## 1. Automatic Adjustments of Pension Benefits

## Source details

**Canonical URL:** [1. Automatic Adjustments of Pension Benefits](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16242.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16242.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16242.pdf.json)

---

### I. Introduction — key premises and purpose
- Asian economies are aging fast, with some countries already having high dependency ratios (Japan) and others facing sharp increases in the next couple of decades (China, Korea, Singapore).
- Two main implications:
  - Pension system deficits can increase very quickly, limiting room for policy action and hampering fiscal sustainability.
  - Projected aging of populations can have large intergenerational equity implications.
- Purpose: explore how incorporating Automatic Adjustment Mechanisms (AAMs)—rules ensuring certain pension system characteristics respond to demographic, macroeconomic and financial developments in a predetermined fashion and without the need for additional intervention—can be part of pension reforms in Asia.
- Historical note: price indexation of benefits introduced in Denmark in 1922 (Gerig, 1960).

### II. Structural framework of AAMs — design features and mechanics
- Four main structural features elevated into legislation:
  - Trigger variables (indicators determining when/how adjustments are made).
  - Adjusting parameters linked to trigger indicators.
  - Frequency of adjustment.
  - Boundaries of adjustments.
- Trigger variables (examples and country links):
  - Ex-ante (prediction-based): age-specific/cohort-specific life expectancy; system dependency (ratio of pensioners to workers); contributory years; projected deficits; benefit indexation; annuity divisor.
    - Finland, Portugal: age-specific life expectancy.
    - Japan: system dependency.
    - Germany, Canada: contributory years.
    - Sweden, Latvia, Poland, Italy: annuity divisor adjustments.
  - Ex-post (state-of-the-world): observed improvements in life expectancy; backward-looking benefit indexation.
- Adjustment parameters (three broad categories):
  - Benefit levels: benefit indexation, valorization of past earnings, reductions reflecting life expectancy improvements or financing shortfalls (examples: Finland, Portugal, Germany, Sweden, Italy, Poland, Latvia, US).
  - Eligibility criteria: retirement age adjustments to reflect life expectancy improvements (examples: Canada, Greece, Finland, Portugal, Denmark, Czech Republic, France).
  - Contribution rates: rarely used automatically; examples include Germany (automatic adjustments if contribution rates fall behind required level) and Canada (contribution rates act as operating variable and trigger; failure to act leads to temporary suspension of indexation).
- Frequency of adjustments:
  - Determined by law; may be annual or less frequent. Infrequent reviews can lead to larger adjustments.
- Boundaries of adjustments:
  - Must specify frequency, individual adjustment limits, compounded impacts, and legal precedence when multiple AAMs coexist.
  - Examples: delay indexation until compounded price inflation reaches threshold; temporary replacement of automatism by discrete decisions if indicators exceed certain levels.

### III. Trade-offs, caveats and considerations
- AAMs face same trade-offs as ad-hoc reforms: fiscal sustainability versus adequacy and equity.
- Benefits of AAMs:
  - Predictability, simplicity, higher likelihood of timely parametric changes, reduced procrastination, improved intergenerational equity, and signaling policymakers’ commitment.
- Caveats:
  - Many AAMs were introduced after major parametric reforms; the true test is whether AAMs operate when indicators call for adjustments.
  - Prediction-based AAMs need feedback mechanisms to correct forecast errors; state-of-the-world AAMs imply delayed reaction and potentially larger/faster adjustments.
- Equity and adequacy concerns:
  - Automatic price indexation may erode pension levels relative to wages over time.
  - Deflation can raise relative pension levels if indexation is not designed for deflationary contexts.
  - Raising retirement ages mechanically may disadvantage groups with lower life expectancy or limited capacity to work longer.
- Interaction with labor markets and institutions:
  - AAMs affect labor supply, savings, and human capital investment; timely information provision and smooth, gradual adjustments are important to enable private responses.

### IV. Applicability in Asian pension systems — system types and policy implications
- Pension system composition in Asia (examples):
  - Defined contribution systems: Hong Kong, Singapore, Australia, Malaysia, Indonesia.
  - Defined benefit systems: Japan, Korea, Thailand, Vietnam.
  - Mixed elements: China, India.
- Aging trajectories and grouping:
  - Aging imminently or already underway: Korea, Hong Kong SAR, Singapore, China, Thailand, Brunei; Japan already aged.
  - Moderate or slow aging: Vietnam, Malaysia (moderate); India, Indonesia, Mongolia, Cambodia (slow).
- Policy implications by country group:
  - Countries aging fast: AAMs can slow accumulation of imbalances but must be complemented with parametric and structural reforms.
  - Relatively young countries: easier to link system parameters to sustainability indicators early; incorporate automatism at system design stage (e.g., NDC schemes).
- Limitations and complementarities:
  - In contexts with adequacy constraints, scope to reduce benefits may be limited, increasing reliance on eligibility adjustments (retirement age).
  - Raising contribution rates can affect labor market and growth; other adjustments should be prioritized when possible.
- Fragmentation:
  - Fragmented systems (separate schemes for civil service, military, state-owned enterprises) complicate AAM application and may require consolidation before adopting AAMs.

### V. Case studies — Japan, Korea, China (high-level findings and mechanics)

- Japan: existing AAM and reform options
  - Universal public pension introduced in 1961; defined-benefit, largely pay-as-you-go.
  - 2004 comprehensive reform introduced “macroeconomic indexing” and gradual increase in pension contributions.
  - Contributions being gradually increased until fiscal year 2017, after which they are fixed.
  - Historical spending:
    - Public pension spending increased from 5.2 percent of GDP in FY1990 to 10.9 percent of GDP in FY2012.
    - Tax-financed spending increased to 2.5 percent of GDP in 2011 due in part to higher government subsidy to basic pension.
  - Indexation mechanics (Box 1):
    1. Two-stage adjustment of annual pension benefits:
       - New pensioners adjusted according to wage increase; others indexed to headline CPI inflation.
       - When price inflation > wage inflation, benefit adjusted using wage inflation.
       - When wage declines while price increases, no adjustment is done for all pensioners.
       - Example: CPI inflation 2.7 percent and nominal disposable wage increase 2.3 percent in 2014 → 2015 pension benefit increased by 2.3 percent for all.
       - Example: CPI inflation 0.8 percent and wage increase negative 0.2 percent in 2015 → 2016 pension benefit remains the same for all.
    2. Macroeconomic indexing reflects changes in number of the insured and average longevity:
       - First activated in 2015 to resolve payouts arising from FY2010–12 price declines.
       - A three-stage reduction completed in April 2015; additional indexing adjustment was 0.9 percent (0.6 percent reduction from fewer insured; 0.3 percent from rise in average lifespan).
       - Under a moderately realistic 2014 re-examination scenario, indexing estimated to stop in 2043, after which the replacement ratio remains 50.6 percent.
       - Under a more realistic scenario, indexing assumed to stop in 2040 as replacement ratio hits 50 percent; if allowed below floor, indexing would stop in 2050 at 45.7 percent.
  - Key recommendations:
    - Make macroeconomic indexing work under low price/wage inflation or deflation (government submitted a bill in March 2016 to allow partial delayed reflection).
    - Further raise pension age, for example by introducing an automatic link to life expectancy at 65. Pension eligibility age being gradually raised from 60 to 65, by 2025 (2030) for men (women) covered by the earnings-linked pension.
  - Observations:
    - Indexing limitations: when pre-indexing benefit increases are smaller than indexing adjustment, adjustment is zero (no negative adjustment); when benefit increases are zero or negative, indexing may not operate, adversely affecting finances.
    - Important to use prudent assumptions in quinquennial re-examinations to improve credibility and reduce reliance on tax revenue.

- Korea: NPS challenges and illustrative AAM scenarios
  - NPS launched in 1988; active members contribute 9 percent of income reported for taxes.
  - System: defined-benefit pension partially funded; accumulated assets around 30 percent of GDP; currently running a surplus.
  - Official projections:
    - NPS system will turn into an overall fiscal deficit in 2044 and pension assets fully depleted by 2060.
  - Illustrative retirement-age AAM simulations:
    - Current phasing raises retirement age to 65 by 2034.
    - Example I (AAM focusing on fiscal sustainability):
      - Trails current phasing until 2034, then locks expected pension duration at 21.8 years.
      - Postpones expected depletion of the NPS by 2–3 years.
    - Example II (AAM focusing on intergenerational fairness):
      - Locks pension duration expected for most recent retirees (1956 vintage cohorts) to later generations via a more gradual schedule aligned with projected life expectancy.
      - Would worsen fiscal position relative to current scheme until around 2060.
  - Observations:
    - Clear scope for further retirement-age reform to balance fiscal sustainability and intergenerational fairness.
    - Benefits and contribution-rate AAMs involve more difficult trade-offs; contribution-rate automatism is rarely used.

- China: fragmentation, projections, and policy options
  - System structure:
    - Separate schemes for enterprise workers (UWS) and Urban and Rural Residents (URR); administered at provincial level.
    - UWS: pay-as-you-go funded by employee and employer contributions; central government covers deficits. After 35 years of contribution, worker earning average provincial wage would receive slightly under 60 percent of average provincial wage.
    - URR: funded individual account + flat pay-as-you-go basic pension financed by state; flat portion set by law at 2½ percent of the urban wage.
    - Coverage and expenditure: UWS ~30 percent workforce contributes; pensioners ~40 percent of population age 60 and older; total expenditure near 3 percent of GDP. URR ~40 percent workforce contributes; 70 percent of population age 60 and older receive basic pensions; total expenditure about 0.3 percent of GDP.
  - Aging projections and fiscal impact:
    - Old age dependency ratio projected to increase from 13 percent in 2015 to near 40 percent in 2050.
    - Assuming unchanged parameters, pension expenditure projected to increase by 5 percentage points of GDP in 2015–50.
    - Combined deficits of UWS and URR projected to widen from 0.6 percent of GDP in 2015 to 5.8 percent of GDP in 2050.
    - Present discounted value (PDV) imbalance 2015–50 estimated at 94 percent of 2015 GDP.
  - Retirement age and AAMs:
    - Statutory retirement ages relatively low (55 years for women and 60 years for men).
    - Expected years in retirement at current ages: men 18½ years, women 25¾ years; by 2050 projected: men 24 years, women over 30 years.
    - Linking retirement ages to life expectancy via AAM would gradually raise retirement ages to 65 for men and 60 for women and improve sustainability:
      - Effect: projected pension system imbalance in 2015–50 declines from 94 to 72 percent of 2015 GDP.
  - Further reforms recommended to fully address imbalances:
    - Sharper increases in retirement ages, particularly for women (e.g., increase retirement ages for men and women to age 67 by 2050).
    - Modify indexing of pensions to price indexing.
    - Change benefit formulas to reduce average benefits by 5 percent.

### VI. Quantitative impacts (illustrative PDV impacts, 2015–50, percent of 2015 GDP)
- Pension imbalance (PDV of expenditure minus revenue): 94
- Impact of measures to restore balance (absolute PDV reductions):
  - Automatic increase in retirement ages: 22
  - Raise expected years in retirement to advanced economy average: 30
  - Change in indexation of benefits to price indexing: 31
  - Reduce benefits by 5 percent: 8

### VII. Selected international examples of AAMs (highlights from Appendix)
- Benefit-level adjustments:
  - Finland: new earnings-related pensions reduced in line with improvements in life expectancy at retirement; life expectancy coefficient calculated at age 62 and compared to 2009; expected benefit reductions of 20 percent by 2050 (other assumptions unchanged).
  - Sweden: annuity divisor translating notional balances into monthly benefits based on cohort-specific life expectancies; automatic balancing mechanism reduces notional returns and benefit indexation until funding gap closed.
  - United States: Social Security Trust Fund depletion triggers immediate downward adjustment (prorating benefits to contribution revenues).
- Contribution-rate adjustments:
  - Canada: every three years review; if legislated contribution rate below sustainability requirement and ministers fail to agree on alternatives, contribution rate increases go into effect; benefit indexation suspended for three years if increases are introduced.
  - Germany: point-value indexation adjusted by a contribution factor and a longevity-based sustainability factor; adjustment shares burden between pensioners and contributors.
- Retirement-age and service-time adjustments:
  - Czech Republic: after retirement age reaches 65 in 2030, automatic increase of 2 months per year in line with projected life expectancy improvements.
  - Norway: flexible retirement ages with actuarial neutrality and normal retirement age adjusted to life expectancy.
  - Spain: Automatic Balancing Mechanism introduced elements linking new benefits to life expectancy from 2019 and linking indexation of existing pensions to prices and the financial position from 2027 with lower and upper sustainability adjustment limits.

### VIII. Conclusions and policy recommendations
- AAMs can increase the likelihood that parametric changes occur early, improve predictability, and enhance intergenerational equity.
- Linking statutory retirement ages to life expectancy or other sustainability indicators is a particularly attractive AAM for Asia given higher expected years in retirement.
- Early implementation of AAMs can prevent the need for sharp future adjustments, especially in less mature systems.
- AAMs are not a silver bullet: for systems with large pre-existing unfunded liabilities, AAMs can slow accumulation of additional liabilities but cannot substitute for major parametric and structural reforms; AAMs help maintain sustainability after such reforms.
- Design considerations to maximize effectiveness:
  - Choose appropriate trigger variables (ex-ante vs. ex-post) and include feedback mechanisms to manage forecast errors.
  - Specify adjustment frequency and boundaries to avoid excessive or inadequate adjustments.
  - Consider distributional consequences (groups with lower life expectancy, informal workers, fragmented schemes) and complement AAMs with measures to protect adequacy for vulnerable groups.
  - Improve transparency and credibility through prudent assumptions in re-examinations and clear legal frameworks for precedence among multiple AAMs.

*Italic: Source: _wp16242 — 1. Automatic Adjustments of Pension Benefits (excerpts from IMF working paper).*

### 1. Automatic Adjustments of Pension Benefits .......................................................................17

### 1. Automatic Adjustments of Pension Benefits

### Content outline
- Includes a set of Figures and Tables and an Appendix related to automatic adjustment mechanisms (AAMs) for pension benefits.
- Figures referenced include: Old Age Dependency Rate and Pension Spending in Asia; Expected Years in Retirement, 2015; Old Age Dependency Rate in Asian Economies 2015–50; Pension Benefit and Social Security Contribution Rates in Asia; Revenue Composition 2014 Re-examination Scenario E; Pension Finances 2014 Re-examination; Scenario E; Pension Contribution Compliance; NPS: Pensioner-to-Active Member Ratio; Illustrative Scenarios with AAMs: Retirement Age and Pension Duration.
- Tables referenced include: Characteristics of AAMs-Trigger and Adjustment Variables; Characteristics of AAMs-Types of Pension Systems; Characteristics of Public Pension Systems in Asia and Pacific Economies; Impact of Options to Restore Pension System Sustainability 2015–50.
- Appendix: Selected Country Cases with AAMs.

### Key themes implied by the unit title and listed components
- Automatic adjustment mechanisms (AAMs) for pension benefits and their design features.
- Triggers and adjustment variables used in AAMs.
- Types of pension systems and how AAMs operate within them.
- Demographic pressures (Old Age Dependency Rate; Expected Years in Retirement) and projected implications through 2015–50.
- Fiscal implications and scenarios for pension finances, including a "2014 Re-examination; Scenario E".
- Empirical indicators: pension benefit and social security contribution rates, revenue composition, pension contribution compliance, pensioner-to-active member ratios.
- Illustrative scenarios connecting retirement age, pension duration, and AAMs.

### Intended analytical outputs (as suggested by listed items)
- Comparative diagnostics across Asian and Pacific economies of demographic and fiscal pressures on pensions.
- Characterization of AAM types, triggers, and adjustment variables for policy design.
- Scenario analysis for pension sustainability over 2015–50, including specific re-examination scenarios.
- Country case studies illustrating implementation of AAMs.

*Source: _wp16242 - 1. Automatic Adjustments of Pension Benefits .......................................................................17*

### References .............................................................................................................

### _wp16242 - References

### I. INTRODUCTION
- Asian economies are aging fast, with significant implications for their pension system finances.
- Some countries already have high dependency ratios (Japan); others expected to experience a sharp increase in the next couple of decades (China, Korea, Singapore).
- The projected pace of aging in many countries in the region surpasses that of other countries that went through a similar transition.
- Two main implications:
  - Pension system deficits can increase very quickly, limiting room for policy action and hampering fiscal sustainability.
  - Projected aging of populations can have large intergenerational equity implications.
- Urgent need to reform pension systems to achieve sustainability and ensure equity.
- Purpose of paper: explore how incorporating Automatic Adjustment Mechanisms (AAMs)—rules ensuring that certain characteristics of a pension system respond to demographic, macroeconomic and financial developments, in a predetermined fashion and without the need for additional intervention—can be part of pension reforms in Asia.
- Historical note: price indexation of benefits introduced in Denmark in 1922 (Gerig, 1960).
- Traditional AAMs aimed at protecting the value of current pensions or accrued entitlements; more recent focus includes fiscal sustainability and adequacy at the bottom of the distribution.
- Social insurance pension systems in many countries are dominated by defined benefit, pay-as-you-go schemes; many systems under previous rules are unprepared for population aging and need parametric or structural reforms.
- Political-economy rationale: automatic adjustments can reduce procrastination, increase likelihood of parametric changes, improve intergenerational equity, transparency and predictability, and serve as a signal of commitment to fiscal sustainability.
- Paper structure: discussion of rationale and structural features of AAMs and considerations for design (Section II); scope for introduction in Asia including case studies for Japan, Korea and China.

### II. AUTOMATIC ADJUSTMENT MECHANISMS IN PENSION SYSTEMS — Rationale and Benefits
- Rationale:
  - Main identifying feature is legal and procedural: making adaptive rules the default so parliaments must act to prevent automatic changes rather than to introduce them.
  - AAMs potentially introduce profound procedural and political change into policymaking and increase likelihood of timely parametric reforms.
- Benefits:
  - Can prevent buildup of intergenerational inequity due to aging.
  - Act as a signaling instrument for policymakers’ dedication to long-term sustainability, improving credibility, confidence and compliance.
  - Commitment to gradual, less painful adjustments versus occasional large reforms.
- Caveats:
  - Many AAMs were introduced after major parametric reforms; true test is whether AAMs operate when indicators call for adjustments.
  - Example: Japan’s AAM reduced policy uncertainty but was allowed to operate only recently due to persistent wage and consumer price deflation.

### II.B. A STRUCTURAL FRAMEWORK — Four main structural features of AAMs
- Key features elevated into legislation:
  1. Trigger variables (indicators determining when/how adjustments are made).
  2. Adjusting parameters linked to trigger indicators.
  3. Frequency of adjustment.
  4. Boundaries of adjustments.

- Trigger Variables:
  - Can be prediction-based (ex-ante) or state-of-the-world (ex-post).
  - Examples of ex-ante triggers:
    - age-specific/cohort-specific life expectancy
    - system dependency (ratio of pensioners to workers)
    - contributory years to length of benefit receipt
    - current deficits
    - future expenditures and deficits
    - benefit indexation
    - annuity divisor
  - Country examples linked to triggers:
    - Finland, Portugal (age-specific life expectancy)
    - Japan (system dependency)
    - Germany, Canada (contributory years)
    - Sweden, Latvia, Poland, Italy (annuity divisor adjustments)
    - US (other reductions in benefits)
    - Canada, Germany, Greece, Norway, Finland, Portugal, Czech Republic (retirement age)
    - Denmark (benefit indexation procedural automatism)
    - France (minimum qualifying service time for full pension)
  - Ex-post triggers: observed improvements in life expectancy; backward-looking benefit indexation.
  - Trade-offs:
    - Prediction-based AAMs may require feedback mechanisms to correct forecast errors.
    - State-of-the-world AAMs imply delayed reaction, possibly necessitating larger/faster adjustments.

- Adjustment Parameters:
  - Three broad categories: benefit levels, eligibility criteria, contribution rates.
  - Benefit levels:
    - Adjusted via benefit indexation, valorization of past earnings, reductions reflecting life expectancy improvements or financing shortfalls.
    - Examples: Germany, Canada, Japan, Portugal link degree of indexation to system dependency.
    - Finland, Portugal and Germany link benefits to life expectancy at retirement through a “life expectancy coefficient” or a “sustainability factor”.
    - US: benefits automatically cut when pension reserves are exhausted.
    - Sweden, Italy, Poland, Latvia (NDC schemes): annuity factor regularly amended to align notional retirement balances with present value of expected payouts.
  - Eligibility criteria:
    - Retirement age often adjusted automatically to reflect life expectancy improvements (Canada, Greece, Finland, Portugal).
    - Denmark: procedural review every five years with possibility of retirement age increases.
    - Czech Republic: automatic retirement increase of 2 months per year based on expected mortality improvements.
    - France: benchmark to keep ratio of expected contributory and beneficiary years constant (2/3-to-1/3 objective).
  - Contribution rates:
    - Rarely used for automatic adjustment.
    - Germany: automatically adjust contribution rates if they fall behind level required for medium-term cash balances.
    - Canada: contribution rates act as operating variable and a trigger; if legislated rate below sustainability level and government fails to act, contribution rates increase and benefits are nominally frozen for three years (until next actuarial assessment).

- Frequency of Adjustments:
  - Determined by law; may be annual or less frequent.
  - Autonomous or contingent on exogenous variables (e.g., annual indexation vs. retirement age increases when life expectancy reaches pre-determined thresholds).
  - Infrequent reviews can lead to larger adjustments.

- Boundaries of Adjustments:
  - Must be specified: frequency, individual adjustment limits, compounded impact.
  - Examples of boundaries:
    - Delay indexation until compounded price inflation reaches a threshold.
    - Replace automatism temporarily by discrete decisions if indicators exceed certain levels (e.g., hyperinflation).
  - Multiple AAMs possible in a system (e.g., retirement age tied to life expectancy and accrual rates tied to unfunded liabilities); legal precedence among mechanisms must be established.

- Applicability by pension system type:
  - Defined benefit and point systems: AAMs can adjust benefits, eligibility criteria, or contributions to improve sustainability or intergenerational equity.
  - Defined contribution and NDC systems: financial sustainability is generally inherent; AAMs can still adjust eligibility or contributions to improve adequacy.
  - Distinction between “strong” AAMs (automatic adjustments) and “weaker” AAMs (indicators trigger a reform process). Examples:
    - Japan: additional reforms called for if replacement rate falls under 50 percent.
    - Canada: failure of government action can trigger temporary suspension of benefit indexation by scheme administrator.

### II.C. CONSIDERATIONS
- AAMs face the same trade-offs as ad-hoc reforms: fiscal sustainability versus adequacy and equity.
- Main benefits of AAMs: predictability and simplicity.
- Equity and adequacy must be considered in design:
  - Automatic price indexation may erode pension levels relative to wages over time, potentially necessitating upward level shifts later.
  - Deflation can raise relative pension levels if indexation not designed for deflationary contexts.
  - Example: Japan’s contingency to call for reforms if replacement rate projected to fall under 50 percent in 5 years or if benefit dispersion becomes large.
  - Raising retirement ages mechanically may disadvantage groups with lower life expectancy or limited ability to work later in careers.
- Interaction with labor market and institutions:
  - AAMs can affect labor supply, savings, and human capital investment.
  - Need timely information provision and smooth, gradual adjustments to enable appropriate private responses.
  - Japan: prevalence of life-time employment could interact with retirement-age AAMs; limited demand for older workers may pose challenges.

### III. AUTOMATIC ADJUSTMENT MECHANISMS IN ASIAN PENSION SYSTEMS
- Significant scope to introduce automatic indexation of retirement age to changes in life expectancy across systems in the region.
- Many Asian pension systems do not yet have automatic links between retirement age and life expectancy; expected years in retirement typically higher in Asia.
- AAMs can help reduce financial imbalances (Korea and China case studies suggest potential benefits).
- Pension system composition in Asia:
  - Defined contribution systems: Hong Kong, Singapore, Australia, Malaysia, Indonesia.
  - Defined benefit systems: Japan, Korea, Thailand, Vietnam.
  - Mixed elements: China, India.
- For defined contribution schemes, AAMs can promote adequacy (for example, index retirement age to life expectancy in Singapore, Malaysia, Indonesia which have relatively high expected years in retirement).
- Aging trajectories in Asia:
  - Countries aging imminently or already underway: Korea, Hong Kong SAR, Singapore, China, Thailand, Brunei; Japan already aged.
  - Moderate or slow aging: Vietnam, Malaysia (moderate); India, Indonesia, Mongolia, Cambodia (slow).
- Policy implications by country group:
  - Countries aging fast: AAMs can slow accumulation of imbalances but need to be complemented with parametric and structural reforms (Korea and China cases).
  - Relatively young countries: easier to link system parameters to sustainability indicators early; priority is to ensure coverage, fiscal sustainability, benefit adequacy and equity; consider incorporating automatism at system design stage (e.g., NDC schemes).
- Some countries may have limited scope to reduce benefits (adequacy constraints), requiring reliance on eligibility adjustments (retirement age). Raising contribution rates can affect labor market and growth—other adjustments should be prioritized, but contribution rate increases may be needed where rates are relatively low and imbalances large.
- Fragmentation:
  - Many Asian countries have fragmented pension systems (separate schemes for civil service, military, state-owned enterprises) complicating AAM application.
  - Examples:
    - Japan consolidated four separate employee schemes in 2015, though dependents and self-employed remain separate.
    - China: distinct systems for enterprise workers and rural/urban residents, administered at provincial level.
    - Malaysia: main system defined-contribution; civil servants’ system defined benefit with large unfunded liabilities.
  - Some consolidation may be needed prior to adopting AAMs.

### IV. CASE STUDIES — Overview and Japan detailed points
- Case studies presented: Japan, Korea, China.
  - Japan: only country in region with an existing AAM.
  - Korea: relatively young pension system but one of the fastest aging countries.
  - China: aging society at a much lower income level; needs comprehensive reform strategy combining AAMs and structural reforms.

- Japan (high-level findings):
  - Universal public pension system introduced in 1961; defined-benefit, largely pay-as-you-go despite GPIF asset holdings.
  - Three categories of insured and three tiers of contributions and benefits:
    - Employees: contributions broadly proportional to earnings, receive basic pension (first tier) and earnings-related benefit (second tier).
    - Dependents of employees: exempt from contributions, entitled to basic pension.
    - Others (self-employed, unemployed): pay fixed amount and receive basic pension.
    - Voluntary schemes provide third tier via employers, insurers, or public entities.
  - 2004 comprehensive reform:
    - Introduced an AAM (“macroeconomic indexing”) and gradual increase in pension contributions.
    - Pension benefits automatically adjusted to demographic changes (size of labor force and longevity).
    - Contribution rates being gradually increased until fiscal year 2017, after which they are fixed.
    - After contribution increases complete, adjustment of the replacement ratio will be key: if replacement ratio expected to go below 50 percent by the next quinquennial re-examination, additional reform measures will be called for.
  - Historical spending and financing:
    - Public pension spending increased from 5.2 percent of GDP in FY1990 to 10.9 percent of GDP in FY2012.
    - Tax-financed spending increased to 2.5 percent of GDP in 2011 due in part to higher government subsidy to basic pension.
    - FY2000–FY2002: pension benefit was kept constant by special laws despite rule of full price indexation.
  - Expected impact of 2004 reform:
    - If implemented as planned, expected to reduce policy uncertainty and contain growth of pension spending relative to GDP (IMF Fiscal Monitor projections show Japan’s increases in pension spending to be much smaller than peer countries).
    - 2004 reform expected to improve declining compliance ratio through positive effects on sustainability and intergenerational equity, though progress has been mixed.
  - Options to improve credibility and reduce reliance on tax revenue:
    - Use more prudent assumptions on key parameters and reasonable scenarios in quinquennial re-examinations.
    - Example of assumption mismatch: inflation was expected to rise to 1.0 percent in both the 2004 reform and the 2009 re-examination, which did not happen.
    - Note: Defined "replacement ratio" as the ratio of household pension benefits of a new pensioner to the wage of a representative household.
  - Pension contribution compliance trend: figure shows National Pension compliance percent series (1986–2014) with values ranging between 50 and 90 percent (exact annual values in figure).

- Additional high-level lessons from analysis (summarized):
  - Important room exists to incorporate AAMs in the region.
  - Linking statutory retirement ages to longevity or other sustainability indicators is an attractive option.
  - Early implementation of AAMs can prevent need for sharp future adjustments, especially in less mature systems.
  - AAMs increase predictability and intergenerational equity and can signal policymakers’ commitment to sustainability.
  - AAMs are not a silver bullet: they slow accumulation of unfunded liabilities but do not substitute for major parametric and structural reforms which establish a new baseline that AAMs can help maintain.

*Italic: Source: _wp16242 - References (excerpts from IMF PDF content provided).*

### 1.0 percent, which is currently beyond reach

### _wp16242 - 1.0 percent, which is currently beyond reach

### Overview
- Asian economies are aging fast, creating rising financial imbalances in pension systems.
- Incorporating Automatic Adjustment Mechanisms (AAMs) can increase the likelihood of parametric changes with limited political trade-offs, improve predictability, and enhance inter-generational equity.
- AAMs are not a silver bullet: they can slow unfunded liabilities accumulation but may not fully fix systems with large unfunded liabilities; more fundamental reforms may be required in those cases.

### Japan: pension challenges and reform options
- IMF (2016) estimates that Japan’s potential growth will gradually decline from the current 0.5 percent to close to 0 in 2030.
- Past downward revisions of projections on the total fertility rate and thus population may have slowed pension reform efforts.
- The 2014 re-examination of pension finances shows that under prudent scenarios the replacement ratio will hit the 50 percent floor around 2035–40.
- Recommended improvements:
  - Make the AAM (“macroeconomic indexing”) work better under a low price/wage inflation or deflation environment. Current AAM is not designed to operate under deflation; government submitted a bill in March 2016 to allow part of the indexing not fully reflected in the pension adjustment to be reflected in later years. Allowing macroeconomic indexing fully under all circumstances would contribute to more sound pension finances.
  - Further raise the pension age, for example by introducing an automatic link to life expectancy at 65. The pension eligibility age is being gradually raised from 60 to 65, by 2025 (2030) for men (women) covered by the earnings-linked pension.
- Expected benefits of raising pension age:
  - Reduce reliance on tax money and/or improve the replacement ratio.
  - Contribute to growth through increasing labor supply.

### Box 1 — Japan: Automatic Adjustments of Pension Benefits (mechanics and effects)
- Two-stage adjustment of annual pension benefits:
  1. Indexing to wages or headline CPI inflation based on rules:
     - Pension amount for new pensioners adjusted according to wage increase; others indexed to headline CPI inflation.
     - When price inflation > wage inflation, benefit adjusted using wage inflation.
     - When wage declines while price increases, no adjustment is done for all pensioners.
     - Example: CPI inflation 2.7 percent and nominal disposable wage increase 2.3 percent in 2014 → 2015 pension benefit increased by 2.3 percent for all.
     - Example: CPI inflation 0.8 percent and wage increase negative 0.2 percent in 2015 → 2016 pension benefit remains the same for all.
  2. Macroeconomic indexing (“indexing”) reflects changes in number of the insured and average longevity.
     - Introduced in 2004 reform; first activated in 2015 to resolve “excessive” payouts from FY2010–12 price declines.
     - A three-stage reduction in benefits in line with past price declines completed in April 2015; additional adjustment due to indexing was 0.9 percent (0.6 percent reduction from fewer insured; 0.3 percent from rise in average lifespan).
     - The indexing allows real-term reductions and will be in use until pension sustainability is secured.
     - Under a moderately realistic 2014 re-examination scenario, indexing estimated to stop in 2043, after which the replacement ratio remains 50.6 percent. Under a more realistic scenario, indexing assumed to stop in 2040 as replacement ratio hits 50 percent; if allowed below floor, indexing would stop in 2050 at 45.7 percent.
- Limitations of indexing:
  - When the increase in pension benefits (before indexing) is smaller than the indexing adjustment, the adjustment is zero instead of negative (e.g., price increase 0.5 percent and indexing 0.9 percent → pension stays same rather than decline 0.4 percent).
  - When increase in pension benefits is zero or negative (as in 2016), the indexing does not work at all, negatively affecting pension finances.
- Policy implication: Allowing the macroeconomic indexing function fully under all circumstances would strengthen pension finances.

### Korea: NPS challenges, AAM potential, and illustrative AAM scenarios
- Korea faces rapid population aging from low fertility and high life expectancy; NPS is the world’s third largest pension fund.
- NPS features:
  - Launched in 1988; active members contribute 9 percent of their income reported for taxes.
  - Retiree benefits linked to average income of active participants, own lifetime income averaged in present value terms, and active participation periods via a pre-fixed formula → defined-benefit pension.
  - Partially funded; currently running a surplus with accumulated assets around 30 percent of GDP.
- Demographic pressures:
  - Higher dependency ratio will increase pensioner-to-active member ratio.
  - Pension payments per retiree set to increase as average active participation periods lengthen as system matures.
- Official projections:
  - NPS system will turn into an overall fiscal deficit in 2044 and pension assets fully depleted by 2060.
  - Long-term outlook suggests potential pension-led fiscal cliff if remedial actions are delayed.
- Past reforms: 1998 reforms increased retirement age and reduced benefits; 2007 reforms further reduced accrual rates. Despite reforms, sizable unfunded liabilities remain.
- AAM suitability:
  - Demographic and macroeconomic projection uncertainty implies AAMs could restrain unfunded liabilities more effectively and predictably than one-off reforms.
  - NPS immaturity is advantageous for AAM adoption: favorable dependency ratios and current cash surpluses give automatic adjustments more scope and may be politically easier to introduce.
  - Designing appropriate AAMs requires careful analysis of imbalances, adequacy, and equity considerations.
- Retirement age AAM (link to life expectancy) is an increasingly popular option:
  - Retirement age is set to increase to 65 by 2034 under current phasing.
  - Simulations (Figures and Examples):
    - Two illustrative examples using retirement-age AAMs show trade-offs between fiscal sustainability and intergenerational fairness.
    - Example I (AAM focusing on fiscal sustainability):
      - Trails current phasing until 2034 (1969 vintage cohorts retire at 65), then locks expected pension duration at 21.8 years (its trough).
      - Simulation indicates such an AAM will postpone expected depletion of the NPS by 2–3 years.
    - Example II (AAM focusing on intergenerational fairness):
      - Locks pension duration expected for most recent retirees (1956 vintage cohorts) to later generations via a more gradual schedule aligned with projected life expectancy.
      - This scheme would worsen the NPS fiscal position compared to the current scheme until around 2060.
  - Observations:
    - Clear scope for further reform in retirement age for fiscal sustainability and inter-generational equity.
    - Trade-off exists: AAM choice must balance fiscal sustainability and fairness; protect system stability and participant expectations.
- Other AAM types (reducing benefits or increasing contributions) involve more difficult tradeoffs:
  - Benefits have already been cut; further reductions risk pension adequacy issues.
  - Contribution rates are rarely used for automatic adjustment given broader ramifications and tax policy implications.

### China: fragmentation, projections, and policy options
- System structure:
  - Separate schemes: UWS (salaried workers in enterprise sector) and URR (Urban and Rural Residents/non-salaried).
  - System fragmented across schemes and geographically; schemes administered at provincial level.
- UWS features:
  - Pay-as-you-go funded by employee and employer contributions; central government covers deficits.
  - Benefits = defined-benefit formula + defined contribution portion functioning as notional DC.
  - After 35 years of contribution, a worker earning average provincial wage would receive slightly under 60 percent of average provincial wage.
  - About 30 percent of workforce contributes; pensioners ~40 percent of population age 60 and older.
  - Total expenditure near 3 percent of GDP.
- URR features:
  - Funded individual account component + flat pay-as-you-go basic pension financed by state.
  - Flat portion set by law at 2½ percent of the urban wage.
  - About 40 percent of workforce contributes; 70 percent of population age 60 and older receive basic pensions.
  - Total expenditure about 0.3 percent of GDP.
- Aging projections and fiscal impact:
  - Old age dependency ratio projected to increase from 13 percent in 2015 to near 40 percent in 2050.
  - Assuming unchanged parameters, pension expenditure projected to increase by 5 percentage points of GDP in 2015–50.
  - Combined deficits of UWS and URR projected to widen from 0.6 percent of GDP in 2015 to 5.8 percent of GDP in 2050.
  - In present discounted value, imbalance between expenditure and revenue over 2015–50 estimated at 94 percent of 2015 GDP.
- Retirement age and AAMs:
  - China’s statutory retirement ages relatively low (55 years for women and 60 years for men).
  - At current ages, expected years in retirement: men 18½ years, women 25¾ years; by 2050 projected to be men 24 years, women over 30 years.
  - Linking retirement ages to life expectancy via AAM would gradually raise retirement ages to 65 for men and 60 for women and improve sustainability.
  - Impact: by linking retirement ages to life expectancy, projected pension system imbalance in 2015–50 declines from 94 to 72 percent of 2015 GDP — improvement but a large imbalance remains.
- More fundamental reforms suggested to fully address imbalances:
  - Sharper increases in retirement ages, particularly for women (e.g., increase retirement ages for men and women to age 67 by 2050).
  - Modify indexing of pensions to price indexing.
  - Change benefit formulas to reduce average benefits by 5 percent.

### Quantitative impacts and illustrative table (2015–50, in percent of 2015 GDP)
- Pension imbalance (PDV of expenditure-revenue): 94
- Impact of measures to restore balance:
  - Automatic increase in retirement ages: 22
  - Raise expected years in retirement to advanced economy average: 30
  - Change in indexation of benefits to price indexing: 31
  - Reduce benefits by 5 percent: 8

### Conclusions and policy recommendations
- AAMs can be a useful part of pension reforms in Asia to increase the likelihood of parametric changes occurring early and to improve predictability and inter-generational equity.
- A particularly important AAM: linking retirement age to life expectancy or other system sustainability indicators — relevant given higher expected years in retirement in Asia.
- AAMs can help prevent the need for sharp future adjustments and serve as a signaling instrument of policymakers’ commitment to long-term pension sustainability.
- For systems with already large unfunded liabilities, AAMs alone are insufficient; more fundamental reforms will be required, with AAMs helping to keep systems sustainable thereafter.
- Japan’s experience: AAM helped prevent further increases in financial imbalances but needs improvements in credibility and effectiveness through more prudent assumptions and better macroeconomic indexing under low inflation or deflation environments.

*Source: IMF working paper content provided in the supplied document.*

### REFERENCES

### _wp16242 - REFERENCES

### Bibliographic references
- Asian Development Bank, 2012, Pension Systems in East and Southeast Asia: Promoting Fairness and Sustainability edited by Donghyun Park, Asian Development Bank: Philippines.
- Gerig, Daniel S., 1960, “Automatic Cost-of-Living Adjustments in Foreign Countries,” Social Security Bulletin, March 1960. Available at: https://www.ssa.gov/policy/docs/ssb/v23n3/v23n3p13.pdf
- Gillingham, Robert. 2010, “Meeting the Long-term Challenge of Old-Age Entitlement Spending: What Can We Learn from the Greenspan Commission”, Working Paper 10-25, Mercatus Center, June 2010.
- International Monetary Fund (IMF), 2016, “Japan: Staff Report for the 2016 Article IV Consultation”, IMF Country Report 16/223.
- Kashiwase, K., M. Nozaki, and K. Tokuoka, 2012, “Pension Reforms in Japan”, IMF Working Paper 12/285.
- OECD, 2013, Pensions at a Glance: Asia/Pacific 2013, OECD: Paris.
- OECD, 2012, “Putting Pensions on Auto-pilot: Automatic adjustment Mechanisms and Financial Sustainability of Retirement-income Systems” in OECD Pensions Outlook 2012, World Bank: Washington DC.
- Soto, Mauricio, and Sanjeev Gupta, Forthcoming, “Pension Reform in China: Toward a Sustainable, Equitable, and Portable System”.
- Tokuoka, K., 2012, “Intergenerational Implications of Fiscal Consolidation in Japan”, IMF Working Paper 12/197.
- World Bank, 2016, Live Long and Prosper: Aging in East Asia and Pacific, World Bank: Washington DC.

### Appendix: Selected country cases with AAMs — Adjustment of Benefit Levels
- Bosnia-Herzegovina
  - 2001: Both confederated entities introduced a strict benefit rationing regime protecting a bracket roughly corresponding to the subsistence minimum; remaining benefits were pro-rated to remaining revenues.
  - Adjustments executed monthly; unpaid pensions were not accounted as arrears.
  - Outcome: extreme AAM did not increase transparency or sustainability; governments induced enterprises to pay contribution arrears and compensated enterprises via tax easements and budget subsidies, resulting in continued nontransparent subsidization.
- Finland
  - From 2010, new earnings-related pensions are reduced in line with improvements in life expectancy at retirement.
  - Life expectancy coefficient calculated at age 62 and compared to that in 2009.
  - Expected result: benefit reductions of 20 percent by 2050, keeping all other assumptions and policies unchanged.
  - Legislated increase of the retirement age to 65; further increases determined by improvements in life expectancy.
- Japan
  - AAM introduced in 2004 and activated in 2015.
  - “Macroeconomic indexing” reflects changes in the number of the insured and average longevity.
  - Indexing allows benefits to be reduced in real terms but pension benefits cannot be cut in nominal terms.
  - If the replacement ratio is projected to drop below 50 percent in the quinquennial re-examination of pension finances, additional reform measures are called for.
- Latvia and Poland
  - Reformed social security pension schemes along the Swedish Notional Defined Contributions (NDC) example.
  - NDC schemes adjust benefit levels to life expectancy at retirement using statistical life tables and equating present value of expected benefit stream with the value of the notional account at retirement.
- Russia
  - Replaced its NDC system with a point system as of January 2015.
  - Points generated by dividing covered earnings with the economy-wide average wage.
  - Point values calculated by the higher of: (a) indexing the previous year’s point value to CPI or (b) dividing Pension Fund revenues (including budget subsidies) by total points.
  - Point value valorization applied to new and existing retirees; inclusion of budget subsidies means government can directly influence the indexation.
- Sweden
  - Key AAM: annuity divisor translating notional individual balances into monthly benefits based on cohort-specific life expectancies.
  - Mortality information regularly updated, reflected in divisor, communicated to members.
  - Automatic balancing mechanism: annual scheme balance sheet compiled; if unfunded liabilities emerge, both notional returns credited to individual accounts and benefit indexation are automatically reduced until funding gap closed.
- United States
  - AAM included in Social Security legislation triggered by depletion of the US Social Security Trust Fund: Social Security prohibited to borrow.
  - From trust fund depletion, pension benefits would be adjusted downward immediately — estimates cited of reductions by as much as 30 percent — effectively prorating benefits to contribution revenues.
  - Legislative design intended to induce earlier, more gradual adjustments by legislators rather than apply the extreme AAM.
- Germany and Spain
  - Both adjust benefit levels and other parameters (Germany: contribution rates; Spain: retirement ages) — discussed in later sections.

### Appendix: Selected country cases with AAMs — Adjustment of Contribution Rates
- Canada
  - One of only two countries (other being Sweden) with automatic adjustments explicitly triggered by long-term sustainability considerations.
  - Every three years there is a review of the financial sustainability of the public, earnings-related scheme.
  - Scheme is partially funded; reserves intended to smooth contribution rates and provide buffer when baby boom cohorts retire.
  - Variable responding to imbalances is the contribution rate: if the legislated rate is below that required for sustainability, the contribution rate is expected to adjust.
  - Projected imbalances first trigger a call on provincial ministers to agree on an alternative; contribution rate increase goes into effect if ministers fail to act.
  - If contribution rate increases are introduced, benefit indexation is suspended for three years (until next actuarial study), sharing burden and pressuring governments to find alternatives.
- Germany
  - Earnings-related public pension scheme operates a point system for over fifty years.
  - Points earned = individual insurable earnings divided by economy-wide average wage.
  - At retirement, points multiplied by point value established for the year.
  - Point value annually indexed to wage growth and adjusted downward by a contribution factor (reflecting necessary contribution rate increases) and a sustainability factor based on improvements in longevity.
  - Index uniformly applied to new pensions and benefits in service (acts as benefit indexation rule).
  - Contribution factor ensures that if contribution rates need to increase, pension benefits decline such that 25 percent of the adjustment is borne by pensioners and 75 percent by contributors.

### Appendix: Selected country cases with AAMs — Adjusting Retirement Ages and Qualifying Service Time
- Czech Republic
  - After current retirement age increase completed (at 65 for both men and women) in 2030, retirement age will continue increasing automatically at a rate of 2 months per year in line with projected improvements in life expectancy at retirement.
- France
  - Gradually increasing retirement ages (not linked to life expectancy via automatic rule).
  - Minimum qualifying service time for a full pension will automatically increase in line with life expectancy to keep ratio of period of benefit receipt to working and contributing constant.
  - “Cost of a point” and point value factors established in accordance with time-bound agreements (as opposed to annual re-negotiations); example: agreement expired end-2012 required upscaling cost of points in line with wages and point value in line with prices.
- Greece
  - Life expectancy adjusted retirement age increases commencing in 2021.
- Norway
  - From 2011 introduced two automatic measures conditioned on changes in life expectancy:
    - Flexible retirement ages observing actuarial neutrality (implying adjustments to life expectancy at retirement).
    - Normal retirement age adjusted to life expectancy.
- Portugal
  - Introduced AAM ensuring new pensions reflect changes in life expectancy at 65, measured between 2006 and the year of retirement.
- Spain
  - Parametric reforms in 2011: retirement age increases and extension of wage histories counted into pension formula.
  - 2013: introduced an Automatic Balancing Mechanism (ABM) with three elements:
    - (a) Starting from 2019, new benefits calculated with life expectancy changes taken into consideration.
    - (b) From 2027 on, indexation of pensions already in payment will no longer be linked solely to prices but to prices and the financial position of the pension system.
    - (c) A lower (-2.25 percent) and upper (+0.5 percent) limit on the sustainability adjustment of pension indexation.

*Source: _wp16242 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16242.pdf_