxes are large.

Economic replacement rates (ERR)
- Compared to the 2009 Ageing Report, the ERR in the 2018 Ageing Report vintage have a higher starting point (around the year 2010) and decline by about 3 percentage points in the long term (Figure 6).
- Adjustment is larger in countries with mature and larger pension systems; in one-third of countries the ERR has increased.
- Diverging replacement rates for current and future retirees indicate wider generational disparities.
- Reform impacts on ERRs are gradual, reflecting grandfathering of the elderly; younger generations will need to work longer and save more to achieve replacement rates similar to current retirees.

Overall assessment with Proportionality Measure (PM)
- Analysis calculates PMs for cohorts born between 1930 and 2000 (individuals who turn age 65 between 1995 and 2065) using 2021 Ageing Report projections (EC 2021).
- Findings:
  - PMs for the current elderly are around or above 2—signaling concerns over sustainability.
  - Post-GFC reform efforts initially nearly halved the gap between actual and sustainable PMs for younger cohorts, reducing the PM to slightly higher than 1.5 (as of 2018 Ageing Report vintage).
  - Subsequent reforms and reversals have eroded part of these gains, increasing the PM for younger cohorts to about 1.7 in the 2021 Ageing Report projections.

### Implications and policy considerations
- Tightening eligibility and strengthening actuarial incentives (penalties/bonuses) are effective to incentivize longer working lives; OECD estimates of actuarially neutral factors at normal retirement age range:
  - Queisser and Whitehouse (2006): 6–9 percent.
  - OECD (2017): 4–7½ percent with 5½ mean.
- Removal of financial disincentives to defer pensions has progressed, but in several countries benefits at normal retirement age do not reach actuarial neutrality and early-retirement incentives remain.
- Parametric reforms (extending pensionable earnings period, reducing accrual rates, adjusting indexation rules, introducing sustainability factors or automatic balancing mechanisms) provide more durable adjustments than ad hoc cuts.
- Protecting lower-income retirees via basic/minimum pensions or social assistance can preserve adequacy but may introduce distortions and disincentives; careful benefit design and work-incentive adjustments (for example, to earnings or asset tests) are important.
- Given limitations in raising contribution-based revenue in many countries, reforms have relied more on adjusting benefits and indexation rules; automatic balancing mechanisms that target system-wide financial balance remain underused but offer broader shock-responsiveness.

*Source: Box 3. Sustainability and the Proportionality Measure (continued), prehfwcgea — PDF chapter excerpt from prehfwcgea - Box 3. Sustainability and the Proportionality Measure (continued).*

### Annex 2 for detailed results. Due to historical data limitations on lifetime

### prehfwcgea - Annex 2 for detailed results. Due to historical data limitations on lifetime

### Key findings on the Proportionality Measure (PM)
- For nearly all countries in Europe the proportionality measure is well above unity.
- For early cohorts who retired around the GFC and the European debt crisis, the PM ranges from 1.4 in Norway and Spain to about 4.5 in Greece, averaging at around 2.0 for countries for whom sufficiently long contribution and benefit history can be constructed.
- Pension systems, as they stand, provide current retirees with lifetime benefits that are two times higher than their lifetime contributions.
- For younger generations, and more characteristic of the pension system’s steady state, the PMs hover around 1.7 on average.
- Only three countries—Estonia, Latvia, and Portugal—could consider their pension systems as sustainable and actuarially fair in the long term: in these countries, for younger cohorts the lifetime contributions broadly balance lifetime benefits.
- Most pension systems are unsustainable on their own contributions and require substantial transfers to maintain financial balance.

### Sustainability benchmarks, debt, and fiscal context
- The paper uses a global Europe-wide benchmark of 1 in determining pension system sustainability.
- Where persistently positive r–g is combined with high debt or persistent pension system deficits, fiscal sustainability may dictate PM benchmarks notably below unity.
- Since most countries in Europe have run pension system deficits for decades (Chapter 5), reducing current and implicit debt would thus require PMs below 1 over the long term.

### Cohort patterns and reform timing
- PMs generally decline over time from high initial levels for older cohorts, indicating pension systems favor early generations.
- Early reformers (Finland, Norway, Sweden) notably reduced PMs for cohorts born in 1930s and 1940s who retired pre-GFC.
- Italy’s pension reforms deliver sustained gains over a longer horizon due to long transition periods.
- Reform impacts generally vanish for cohorts born in the late 1960s when PMs broadly stabilize.
- In Greece and Portugal, reform gains become evident for cohorts born around 1955 who retire after the GFC and European debt crisis.
- At the European level, some countries (notably Ireland and Luxembourg) show PM increases for younger cohorts, offsetting gains elsewhere.

### Impact of reforms since the GFC (vintage comparisons and magnitudes)
- Pre-GFC (2009 Ageing Report vintage): PMs remained broadly stable around 2 on average across most cohorts; pension systems treated generations broadly equally but provided excessive returns (roughly twofold benefits over contributions).
- Reforms enacted since the GFC brought Europe halfway toward achieving sustainability.
  - Comparing 2009 and 2018 Ageing Report vintages for cohorts born in 1990: PMs decline substantially for younger cohorts; PMs declined in two-thirds of countries.
  - Cross-country variation in PMs (measured by standard deviation) for younger cohorts declined by half.
  - Long-term convergence in the mean and median PMs toward 1.6 and 1.5, respectively, in the 2018 vintage.
  - Even after 2018 reforms, pension systems in Europe had reached only halfway toward achieving sustainability.
- By the 2021 Ageing Report vintage:
  - For younger generations the PMs average around 1.7, indicating the distance toward sustainable PM benchmarks has declined by about one-third when compared to the 2009 Ageing Report benchmark as well as to the older generations.
  - Subsequent reforms and reversals since 2018 have eroded some earlier gains; recent years preceding COVID-19 witnessed a slowdown in reform momentum and several reversals.

### Reforms, reversals, and policy changes (selected examples and quantitative impacts)
- Retirement ages:
  - Capping, postponing, reversing, or abolishing earlier increases in retirement ages occurred in multiple countries (Czech Republic, Italy, the Netherlands, Poland).
  - Germany and Portugal allowed earlier retirement for the insured with long contributory history.
- Pension benefits:
  - Greece reversed the 2016–17 reforms aligning benefit formulas of existing retirees with new retirees; reinstated 70 percent entitlement of the deceased’s pension (instead of reformed 50 percent); increased accrual rates for longer contributory periods after a 2019 court ruling; provided an Easter bonus for retirees in 2019.
  - Hungary reintroduced a 13th pension; Italy introduced a loosely targeted 14th pension payment and provided temporary cash benefits to elderly workers.
- Private pension and reserved funds:
  - Hungary closed mandatory private DC funds and transferred almost all assets to the state.
  - Ireland transferred the National Pension Reserve Fund to the state.
  - Baltic states and Poland temporarily or permanently transferred some contributions from DC accounts to the state pension pillar.
  - Estonia allowed optional dismantling of privately managed DC plans returning assets to individual account holders.
- Automatic adjustment mechanisms:
  - Italy cancelled the automatic adjustment of retirement ages to life expectancy for 2019–20; similar cancellations or softenings occurred in the Slovak Republic and the Netherlands.
  - Spain postponed introduction of the sustainability factor and temporarily suspended the automatic adjustment mechanism linking pensions to the financial balance of pension and social security system.
  - Permanent removal of such mechanisms would raise long-term PMs substantially (for Spain, removal of IRP would raise the long-term PM by about half index points as a cumulative effect).
- Pension contributions and financing mix:
  - Lithuania (2019 reform): eliminated nearly all employer social contributions (28.9 percent of wages, of which 22.7 percent of wages was earmarked for pensions); raised wages by 1.289 to keep gross wages constant; employee contributions set to 8.72 percent of the new wage with remainder financed by personal income taxes; reform of the second pillar led to an upward revision in public pension spending projections (EC 2021).
  - Hungary reduced social contributions from 27 to 15.5 percent in four steps between 2017–20.
  - Greece reduced pension contributions for the self-employed and provided a temporary subsidy on employers’ pension contributions for youth.
- Specific quantified impacts:
  - Greece: planned recalibration of pensions of existing retirees had been in the order of 1 percent of GDP before reversal.
  - Slovakia: cancellation of the automatic adjustment mechanism is the largest contributor to an increase in the long-term PM by nearly ¾ points.
  - Hungary: permanent reintroduction of the 13th pension and residual reduction in contribution rates contributed to PM increases.

### Distributional and intergenerational equity implications
- Reforms generally favored the current elderly over the current youth; measures implemented after the GFC often fully protected older generations at the expense of younger cohorts.
- Some reform strategies promoted greater equity across generations (example: Greece’s initial accelerated 2016 reform with full application to current retirees before reversal; Latvia’s 1996 NDC public pension scheme applied to entire working-age population from age 15 and converted past rights into NDC capital at reform date).
- Recent reversals and benefit restorations have tended to benefit older cohorts, eroding earlier progress toward intergenerational equity.

### Measurement and additional metrics
- Real internal rates of return (IRRs) provide a consistent assessment of sustainability across countries and Ageing Report vintages; IRRs derived under the same assumptions as PMs are reported in Annex 3.
- Changes in Ageing Report vintages capture not only reforms but also changes in demographic and macroeconomic assumptions, modeling improvements, and coverage changes; not all changes in PMs across vintages can be attributed solely to reforms.

*Source: Annex 2 excerpt from prehfwcgea - Annex 2 for detailed results. Due to historical data limitations on lifetime (PDF chapter/section).*

### 3.4 percent in Germany to 7 percent in Greece, exceeding the average real GDP growth

### 3.4 percent in Germany to 7 percent in Greece, exceeding the average real GDP growth

### Box 4 — Real Internal Rates of Return (IRRs) of Pension Systems: key findings
- IRRs for cohorts in several countries exceed contemporaneous real GDP growth: "3.4 percent in Germany to 7 percent in Greece," exceeding the average real GDP growth "by about 1.8 and 6 percentage points, respectively."
- At steady state (for younger birth cohorts born from 1980 onward) Estonia and Latvia have the lowest IRRs ("less than 1 percent") which are below their long-term GDP growth rates.
- IRRs based on the 2021 Ageing Report vintage are generally lower for younger cohorts compared to IRRs based on the 2009 Ageing Report, but they "still remain notably above their country-specific GDP growth rates."
- Interpretation: despite reforms over the past decade, "in most European countries more efforts are needed to ensure sustainability of pension schemes."

### Incentives, design features, and behavioral effects
- Aggregate metric limitation:
  - The Pension Model (PM) is "an indicator of sustainability and fairness" but masks within-generation imbalances linked to incentive compatibility, adequacy, and affordability.
- Actuarial adjustments and retirement decisions:
  - Many countries apply actuarial adjustments that are "less than what is required for actuarial neutrality" (examples: Belgium, Bulgaria, Italy (DB scheme), Luxembourg, Malta, Romania).
  - Penalities above actuarial neutrality (examples: Portugal, Spain) can delay claiming but risk inadequate pensions for those forced to exit early.
- Minimum pensions and years of contributions:
  - Several countries (Belgium, Greece, Latvia, Italy, Spain, Sweden) provide minimum pensions conditional on a minimum number of years of contributions.
  - For low earners, minimum pensions can create incentives "not to contribute beyond the minimum required years"; evidence of bunching at the minimum years of contribution is noted (Greece example before 2016–17 reforms: strong incentive to retire after 15 years).
- Interactions with other social programs:
  - Pension systems interact with means-tested social assistance, unemployment insurance, parental leave; these interactions can create lifetime incentives to limit contributions (Montenegro example: higher returns for those with fewer years of contributions and lower earnings; incentives to limit contributions to no more than 15 years).

### Are social contributions distortive? Financing and labor market impacts
- Pension financing metrics:
  - "Average payroll tax rate earmarked for pensions is 22 percent" (Europe).
  - "Average collection is 7 percent of GDP in revenue—about 0.32 percentage points of GDP per point of payroll tax."
  - Contributions cover on average "70 percent of pension spending," with the remainder largely financed by general revenue.
- Labor market effects:
  - Taxes on labor create a wedge between labor cost and take-home pay, affecting labor demand and supply.
  - Cross-country and empirical evidence: a "10 percentage points change in the after-tax wage can raise employment by 2–5 percent (IMF 2012)."
  - Macro evidence shows a negative association between employment rates and pension contribution rates; micro evidence indicates stronger responses at the extensive margin for low-income workers, women, single parents, second earners, and older workers.
- Contributions perceived as taxes vs. savings:
  - Where pension contributions are perceived as actuarial savings (strong contribution–benefit link), distortionary effects on labor supply can be offset.
  - A contemporaneous share of contributions in pension spending can proxy the "savings" component; this share is positively though weakly correlated with employment.

### Pension adequacy and poverty mitigation
- Role of public pensions:
  - "Old-age pension accounting for about 72 (85 including survivors’ pensions) percent of single-person retirees’ total income in the EU on average."
- Trade-offs and projected trends:
  - Past reforms aimed at sustainability imply "expected gradual reduction of replacement rates" which may weigh on adequacy.
  - To offset reductions in replacement rates, many countries expanded basic or minimum pension allowances or provided benefits to retirees outside the pension system.
  - Trade-off: "reducing overall replacement rates while increasing the protection at the lower tail of the benefit distribution necessarily implies flattening of pensionable earnings—benefits profile," which raises efficiency distortions.
- Empirical relationship:
  - "The generosity of pension systems (measured by replacement rates) is inversely linked with poverty risks at retirement."
  - Incidence of elderly poverty varies: advanced European economies with mature pension systems tend to have lower elderly poverty relative to working-age population; less mature systems often show the opposite.
  - Poverty depth of the elderly remains on average below that of the working-age population throughout nearly all of Europe.

### Policy-relevant implications and considerations (derived from analysis)
- Strengthen actuarial neutrality where adjustments are insufficient to avoid incentives for early claiming.
- Reassess minimum-pension and minimum-contribution thresholds to avoid creating perverse bunching incentives among low earners.
- Improve the perception and reality of a contribution–benefit link (increase the "savings" component) to mitigate distortionary labor market effects of payroll taxes.
- Balance sustainability and adequacy: when lowering replacement rates for fiscal sustainability, expand targeted minimum or basic pensions to protect low-income retirees while minimizing flattening of earnings–benefits profiles that reduce incentives to contribute.
- Consider interactions across social programs (means-tested benefits, unemployment insurance, parental leave) to prevent unintended incentives to limit contributions or exit the labor force prematurely.

*Sources: 2009 and 2021 Ageing Report; Eurostat; and IMF staff calculations.*

### 1. At Risk of Poverty Rate

### 1. At Risk of Poverty Rate

### Definition
- "At Risk of Poverty Rate" — (Percent of population)

### Measurement
- The metric is measured as a percent of the population.

*Source: prehfwcgea - 1. At Risk of Poverty Rate*

### 2. Poverty Depth

### 2. Poverty Depth

### Poverty by tenure and age
- Homeownership and imputed income:
  - On average, the elderly are more likely to be homeowners compared to younger people: in OECD, in 2011 on average 77 percent of heads of households aged 55 and older were homeowners compared to 60 percent among those younger than 45 (OECD 2013).
  - Taking into account imputed rent as an unearned income of households significantly reduces the risk of poverty for retirees.
- Risk of poverty and social exclusion (AROPE):
  - In the majority of European countries, AROPE is much more prevalent among the working-age population when compared to the elderly.
  - The AROPE gap between the working-age population and the elderly is especially high in advanced economies, reaching nearly 4.5 percentage points in 2019 for the euro area or countries with more mature pension systems.
- Income sources for retirees:
  - Additional sources (work income, imputed income from home ownership, rental, dividends, interest, profits) vary with household characteristics; work income accounts for about 4 percent of total income for a single retiree person, and almost 16 percent in a household consisting of a couple with at least one member aged 65 and older.

### Adequacy, incentives, and pension design
- Replacement and benefit trajectories:
  - Replacement and benefit rates in public pension systems are set to decline in decades to come; the EC 2018 Pension Adequacy Report projects that pension systems will be on average less adequate in 2056 compared to what was observed in 2016.
- Policy priorities within contributory systems:
  - Prefer to provide income support outside contributory public pension systems in earnings-related systems aiming for strong contribution–benefit links.
  - Priority: ensure incentives (especially at lower-income levels) to lengthen formal work careers and increase retirement ages to increase pension income at retirement.
- Addressing low public pensions:
  - Low public pensions reflect lack of employment income, short careers, or informal work. First-order importance: policies to reduce structural unemployment and incentivize formal participation.
- Options for countries with proportionality measures below 1:
  - Such countries (cohorts receive lifetime benefits below lifetime contributions) may afford to increase pension expenditures while improving incentives and adequacy, provided long-term sustainability is ensured.
- Role of private pensions:
  - Governments could explore measures to encourage workers to save through private pension schemes to complement public pensions and diversify income sources.
  - Caveat: Private pension schemes might be costly to participants and provide limited risk-diversification, highlighting the importance of strong regulation.

### Financing high pension spending and short-term imbalances
- Current deficits and historical peaks:
  - The average pension system deficit in Europe is currently about 2½ percent of GDP after peaking at about 3 percent of GDP in the aftermath of the European sovereign debt crisis—double pension system deficits before the GFC.
  - This compares to a primary general government deficit of about ½ percent of GDP that countries in the sample ran on average over the past two decades and before the COVID-19 pandemic.
- Projections and spending levels:
  - According to AWG projections, pension reforms have contained average pension spending around or above 12 percent of GDP in Europe between now and 2070.
  - Average pension spending was 9 percent of GDP in 2000.
  - The number of countries with pension spending over 10 percent of GDP is projected to increase from 9 out of 27 in 2000 to 18 out of 28 in 2050.
  - Annual Europe-wide pension system deficit will stabilize at about 4 percent of GDP by 2050s.
- Cumulative deficits and sovereign debt:
  - Cumulative pension deficits since the 1980s are on the same order of magnitude as gross sovereign debt (before the COVID-19 crisis).
  - Within the next 20 years, pension system deficits increase mostly in countries with mature pension systems; after the 2040s, spending pressures are stronger in currently maturing systems.
- Links to macro performance and crowding out:
  - Historical association: countries that spent higher shares of national income on pensions tended to experience lower real growth rates (Figure 24; regression R^2 = 0.7225 as presented).
  - Countries with relatively high pension spending tend to have higher deficits and debt (Figure 25; regression R^2 = 0.459 as presented).
  - Pension spending has been negatively associated with gross fixed capital formation since euro inception (Figure 26; regression R^2 = 0.2361 as presented).
  - Consequences: pension spending may displace productive spending, contribute to fiscal vulnerabilities, limit shock-response scope, crowd out public and private investment, and weigh on long-term growth.
- Reform reversals and examples:
  - Several emerging Europe countries could not sustain persistent pension deficits tied to private pension deviations, leading in some cases to reversals (example: Hungary closed the mandatory private scheme introduced in 1998; in 2010 Hungary eliminated obligatory participation and made return to public scheme possible; by 2011 Hungary closed the mandatory private scheme).

### Political economy considerations
- Aging and median voter pressures:
  - Old-age dependency rates are projected to almost double in several countries; by 2040 the median age projected to reach 52 years for men and 55 years for women—an increase by 19 years compared to early 1990s (Carone and others 2016).
  - As populations age, median-voter pressure to roll back reforms is likely to intensify.
- Risks of reform reversal:
  - Reversals are more likely where pension systems are perceived as treating generations unfairly (declines in proportionality measures). Gradual or phased-in reforms that protect current retirees can increase the ease of reversals.
- Enfranchising younger cohorts and political participation:
  - Greater political engagement of the young can induce fiscal discipline; evidence: a 1 percent increase in young voter turnout increases allocation to higher education by 0.77 percent (Bertocchi and others 2020).
  - Policies to enfranchise young voters (preregistration, lowering voting age) could improve acceptability and durability of reforms.
- Financial literacy and institutions:
  - Fornero and Lo Prete (2019): no robust average relationship between pension reforms and re-election in advanced economies, but the electoral impact is contingent on economic and financial literacy—reforms can have electoral costs where literacy is very low and benefits where literacy is very high.
  - Executive constraints, political competition, and inter-jurisdictional fiscal federalism increase likelihood of reforms (Verbič and Spruk 2019).
- Reform design and shared principles:
  - Reforms that strengthen contribution–benefit links and are perceived as equitable are more durable; categorical or redistributive retrenchments lacking broad backing risk reversal.
  - Dedicated commissions and broad societal dialogue can facilitate reform implementation and durability, though they take time to build.

### Policy implications and recommendations
- Maintain reform momentum:
  - Annual pension system deficits currently about 2½ percent of GDP are projected to increase to 4 percent of GDP in the next three decades—this is not a time to stall reform momentum.
  - Authorities need to diligently implement further reforms and ensure a robust fiscal position, especially in countries with high pension spending and high government debt.
- Address intergenerational divides:
  - Younger generations face lower lifetime pension balances; the authors estimate initial impact of reforms reduced average PM from above 2 for the elderly today to nearly 1.5 for those in their 20s today, reverting to about 1.7 supported by reform reversals.
  - Resolute reform implementation should be accompanied by measures to increase opportunities for youth: structural labor market reforms, greater protection against poverty at younger ages, investment in education and training, and reforms to non-pension benefits (including unemployment insurance).
- Burden-sharing:
  - Addressing imbalances ideally requires burden-sharing by both current and future retirees; burden-sharing by current retirees is more important the higher current pension outlays or public debt.
- Balance sustainability, efficiency, and adequacy:
  - Design pension benefits to minimize labor market distortions (e.g., actuarially neutral adjustments for early retirement).
  - Where social/minimum pensions provide poverty alleviation, ensure careful interlinking with earnings-related pensions to smooth participation tax rates and avoid bunching effects.
  - Consider income support outside pensions to preserve contribution–benefit links while addressing poverty.
  - Develop private pension schemes with strengthened regulation and consumer protection to diversify retirement income.
- Improve reform durability via design and education:
  - Reforms emphasizing shared principles of equity and fairness and enhancing financial literacy can reduce pressures to reverse reforms and improve political acceptability.

### Data sources, definitions, and technical assumptions
- Projection vintages: follow the 2009, 2018, and 2021 versions of the Ageing Report (EC 2009, 2018a, 2021).
- Population data: historical and forecasts from the UN population database (medium variant).
- Number of pensioners for 2007–(2060) 2070 taken from respective Ageing Report vintages and extended backward until 2005 with Eurostat growth rates of number of retirees.
- Workers defined as population aged 22 to the calculated effective retirement age; everyone above the calculated retirement age considered a pensioner.
- Effective retirement age: calculated by distributing the number of pensioners for each year to the population of that year starting from age 100 until all pensioners accounted for; extended backwards with growth rates of statutory retirement ages from the OECD for 1950–2005 and kept constant for (2060) 2070–2100.
- Model sensitivity: tested to labor market exit age as reported in respective Ageing Report vintage, supplemented with historic OECD data, yielding very similar results.
- Note on retirement age changes: changing retirement age adjusts lengths of working and retirement lives but does not alter annual contributions and benefits taken from Ageing Report vintages and has very little impact on lifetime contributions and benefits of cohorts (explanatory notes provided in text).

*Source: https://www.imf.org/-/media/files/publications/dp/2021/english/prehfwcgea.pdf — Chapter 2: Poverty Depth.*

### Annex 1. Data Sources and

### Annex 1. Data Sources and Technical Assumptions

### Pension contributions and expenditures: sources and temporal extensions
- Pension contributions for 2007 – (2060) 2070 follow the respective Ageing Report vintages while netting out any state transfers and third-party contributions as defined in the 2018 and 2021 Ageing Report vintages.
- Pension contributions are extended backward with growth rates of total social contributions from the OECD for 1965–2016.
- Where OECD data was not available, pension contributions are assumed constant as a share of GDP prior to 1970 and after (2060) 2070 until 2100.
- Pension expenditures for 2007 – (2060) 2070 follow the respective Ageing Report vintages.
- Pension expenditures are extended backward with growth rates of actual pension spending data from Eurostat for 1990–2007 and from the OECD for 1980–90.
- Pension expenditures are extended forward for (2060) 2070–2100 with growth rates of old-age dependency ratio.

### Cohort definitions and lifetime calculations
- Cohorts are defined by their birth year.
- Cohort pension contributions and cohort pension expenditures are calculated by adding all contributions and expenditures over the lifetime of a cohort (22–100).
- The proportionality measure for a given cohort is calculated by dividing lifetime cohort benefits received by lifetime cohort pension contributions paid.

### Distribution of contributions and expenditures across cohorts
- Distribution of pension contributions to different cohorts for a given year:
  - Assigned based on an interaction term between the size of the respective cohort in total workers and country-specific age-income profiles over the working life of a worker.
  - Age-income profiles are derived from the Eurosystem Household Finance and Consumption Survey microdata (wave 3).
  - Age-income profiles are kept constant before the age of 20 and after the age of 67 when smaller sample size can give rise to a selection bias.
- Distributions of pension expenditures to different cohorts for a given year:
  - Assigned based on the size of the respective cohort in the total number of pensioners that underlies reported baseline results.
  - Robustness test: pension expenditures were further distributed based on a notional age profile, constructed by dividing spending after retirement by a cohort-specific index of compounded real GDP growth rates.
  - Purpose of adjustment: by distributing pension spending from older to younger retirees this adjustment allows for declining spending profile as retirees get older since pensions in payment in most countries are indexed with less than nominal wage growth.
  - Result: The resulting PM values show minimal changes for mature pension systems and for all European countries covered decline toward zero in the long term.

### Interpolation and aggregation methods
- Linear interpolation is applied to data between 5-year forecasts of the Ageing Report.
- Country and time series sources explicitly cited in this annex: Ageing Report vintages; Eurostat; OECD; Eurosystem Household Finance and Consumption Survey (wave 3).

### Stability and magnitude of age-distribution effects
- The age distribution of pension spending remains relatively stable over time and across cohorts.
- Since the total annual pension spending in the model is constant, given by the Ageing Report vintages, the impact on the PM is minimal for mature pension systems and noticeable only for less mature pension systems at their transition period.
- For mature pension systems the historical impact (for older cohorts) is on average less than 0.15 index points.
- For all countries the impact declines in the long term, averaging at about 0.05 for younger cohorts.

*Source: Annex 1, “Data Sources and Technical Assumptions,” PENSION REFORMS IN EUROPE: HOW FAR HAVE WE COME AND GONE (IMF staff).*

### References

### References

### Books and book chapters
- Galasso, Vincenzo. 2006. The Political Future of Social Security in Aging Societies. Cambridge, MA: MIT Press.
- Gruber, Jonathan, and David Wise. 2004. Social Security Programs and Retirement around the World. Chicago, IL: University of Chicago Press.
- Holzmann, Robert, Edward Palmer, Robert Palacios, and Stefano Sacchi. 2020. Progress and Challenges of Nonfinancial Defined Contribution Pension Schemes: Volume 1. Addressing Marginalization, Polarization, and the Labor Market. Washington, DC: World Bank.
- Kohli, Martin, and Camila Arza. 2011. “The Political Economy of Pension Reform in Europe.” In Handbook of Aging and the Social Sciences, edited by R. H. Binstock and L. K. George, seventh edition, 251–64. San Diego, CA: Academic Press.
- Lindbeck, Assar, and Mats Persson. 2003. “The Gains from Pension Reform.” Journal of Economic Literature XLI: 74–112.
- Palmer, Edward, Sandra Stabina, Ingemar Svensson, and Inta Vanovska. 2006. “NDC Strategy in Latvia: Implementation and Prospects for the Future.” In Pension Reform: Issues and Prospects for Non-Financial Defined Contribution Schemes, edited by Robert Holzmann and Edward Palmer, 397–424.

### Journal articles
- Fornero, Elsa, and Anna Lo Prete. 2019. “Voting in the Aftermath of a Pension Reform: The Role of Financial Literacy.” Journal of Pension Economics & Finance 18 (1): 1–30.
- Freudenberg, Christoph, Natalie Laub, and Tim Sutor. 2018. “Pension Decrement Rates across Europe – Are They Too Low?” The Journal of the Economics of Aging 12: 35–45.
- Gruber, Jonathan, and David Wise. 1998. “Social Security and Retirement: An International Comparison.” The American Economic Review 88 (2): 158–63.
- Jiménez-Martín, Sergi. 2014. “The Incentive Effects of Minimum Pensions: Extended Version.” Working Paper 2014–04, FEDEA.
- Jordà, Òscar, Katharina Knoll, Dmitry Kuvshinov, Moritz Schularick, and Alan M. Taylor. 2019. “The Rate of Return on Everything, 1870–2015.” The Quarterly Journal of Economics 134 (3): 1225–98.
- Knell, Markus. 2010. “How Automatic Adjustment Factors Affect the Internal Rate of Return of PAYG Pension Systems.” Journal of Pension Economics and Finance 9 (1): 1–23.
- Nickell, Stephen. 2003. “Employment and Taxes.” CESifo Working Paper 1109, Center for Economic Studies and ifo Institute, Munich.
- Samuelson, Paul A. 1958. “An Exact Consumption-Loan Model of Interest With or Without the Social Contrivance of Money.” Journal of Political Economy 6: 467–82.
- Song, Zheng, Kjetil Storesletten, and Fabrizio Zilibotti. 2012. “Rotten Parents and Disciplined Children: A Politico-Economic Theory of Public Expenditure and Debt.” Econometrica 80 (6): 2785–803.
- Soto, Mauricio. 2017. “Pension Shock.” Finance & Development 54 (2): 14–15.
- Tabellini, Guido. 1991. “The Politics of Intergenerational Redistribution.” Journal of Political Economy 99 (2): 335–57.
- Verbič, Miroslav, and Rok Spruk. 2019. “Political economy of pension reforms: an empirical investigation.” European Journal of Law and Economics 47: 171–232.
- Wilke, Christina Benita. 2005. “Rates of Return of the German PAYG System – How They Can be Measured and How They Will Develop.” MEA Discussion Paper Series, Munich Center for the Economics of Aging, Munich.
- Wong, Boris. 2016. “Formal Sector Labor Supply Responses to the 2008 Chilean Pension System Reform.” Manuscript, University of California San Diego, San Diego, CA.
- Zimčík, Petr. 2017. “Tax Wedge in OECD Countries - A New Evidence.” Organisation for Economic Co-operation and Development, Paris.

### Working papers and policy papers
- Égert, Balázs. 2012. “The Impact of Changes in Second Pension Pillars on Public Finances in Central and Eastern Europe.” OECD Economics Department Working Paper 942, Organisation for Economic Co-operation and Development, Paris.
- Knell, Markus. 2005a. “Demographic Fluctuations, Sustainability Factors and Intergenerational Fairness – An Assessment of Austria’s New Pension System.” ONB Monetary Policy & The Economy Q1/05, Oesterreichische Nationalbank, Vienna, Austria.
- Knell, Markus. 2005b. “On the Design of Sustainable and Fair PAYG Pension Systems When Cohort Sizes Change.” ONB Working Paper 95, Oesterreichische Nationalbank, Vienna, Austria.
- Mauro, Paolo, and Jing Zhou. 2020. “r-g<0: Can We Sleep More Soundly?” IMF Working Paper 20/52, International Monetary Fund, Washington, DC.
- Queisser, Monika, and Edward Whitehouse. 2006. “Neutral or Fair? Actuarial Concepts and Pension-system Design.” OECD Social, Employment and Migration Working Paper 40, Organisation for Economic Co-operation and Development, Paris.
- Szikra, Dorottya. 2018. “Reversing Privatization and Re-Nationalizing Pensions in Hungary.” ESS Working Paper 66, International Labor Organization, Geneva.
- US Government Accountability Office. 2008. “Entitlement Reform Process: Other Countries’ Experiences Provide Useful Insights for the United States.” Report to the Committee on the Budget, US Senate, GAO-08–372, Washington, DC.
- Velasquez, Agustin, and Svetlana Vtyurina. 2019. “How does Taxation affect Hours Worked in EU New Member States?” IMF Working Paper 19/130, International Monetary Fund, Washington, DC.

### Institutional publications, guides, and reports
- European Parliament. 2007. “The Role of Minimum Income for Social Inclusion in the European Union.” Policy Department Economic and Scientific Policy, European Parliament, Brussels.
- Eurostat and European Central Bank. 2011. “Technical Compilation Guide for Pension Data in National Accounts,” Eurostat Methodologies and Working Papers, Luxembourg.
- Holzmann, Robert. 2002. “The World Bank Approach to Pension Reform.” International Social Security Review 53 (1): 11–34.
- International Monetary Fund (IMF). 2011. “The Challenge of Public Pension Reform in Advanced and Emerging Economies.” IMF Policy Paper, Washington, DC.
- International Monetary Fund (IMF). 2012. “Fiscal Policy and Employment in Advanced and Emerging Economies.” IMF Policy Paper, Washington, DC.
- International Monetary Fund (IMF). 2017. “Reforming the Greek Pension System.” IMF Country Report 17/41, Washington, DC.
- International Monetary Fund (IMF). 2019. “The Pension System in Luxembourg.” IMF Country Report 19/131, Washington, DC.
- Organisation for Economic Co-operation and Development. 2005. “Pensions at a Glance 2005: Public Policies Across OECD Countries.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2007. “Pensions at a Glance 2007: Public Policies Across OECD Countries.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2009a. “Pensions at a Glance 2009: Retirement-income Systems in OECD Countries.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2009b. “The Jobs Crisis: What Are the Implications for Employment and Social Policy?” In OECD Employment Outlook 2009: Tackling the Jobs Crisis, 17–115. Paris: OECD Publishing.
- Organisation for Economic Co-operation and Development. 2011a. “Pensions at a Glance 2011: Retirement-income Systems in OECD and G20 Countries.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2011b. “Taxation and Employment.” OECD Tax Policy Studies 21.
- Organisation for Economic Co-operation and Development. 2013. “Pensions at a Glance 2013: OECD and G20 Indicators.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2015. “Pensions at a Glance 2015: OECD and G20 Indicators.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2017. “Pensions at a Glance 2017: OECD and G20 Indicators.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2019. “Pensions at a Glance 2019: OECD and G20 Indicators.” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development. 2020. “Guidelines for the OECD Table on Social Insurance Pension Schemes (Table 2900).” OECD Publishing, Paris.
- Organisation for Economic Co-operation and Development and International Monetary Fund. 2015. “Fiscal Policy and Growth: Why, What and How?” OECD Publishing, Paris.
- US Social Security Administration. 2018. “Social Security Programs Throughout the World: Europe.” Washington, DC.
- US Social Security Board of Trustees. 2020. “The 2020 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds.” Washington, D.C.
- Velasquez, Agustin, and Svetlana Vtyurina. 2019. “How does Taxation affect Hours Worked in EU New Member States?” IMF Working Paper 19/130, International Monetary Fund, Washington, DC.

*Content extracted from the References section of the provided PDF.*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2021/english/prehfwcgea.pdf_