RESEARCH DOSSIER
Public Pension Minimum Retirement Ages and Early Exit
Abstract. Public pension eligibility ages govern the timing and value of public old-age income, but a single national “retirement age” conceals several distinct legal thresholds and behavioral outcomes. This monograph examines the minimum age at which a public pension may be claimed, the age for an unreduced entitlement, contribution- and occupation-specific exceptions, actuarial reductions for early claiming, and the relationship between benefit receipt and labor-market withdrawal. It analyzes international minimum social security standards, official cross-country indicators, administrative pension rules in the United States, United Kingdom, Germany, Sweden, Finland, and Australia, and causal evidence from pension reforms. The analysis distinguishes pension design from employment constraints and assesses reforms against coverage, lifetime benefit valuation, fiscal substitution, distributional adequacy, and legal predictability. OECD data for 2024 describe an average normal pension age of 64.7 years for men and 63.9 for women under a specified full-career model, while the ILO records substantial global gaps in older-age pension receipt. Evidence from Austrian early-eligibility reforms indicates that higher minimum ages can increase employment yet also increase unemployment-program reliance. The central finding is that retirement-age policy operates through interacting thresholds, benefit formulas, employment opportunities, and social protection pathways. Sound evaluation requires exact statutory cohorts, separate measures of claiming and work exit, and government-wide rather than pension-only accounting. Original contributions consist of comparative interpretation, a source-dated analytical taxonomy, reproducible illustrative valuation, and an evidence-sensitive evaluation agenda; no new proprietary theory or causal estimate is asserted.
Keywords: public pensions; minimum pension age; early retirement; pension eligibility; statutory retirement age; normal retirement age; early claiming; benefit reductions; actuarial neutrality; labor supply; old-age income adequacy; social security; pension reform; comparative pension law; distributional analysis; social insurance.
Introduction
The legal age of pension entitlement is an allocation rule for time and risk. It determines when accumulated rights or citizenship-based guarantees can be converted into cash income; it also shapes the years over which pension liabilities are paid, the incentives to remain employed, and the period during which people who cannot work require other protection. An age rule consequently affects the pension institution, the individual household, the employer, and neighboring public programs. The economically relevant question is not simply what the legal retirement age should be, but which entitlement opens at what age, to whom, under what conditions, and with what consequences for income and employment.
The OECD's current retirement age definitions demonstrate that even the internationally standardized normal retirement age is a modeled concept: it concerns eligibility without penalties across mandatory schemes after an assumed full career beginning at age 22. This differs from a national statute naming a standard pension age, an early-claiming threshold in one component, or an observed average age of leaving the labor force. The OECD's effective labor-market exit indicator measures yet another phenomenon and therefore cannot be substituted for an administrative series of first pension claims.
The international normative baseline is equally differentiated. Article 26 of the ILO's Social Security Convention No. 102 permits a prescribed old-age benefit age normally not above 65, with a higher age subject to consideration of older persons' working ability; its application is conditioned by national ratification and the branches accepted. The Social Protection Floors Recommendation No. 202 separately articulates basic income security for older persons, emphasizing nationally defined guarantees. These instruments have different normative characters: the convention creates treaty obligations for ratifying members to the applicable extent; the recommendation supplies international guidance. Neither implies a uniform global early retirement option.
The OECD review of recent pension reforms documents statutory and incentive reforms over September 2023–September 2025, supplying historical context rather than verification of every subsequent amendment. The empirical scope is worldwide in its conceptual and comparative foundations, while close legal readings use six deliberately contrasting jurisdictions and their specified public pension components. The time anchor is the latest OECD comparative edition examined here, published in 2025 and largely describing 2024 rules and behavior, combined with official rules inspected on November 17, 2026. Legal claims are explicitly attached to the cohort, component, and official source. A live administrative guidance page can reflect a recent revision; it should not be read retroactively as the statute that governed earlier cohorts.
This study uses comparative doctrinal analysis, institutional process tracing, formal cash-flow reasoning, interpretation of aggregate indicators, and critical appraisal of published causal designs. An illustrative annuity comparison is computed arithmetically without importing unverified mortality tables or falsely treating a hypothetical worker as a representative beneficiary. Scholarly literature is used for its actual examined design and findings. The original synthesis establishes the interactions among legal thresholds, labor-market feasibility, and retirement-income adequacy, without representing an invented theory as established science.
Conceptual Foundations and Historical Development
The several ages within a pension system
A public pension system can specify a minimum claim age, a normal or full-benefit age, a compulsory or default employment-retirement age in a different legal regime, and component-specific entitlement dates. A claimant may encounter contribution-period requirements, residence requirements, accrued-balance minima, disability certification, occupational exposure conditions, or minimum benefit tests. Age alone may be necessary and insufficient for eligibility. Where access is contingent on years of insurance, the relevant legal rule is a surface across age and contribution duration rather than a single number.
For analytical clarity, the earliest eligibility age is the minimum age at which a specified benefit component may legally begin for a claimant satisfying all other conditions. A normal pension age is the age at which an identified benefit may be drawn without the reduction imposed solely for early commencement; it is not necessarily the first age of eligibility. An effective claiming age is the observed age when pension payments actually commence. An effective labor-market exit age measures withdrawal from labor-force participation under the statistics producer's method. An individual may stop working and wait to claim, claim while remaining employed, move into unemployment or disability protection, or begin receiving a partial pension while continuing paid work. The U.S. Social Security Administration explicitly distinguishes benefit claiming from stopping work.
Retirement income can be contributed social insurance, tax-financed minimum protection, a social pension, occupational pension income, or private savings. Comparing a national old-age assistance threshold with the minimum access age in an earnings-related program without noting that difference manufactures an artificial convergence or divergence. The OECD's current-age comparison makes a particularly important methodological distinction: benefits in basic residence-based and safety-net components often cannot start before their standard statutory age, even where earnings-related components permit earlier claims.
The label “early retirement” may refer to at least four events: an earnings-related pension claimed with a permanent reduction; an unreduced pension granted to a long-career worker at an otherwise early age; an occupational or disability-related exception; or a person leaving employment before any public pension becomes payable. Those cases differ in law, cost, expected duration, and exposure to income poverty. A reliable research record names the event and the scheme rather than allowing one expression to carry all meanings.
Institutional history and the pension-age bargain
Age-based old-age protection developed as systems converted lifetime income uncertainty into contingent public entitlements. Increasing life expectancy and changing employment histories subsequently challenged the relationship between the contribution base, the pension accumulation period, and the expected duration of benefit payments. The question of an appropriate age became entangled with coverage of physically demanding occupations, uneven life expectancy, women's unequal contribution histories, and the distribution of work opportunities late in life. These are observable institutional tensions, not assumptions about identical aging or identical capacity to work across a population.
A pivotal illustration of legislative cohort staging is the United States. The SSA retirement-age explanation records the phased increase in the full retirement age enacted in the 1983 amendments, eventually reaching 67 for people born in 1960 or later, while retaining an earliest retired-worker claiming age of 62. This reform altered the benefit schedule relative to the minimum claiming threshold. The distinction matters: an unchanged first claiming age can coexist with a reduction in benefits available at that age because the distance to the full-benefit age has increased.
The United Kingdom illustrates a different design path in its basic public entitlement. The official State Pension age timetable sets out a legislated increase from age 66 to age 67 over 2026–2028 with birth-cohort-specific monthly increments. Access to a new State Pension depends on reaching the applicable State Pension age and satisfying relevant National Insurance qualifications. Unlike a contributory early-claim option with permanent penalties, the standard new State Pension has no ordinary early-claim route before State Pension age. These differences make retirement-age reform historically specific: a change can move a binding eligibility gate, shift the unreduced reference point, or both.
International standards provide a separate history of normative commitments. Article 26 of ILO Convention No. 102 expresses an age threshold in relation to working ability; the later Recommendation No. 202 connects older-age income security with an accessible minimum floor. The analytical implication is that eligibility age, adequate guaranteed income, and ability to remain employed belong in the same evaluation, although they are governed through different institutions.
A comparative unit of observation
The unit of an age-policy observation is defined here as jurisdiction × pension component × birth cohort × legal effective date × contribution history × claimant characteristic. For each observation, record: minimum claim age, unreduced benefit age, vesting or residence condition, early-claim adjustment, late-claim credit, treatment of employment income, partial-claim options, special eligibility categories, administrative take-up, and appeals or transition provisions. The model does not assert that this collection is a new theoretical system; it is a practical audit specification for comparing unlike legal entitlements without conflating them.
A table with only country and a single “retirement age” cannot establish these distinctions. Nor can a legal entitlement table describe actual retirement choices. The OECD's separate treatment of future retirement ages and labor-market exit behavior supplies an empirical foundation for treating legal rules and population outcomes as different objects.
Statutory Age Rules and Comparative Legal Design
International norms and the limits of harmonization
The ILO's Convention No. 102, Article 26 frames the insured contingency as survival beyond a prescribed age and permits prescribed employment-related conditions for suspension or reduction. The specified threshold is normally no more than 65, with a higher threshold allowed where the competent authority gives due regard to elderly persons' working ability. The provision must be applied within the convention's ratification and coverage structure. It does not establish a general age-65 entitlement for every resident of every member country, and it does not make age 65 the universal legal minimum for private or occupational pensions.
The social protection floor guarantees address a separate problem: a formal pension rule may exist while older people lack sufficient income. The ILO recommends basic income security for older persons, based on nationally determined levels and implementation by law. The legal force of national rights, the financing base, the availability of an administrative remedy, and the breadth of coverage become critical in assessing real access. A pension age cannot substitute for proof of entitlement or adequacy.
Comparative OECD evidence and denominator integrity
For a full-career model beginning at age 22, the OECD reports the average 2024 normal retirement age as 64.7 for men and 63.9 for women in Pensions at a Glance 2025. These are averages of scheme-defined modeled outcomes across OECD countries, not averages of the ages at which persons actually claimed. In the OECD's prospective analysis, the average early eligibility age excluding special very-long-career cases was 62.5, compared with 64.7 for the reference normal age; it is projected under then-current rules to rise to 63.9 versus 66.4 for careers beginning in 2024. The same prospective-age analysis reports a mean effective annual early-claim penalty of 4.4% in specified contributory basic, defined-benefit, and points schemes. The figures describe different cohorts and modeled conditions; subtracting them without acknowledging those definitions would be misleading.
Multiple components can become payable at different dates. The OECD identifies that the operative normal retirement age in its measure can reflect the latest component required for an unreduced full bundle. Some long-career provisions also grant unreduced access earlier than the generic age. This creates a distributional condition: the person with uninterrupted full-time employment may qualify earlier than a same-age individual with fragmented employment even if the latter has greater economic need. Cross-country comparisons must state both the career-entry assumption and the crediting treatment of care, unemployment, and disability periods.
United States: early claim, full benefit, and earnings histories
The U.S. retired-worker old-age insurance benefit generally permits claiming from 62 once contribution conditions are satisfied. Full retirement age is 67 for those born in 1960 or later. The official SSA benefit-age schedule illustrates the permanent reduction that accompanies early commencement. The SSA program explainer clarifies the minimum insured-service requirement and the interaction of cohort-specific full retirement age with the 1983 reform. At age 62 with a full retirement age of 67, the reduction reaches 30% of the primary insurance amount under the normal reduction schedule.
The formula is time-specific: SSA's early-or-late retirement calculation applies 5/9 of one percent per month for the first 36 early months and 5/12 of one percent for months beyond 36. For a 60-month early claim, 36 × 5/9% plus 24 × 5/12% equals 20% + 10% = 30%. This is a benefit reduction relative to an individual primary insurance amount and is not a universal statement that lifetime discounted benefits are 30% lower. For the same cohort, continued work can affect the underlying earnings computation because the SSA uses a 35-year earnings history. The eventual monthly benefit therefore combines eligibility timing with both earnings accrual and early/late adjustments.
Early claiming does not require complete labor-market exit. The SSA working-and-benefits rules provide for the possibility of work after commencement, with an earnings test before full retirement age and later recomputation where benefits were withheld. A casual comparison of all early claimants with nonclaimants therefore cannot treat pension commencement as retirement from work or interpret a benefit reduction in isolation from the earnings test.
United Kingdom: a binding public entitlement age
The UK new State Pension eligibility rules require reaching State Pension age and normally at least 10 qualifying National Insurance years to receive any new State Pension. The statutory age increase from 66 to 67 is phased by date of birth between 2026 and 2028, with a later legislated increase to 68 in 2044–2046 as represented in the currently published timetable. The latter is a statement of enacted schedule, not a projection about whether future legislatures will amend it.
An English-language phrase such as “early retirement at 60” in a UK household survey may reflect an occupational or private pension, savings, an incapacity-related payment, or labor-force exit without a State Pension. The ordinary new State Pension cannot simply be claimed early in exchange for an actuarial reduction. A UK statutory-age increase therefore changes the earliest new State Pension access directly, while the U.S. full-retirement-age change can leave age-62 access open but reduce the amount at that age. This is a substantive institutional difference with consequences for income bridging and substitution into other programs.
Germany: contributory careers and permanent deductions
The German statutory system provides distinct old-age routes for insured persons with qualifying periods. According to the public pension agency's long-career pension guidance, a qualifying history of 35 years can permit early retirement from 63, with a permanent 0.3% reduction for each month before the applicable unreduced age. A 48-month advancement therefore generates 14.4% reduction under that schedule. Those with 45 years of qualifying periods face a distinct unreduced route; the agency's long-insured benefit description notes an eventual age of 65 for persons born in 1964 or later in the especially long-insured category.
The legal significance is that counting periods can matter as much as chronological age. A person with 35 years and one with 45 years of qualifying periods do not necessarily face the same earliest unreduced pension age. Terms such as compulsory contributions, periods of child-rearing, certain unemployment periods, and credited periods must be read in the agency's exact eligibility rules rather than inferred from uninterrupted employment alone. An analysis of statutory age by sex or occupation without contribution histories risks attributing to the pension-age parameter what is actually determined by eligibility credits.
Sweden: cohort-linked reference ages and benefit layers
The Swedish Pensions Agency describes its recommended retirement age as an age linked to life expectancy. This benchmark governs component-specific access and is not an instruction that everyone must stop working. For the period 2026–2031, the agency identifies 67 as the established recommended age. Its individual cohort table explains that the earliest income- and premium-pension claim can precede the applicable recommended age, while guarantee pension and certain supplements become available at the recommended age, subject to their conditions.
A single Swedish age number misses the coexistence of an earnings-linked benefit that may be drawn early and a public income floor with a later access threshold. The comparative importance is the possibility of drawing one public pension component while remaining exposed to an income shortfall until the guarantee component becomes available. Since the agency distinguishes legally established ages from forecasts for younger cohorts, any quoted future age must also carry a status label: enacted, predetermined administrative age, or projected under the indexing formula.
Finland: partial benefit access while work continues
Finland permits flexible partial commencement of an accrued earnings-related old-age pension. The Finnish Centre for Pensions states that partial old-age pension rules allow 25% or 50% of accrued pension to be drawn under age-specific eligibility, with a permanent 0.4% monthly reduction on the portion taken before the claimant's normal retirement age. The 2026 national guidance on claiming partial pensions confirms the component percentages and the requirement for a separate later application for the full old-age pension.
Partial claiming separates two decisions ordinarily collapsed in simplified models: how much of the accumulated pension right to liquidate, and whether to keep working. A person can receive 25% of an accrued benefit without an equivalent 25% cut in work hours; claiming may finance debt repayment, caregiving, a voluntary hours reduction, or a period of unemployment. It is consequently necessary to record benefit claiming, earnings and hours worked separately. The Finnish rule creates a practical test of why “early exit” is not a reliable synonym for “early pension.”
Australia: means-tested age pension and private accumulation
Australia's public Age Pension is generally payable at 67 or older, subject to residence, income, and asset rules, as explained in the official Services Australia eligibility guidance. The regulator's retirement-income explanation makes clear that private superannuation access can begin at different ages. The public Age Pension age is therefore not an all-encompassing retirement or pension-withdrawal age. An empirical analyst observing retirement at 60 must identify whether income arose from superannuation, employment, savings, another social payment, or household transfers.
The means test introduces another difference from a proportional contributory benefit: an individual's own resources affect the public Age Pension amount and even entitlement. Minimum age rules interact with asset decumulation and household income, so the incentive to work longer cannot be read directly from the public age threshold. It also illustrates why public program access, private retirement wealth, and actual withdrawal from employment must be modeled as separate variables.
Comparative interpretation
The six systems demonstrate several legal mechanisms. The United States offers early claiming with cohort-dependent lifetime reductions; the United Kingdom uses a relatively firm access age for its ordinary new State Pension; Germany adjusts access through credited career duration; Sweden links reference ages with differently timed benefit layers; Finland permits partial receipt with persistent reductions on the advanced portion; and Australia's means-tested public pension coexists with separate private access ages. None of these arrangements can be represented faithfully by an undifferentiated national “retirement age.”
The comparative conclusion concerns institutional structure rather than a ranking. A lower earliest age may improve liquidity while reducing monthly income; a higher age may preserve later monthly adequacy but require credible bridging support; an earlier unreduced age for very long careers may reward long contribution records while excluding persons with care interruptions. Every evaluation requires the actual law, its relevant date, the exact benefit component, the insured history, and the period over which the effect is measured.
Claiming Behavior, Incentives, and Actuarial Valuation
The choice set under a legal minimum
Let a person reach age a with an accumulated pension right, a current employment income option, and a future stream of possible public pension benefits. Eligibility rules define the feasible claim dates. The individual may compare a larger current pension stream with a smaller present stream and a higher eventual payment, depending on the system. The comparison is affected by expected survival, health, liquidity, the discount rate, household labor supply, taxes, means tests, and the ability or obligation to continue working. The legal minimum truncates the choice set: benefits cannot be obtained in the covered scheme before the stipulated age merely because a claimant would prefer them earlier.
The economist's choice set is consequently conditional. A person may be legally free to work longer but face involuntary dismissal, health limitations, caring responsibilities, insufficient demand for older workers, or a collective contract that restricts employment. A person with no liquid savings may attach disproportionate value to the earliest available pension even when delayed claiming could raise lifetime expected income. Conversely, a wealthy household may be able to defer without hardship and optimize tax or survivor-benefit outcomes. An adjustment mechanically calculated to approximate actuarial neutrality for a representative cohort does not guarantee neutrality for any one individual.
The foundational empirical insight of Coile and Gruber (2007) is that retirement behavior responds to the structure of future incentives rather than only this year's cash flow. Their forward-looking Social Security research develops measures of prospective entitlement changes and identifies incentive-related retirement responses using the U.S. Health and Retirement Study. The bibliographic landing page distinguishes the 2000 working paper from the 2007 published version; this monograph credits the latter for its scholarly result but links the openly accessible underlying working-paper record.
A transparent cash-flow formulation
Suppose the earliest admissible commencement age is a0, the unreduced reference age is af, and a person chooses a claim age c at or after a0. Let B(c,t) denote the real periodic payment at age t, S(t|a) the conditional probability of surviving to age t given survival at current age a, and r a real discount rate. A stripped-down expected present value is
PV(c|a) = Σ[t≥max(c,a)] S(t|a) · B(c,t) / (1+r)^(t−a).
The sum is over payment periods, and B may incorporate early reductions, delayed credits, indexation, or changes in entitlements from additional work. If tax treatment, household survivor provisions, benefit earnings tests or other public payments matter, their state-contingent effects should enter B or a separately defined household cash-flow function. This expression is a conventional discounted contingent-cash-flow calculation. It is not a new pension valuation theory and does not establish an optimal claim age without estimates of its inputs.
A comparison of two claim dates c1 and c2 asks whether PV(c1|a) exceeds PV(c2|a) under a specified probability and discount model. Higher mortality generally makes earlier cash flows relatively important; lower mortality favors the longer receipt of a later higher amount, conditional on a particular schedule. An increase in the discount rate also tends to privilege earlier transfers. An individual's private valuation may differ from a social-planner valuation because public financing, taxation, work incentives, spillovers, and insurance against longevity risk affect the societal outcome.
An early adjustment has at least two distinct effects. The immediate arithmetic reduction multiplies the pension amount by a factor below one. If the same person also retires earlier, the future accrual base may be lower because fewer contribution years or lower highest-year earnings enter the formula. The U.S. 35-year computation explanation illustrates the second effect. Treating the published reduction percentage as the complete difference between early and late claiming can therefore understate or overstate the combined income consequences of altered work histories.
A reproducible U.S. timing example
Consider a purely hypothetical U.S. claimant born in 1960 or later, with a primary insurance amount of $2,000 per month at age 67. Hold the primary insurance amount fixed solely for comparing claim timing. Under the SSA early-retirement schedule, commencement at 62 pays 70%, or $1,400 per month, before any other adjustments. Starting at 67 pays the assumed $2,000. The later claim yields $600 more per month once commenced, while the earlier claim supplies five years of payments first.
Without mortality, inflation adjustment, investment earnings, taxes, or benefit interactions, the age-62 claim pays 60 × $1,400 = $84,000 before the person's 67th birthday. After age 67, the age-67 claimant receives $600 per month more. The simple cumulative-cash break-even requires $84,000 ÷ $600 = 140 months, or 11 years and 8 months after age 67: approximately age 78 years and 8 months. The arithmetic describes an illustrative nominal-equal-dollar comparison and is neither a present-value equality nor a recommendation. Introducing survival probabilities, earnings credits, taxes, cost-of-living adjustments, or different benefit histories changes the comparison; changes in household survivor rights can change it further.
The example reveals why a monthly-benefit maximization objective and a liquidity or consumption objective can select different claim dates. Maximizing the payment at age 80 is not identical to maximizing expected household welfare at age 62. When workers lack an income bridge between job loss and their normal retirement age, a legal increase in the earliest age changes feasibility itself. A statutory reform should therefore be evaluated for its effects on both the payment schedule and the household's affordable transition path.
Benefit reductions and the meaning of neutrality
In defined-benefit pension design, an early commencement factor can be calibrated to reflect the extra expected years of payment. It is actuarially neutral under a specified set of assumptions when the discounted expected value of entitlements is equivalent across claiming dates, holding the pension base and relevant state contingencies constant. Age-specific survival, mortality improvements, indexing, payment timing and discounting are necessary inputs. There is no reason to assume a legislated flat monthly reduction rate produces exact neutrality for every birth cohort, health group, or household composition.
The OECD finds an average annual early-claim penalty of 4.4% in its stated set of contributory pension arrangements, but this is a comparative summary of effective scheme adjustments rather than a uniform actuarial valuation rule. It cannot be used without revaluation to calculate the lifetime effect of a particular country's reform. Likewise, a public pension could be designed with no actuarial penalty for long careers, a partial penalty on one component, or a variable deferred credit. Actual incentives depend on tax treatment, accrual, and the interaction of multiple programs.
For Germany's 35-year route, the 0.3% per-month permanent adjustment translates to 3.6% over twelve early months if the pension otherwise is held fixed. In Finland, the 0.4% per-month early deduction applies only to the selected portion of accrued pension. These examples are not comparable until the retirement-age reference, contribution years, accrual changes, payment base and affected fraction are normalized. A nominally larger percentage reduction in Finland need not cause a larger reduction in the entire future pension because only part of the accrued benefit is advanced.
For a separate illustrative Finnish calculation, assume an accrued old-age pension of €2,000 per month, a hypothetical age eligibility allowing a 24-month advancement, and election of a 25% partial pension. The advanced slice is €500 per month. A 0.4% monthly reduction for 24 months is 9.6% on that slice, so its immediate amount is €500 × 0.904 = €452 per month. The unadvanced accrued slice is not subjected to that early-claim reduction merely because the individual claimed a quarter. Later accrual and life-expectancy coefficients have deliberately been omitted from this teaching calculation; actual full pension outcomes depend on those rules.
Dynamic incentives, learning, and liquidity
People respond to the entire schedule of pension wealth, including the possibility of waiting one additional month or year, not merely a headline eligibility age. In their Austrian study of benefit incentives, Manoli and Weber (2016) use discontinuities and bunching to estimate responsiveness. Their published incentive-response analysis reports participation semi-elasticities roughly 0.1 to 0.3 under its design and emphasizes adjustment frictions and the timing of incentives. The examined employer-provided retirement benefit rules are not equivalent to all public old-age entitlements, so the result is evidence about behavioral mechanisms rather than a transferable universal pension-age elasticity.
Administrative complexity and imperfect information can affect whether an eligible person claims promptly, late, or at all. The same nominal option is valuable differently to an informed worker who can finance deferral and to a worker who lacks accessible advice or cannot withstand temporary income loss. The government can modify observed behavior by clarifying eligibility, simplifying applications, offering partial pensions, altering earnings tests, improving disability assessment, or requiring minimum notice of age reforms. These are policy implementation channels, not claims that pension information alone determines retirement.
Risk transfer across the life cycle
Public pensions insure longevity and, through redistributive or indexed components, can also protect against low lifetime earnings and inflation risk. An early reduced pension reallocates cash from later ages to earlier ages and can magnify financial vulnerability among very old survivors when the reduction persists. Deferral may improve longevity protection but expose a claimant to an unprotected interval before entitlement. The preferred allocation cannot be inferred from one average longevity estimate: health, marital status, mortality gradients, and survivor rights can change the practical welfare balance.
For this reason, short-term liquidity, long-term adequacy, and lifetime actuarial value must be reported as separate evaluative dimensions. A policy can be actuarially near-neutral on average and still have severe consequences for the subgroup whose health prevents continued employment. A policy can also preserve a minimum safety-net benefit at the normal age while leaving an earlier contributory benefit available at a reduced level. Such design interactions are central to understanding distributional outcomes of retirement-age policy.
Adequacy, Labor Supply, and Social Protection Substitution
Eligibility is not effective income security
Worldwide, pension receipt remains incomplete even when legislation provides age-based entitlements. The ILO report executive summary identifies global gaps and warns that the extent of pension coverage and the adequacy of benefits must be measured separately. In its World Social Protection Report 2024–26, the ILO estimates that 79.6% of persons above the statutory retirement age receive some form of old-age pension, leaving more than 165 million older persons beyond statutory age without a pension. This measure combines contributory and noncontributory channels. It is a population coverage measure, not evidence that 79.6% receive an adequate benefit or that everyone below a statutory age is without another form of income support.
The ILO reports that legal pension coverage and current contribution are different statistical quantities: workers may hold formal rights but lack qualifying contribution periods, or may participate in schemes without sufficient years to vest. The same ILO older-persons evidence describes substantial gender differences in contributory pension receipt and highlights the roles of precarious employment and migration. Therefore, when a government raises a minimum retirement age, a coverage analysis should examine people who newly fall between their labor-market exit and legal entitlement, alongside people whose formal pension rights never mature into effective receipt.
A pension's adequacy is also distinct from the right to receive it. Relative old-age poverty, measured through equivalized disposable household income, captures income shortfalls across pensions, other transfers, employment earnings and household resources. OECD's old-age poverty comparison reports a 14.8% average relative poverty rate among persons older than 65, compared with 11.5% for the population overall, in its latest available comparative series. The report also estimates higher old-age poverty among women, 16.9% compared with 11.7% among men. These population values are contextual distributional evidence, not causal estimates of the effects of any particular retirement-age law.
A social protection floor can partly decouple the basic anti-poverty guarantee from accumulated contributions. The ILO's Recommendation No. 202 income guarantee concerns nationally determined basic income security for older people. Its relevance to early exit is practical: people whose employment ends before the public old-age entitlement begins may need unemployment, disability, sickness, social assistance, or other lawful bridging income. The statutory pension-age change must be analyzed with those legal access channels and their benefit adequacy, not in isolation.
Labor-market exit versus pension claiming
The OECD estimates that the average effective age of labor-market exit in 2024 was 64.7 for men and 63.6 for women across OECD members. Its exit-age methodology and results construct a synthetic-cohort measure from changes in labor-force participation rates, rather than directly observing a universal individual retirement event. The report also notes that normal-age and effective-exit-age indicators are only moderately correlated, around 0.46 across countries. An observed national exit age above normal pension age can reflect continued work out of preference, low pensions, inadequate savings, or other institutions; it does not necessarily indicate a successful retirement-age incentive.
The same OECD source documents considerable international and gender variation in the timing of exit. Descriptive cross-country differences cannot be interpreted causally because employment conditions, education, health, the composition of retirement schemes, and benefit generosity co-vary with statutory age. A policy assessment should examine employment at 55–59, 60–64, 65–69 and other relevant groups, while tracking how the denominator is constructed and whether affected cohorts have actually crossed the new legal threshold.
Employment can rise without working hours, earnings quality or lifetime welfare rising proportionately. Some persons move into part-time work, jobs with lower wages, or intermittent attachment; others wait unemployed until the new threshold. A simple percentage-point increase in employment is useful but insufficient. An integrated measure includes earnings, working hours, health, job stability, disability and unemployment take-up, net taxes, total transfers, and poverty risk in the transition period.
The Austrian natural experiment: employment and substitution
Staubli and Zweimüller (2013) analyze Austrian reforms that raised the early retirement age from 60 to 62 for affected men and from 55 to 58.25 for affected women over the studied period. Their Journal of Public Economics study reports employment increases of 9.75 percentage points for men and 11 percentage points for women, while unemployment-insurance use rose by 12.5 and 11.8 percentage points, respectively. It finds comparatively small disability-claim spillovers in that setting and stronger employment responses for healthier and higher-wage workers.
This study is important for its identification strategy and multidimensional outcome accounting. It draws on reform-induced changes in eligibility across affected cohorts rather than interpreting a simple cross-country pension-age correlation. Yet its estimates arise from a specific Austrian institutional environment, cohort windows, benefit system, and labor market. They should not be applied mechanically to the present-day United States, Sweden, the United Kingdom, or countries with extensive informal employment. The employment effect and unemployment effect also describe partially different groups and transitions, so they cannot be interpreted as complementary percentages that add to 100.
One substantive implication follows directly: pension-account expenditure savings can overstate net public-budget savings when blocked early claims increase unemployment or disability support and when additional employment changes income-tax and social-insurance receipts. The published Austrian fiscal analysis explicitly considers spillovers and incremental tax revenue. A valid fiscal evaluation records changes in all major programs and associated receipts, using consistent horizons and avoiding double counting of transfers between government levels.
Later eligibility, job constraints, and poverty risks
Raising the earliest public pension age is most consequential for people who would otherwise have claimed at the prior age. For those able and willing to remain employed, a higher threshold can extend earnings and contribution accrual. For those unable to find work, it can create a cash-flow gap. For people in poor health, prolonged work may be infeasible; for caregivers, ongoing paid employment may compete with unpaid family responsibilities. These heterogeneous constraints help explain why a reform can raise the employment rate among affected high-wage workers and still deepen hardship for persons at the labor-market margin.
The distributional analysis must consider lifetime and pre-retirement inequality separately. Individuals with shorter healthy life expectancy may have less opportunity to recoup a later pension start. Individuals with high lifetime earnings can more easily self-finance deferral. Yet a long-career early access route can favor uninterrupted contributors over workers with lower earnings or care-related absences. Gender-based age differences can alter the distribution of benefit receipt, but equal statutory ages do not by themselves equalize pension amounts or employment histories.
The OECD's gender pension gap analysis documents unequal pension receipt patterns and discusses shorter or fragmented women's earning histories, survivor coverage, and differing retirement durations. Its 2025 edition uses country-specific data years, not a single 2025 statistical observation everywhere. These data support inclusion of gender and career interruptions in an assessment, while providing no direct estimate of what raising a specific pension age will do to the gender gap.
An adequate empirical design would estimate at least three effects: immediate income at the former eligibility ages, retirement income in later life, and the overall net-present-value change including work and transfers. Household-level outcomes matter where a spouse's employment or benefits buffer an individual's loss. The analysis should separate people whose benefits are delayed from people who qualify through an exception, whose contributions are credited, or whose entitlement exists but is not taken up.
Partial claiming and the work-retirement continuum
The Finnish partial pension framework offers a counterpoint to binary retirement models. If an individual may claim a quarter or half of an accrued pension while working, the legal age governs a liquidity option as much as the timing of labor exit. The 2026 public guidance on partial pension applications explains the election and resulting permanent reduction. Observing a pension payment in administrative data need not imply reduced employment or any workplace separation.
In Sweden, the first access date for an earnings-related public component and the age of guarantee pension access can differ under cohort-specific age rules. This makes adequacy analysis dependent on the composition of benefits and other resources. In Australia, the 67-year public Age Pension age sits alongside separate superannuation access dates as identified in the Australian retirement-income guidance. The benefit source matters: analyzing an early “pension withdrawal” without classifying the payer and program may assign the effect to the wrong legal reform.
Public finance and the full budget constraint
A public pension reform is commonly assessed for solvency, financing balance, and implicit liabilities. Restricting early payments may reduce immediate pension outlays and can increase pension contributions when people stay employed. However, an extended eligibility gap may increase unemployment or disability payments, social assistance, healthcare needs, and household reliance on informal transfers. The net fiscal effect also depends on whether replacement employment is created, whether older workers displace or complement other labor, and whether the pension adjustment permanently changes eventual benefits.
Let ΔP denote the change in pension expenditure, ΔU unemployment expenditure, ΔD disability and sickness benefits, ΔA social assistance, ΔT tax and contribution revenue, and ΔC administrative or other public costs. An illustrative government-wide budget change is ΔG = ΔP + ΔU + ΔD + ΔA + ΔC − ΔT, computed over a common discount horizon and clearly defined consolidation boundary. An apparent pension-saving reform has ΔP below zero, but the sign and size of ΔG depend on all terms. This identity is accounting, not an estimated effect. The Austrian evidence shows why the terms are empirically nonzero and why their values must be measured rather than assumed.
The intergenerational balance also depends on future promises: delaying a claim may raise annual payments through actuarial credits, potentially offsetting part of near-term cash savings. Linking ages to longevity may change both lifetime entitlements and expected years in employment. Distributional accounting requires comparing the same cohort under a credible counterfactual schedule while preserving real indexation, taxes and baseline earnings assumptions. Cash flow in the first budget year is not equivalent to a lifetime saving.
Intergenerational fairness and transitional rights
Reforms applied to people near their expected claim age can disrupt long-standing employment, saving, and caregiving plans. Legally phased cohort schedules and transparent notice support planning even when an individual dislikes the result. The UK cohort-based statutory timetable and the Swedish advance setting of reference ages illustrate two institutional approaches to predictability. Whether they are substantively fair requires a broader assessment of expectations, legal rights, health, lifetime work histories, and transitional protection.
A long-run fiscal rationale by itself cannot identify which birth cohorts should absorb the adjustment. People within the same cohort also have different capacities to continue employment and different expected pension durations. Comparing actuarial value across cohorts without decomposing those differences risks treating changes in health and contribution opportunities as mere preference variation. A scientifically responsible reform appraisal therefore reports the age rule, the transitional cohorts, the binding nature of the rule, the compensating measures, and the distribution of expected losses and gains.
Evidence Quality, Identification, and Sensitivity
The denominator problem in retirement-age statistics
Three denominators are often mistaken for one another: people who meet the statutory age condition, people entitled after all contribution or residence conditions, and people actually receiving benefits. A fourth set comprises people who stop working at a particular age. For contributory schemes, one can legally reach the minimum age but fail the contribution qualification. For means-tested programs, one can reach the age but have assets or income that prevent payment. For early earnings-related pensions, a person may satisfy eligibility and deliberately defer payment. A percentage of recipients over a given age cannot distinguish those mechanisms without further administrative data.
The ILO's pension coverage analysis distinguishes legal contributory coverage, active contributors, and older pension beneficiaries. Its 79.6% recipient indicator should be quoted with the phrase “persons above statutory retirement age” and with the definition of pension receipt. It should not be used to infer that the remaining population is formally ineligible, nor extrapolated to a precise number of potential early claimants. Legal age distributions and participation records are needed to produce the latter estimate.
The OECD effective-exit estimation method similarly uses changes in labor-force participation rather than universally observed individual claiming events. Its comparisons are suited to trends and relative differences within the defined methodology, with caution about absolute age interpretations. Pension-age legislation, labor-force-exit statistics, pension claiming data, and pension coverage indicators have different estimands. Mixing them would create a false apparent precision in age-policy effectiveness.
Legislative source dating and cohort mapping
For each policy, the applicable legal rule should be established from an enacted text or an official administering agency's description, with the legal effective date and the cohort to which it applies. The United Kingdom's birth-date age schedule shows why a statement such as “UK pension age is 67 in 2026” loses precision: during the phase-in, people reach entitlement at different ages depending on birth date. Sweden's recommended-age cohort table distinguishes determined ages from later forecasts. Germany's long-career eligibility rules require qualifying insured periods, making age alone incomplete.
An audit trail should record: official name and date of an act or rule; date of amendment and commencement; currentness of the administrative guidance; affected birth cohorts; relevant scheme; exact minimum and normal ages; other eligibility and credited-service conditions; and whether the reported rule is enacted or forecast. When public guidance summarizes detailed legislation, the summary can establish the currently administered rule, but it should not substitute for legal text in litigation or a jurisdiction-specific legal opinion. This dossier provides international analytical research, not advice on an individual's entitlement.
Causal designs and counterfactual retirement outcomes
A credible causal claim asks what the same person or cohort would have done under a different pension-age schedule. Retirement ages are often raised during periods of population aging, labor-market change and fiscal reform; before-and-after population averages therefore confound the change in pension law with other contemporaneous forces. Comparisons between countries further confound employment practices, coverage, health, contributions, and private pension design. A raw correlation between legal and effective age is descriptive evidence, not the estimated treatment effect of a reform.
Potential identification strategies include birth-cohort discontinuities at eligibility thresholds, difference-in-differences with valid parallel-trend assumptions, regression discontinuity where manipulation around an age threshold is limited, and administrative event studies tracking benefit and labor-force transitions. Each design answers a local question about affected cohorts, policy margins, and an observed period. Compliance, anticipation, benefit claiming deferral, sample selection and spillovers can defeat simple interpretations. A study that estimates employment within ages 60–62 after an age-eligibility reform does not necessarily estimate lifetime earnings or old-age poverty after age 80.
Mastrobuoni (2009) examines U.S. cohorts affected by changes in the normal retirement age and uses cohort discontinuities to investigate labor-supply responses. The published cohort-discontinuity study is relevant precisely because the American early retirement age remained available while the full-benefit reference age changed. It provides an identification contrast with studies of a binding minimum age: the legal margin, and therefore the affected choice set, differs. A causal estimate about the former cannot be treated as the effect of legislatively prohibiting early claims.
Staubli and Zweimüller (2013) identify consequences of raising the actual early eligibility threshold using Austrian reforms and report both employment and unemployment spillovers. Their results give stronger evidence for substitution mechanisms than a cross-sectional comparison of pension ages. Manoli and Weber (2016) investigate retirement responsiveness to financial incentives via benefit discontinuities in Austria. These papers examine distinct policies and local populations. An integrative reading finds that policy responses depend on whether the constraint is legal availability, the payoff to delayed work, or institutional barriers; the evidence does not establish one universal elasticity for all pension reforms.
A structured sensitivity audit
For any projected individual claiming value, conduct at least the following variations. First, change the age-specific survival assumption, preferably with cohort and socioeconomic stratification rather than only national period life expectancy. Second, vary real discount rates and pension indexation paths. Third, distinguish uninterrupted contributions from care, unemployment, migration, or informal-work interruptions. Fourth, model immediate and delayed claiming with and without continued work. Fifth, examine households with and without liquid wealth and survivor beneficiaries. Sixth, test the incidence of means tests and bridge benefits during the statutory gap. Seventh, vary the legislative transition timetable and permitted exceptions. Eighth, separate a pension-account perspective from consolidated-government and household perspectives.
These dimensions are connected. Higher discount rates change the present value of an early pension and also the value of additional contribution years. A later claiming age may generate higher annual payments yet lower lifetime received cash for a person with short survival. If earnings rise during continued employment, a pension system linked to final or best earnings may increase the accrual base independently of the delay bonus. In a means-tested system, drawing down private wealth before old-age entitlement can change the public benefit amount. Quantitative estimates therefore require a fully specified counterfactual, not one assumed representative life history.
Consider again the hypothetical American $1,400 early pension and $2,000 later pension. The nominal cumulative break-even at approximately 78 years and 8 months depends on payments being compared in the same unindexed dollars and on survival through the break-even date. A real discount rate above zero places more weight on the five earlier years and pushes the equivalent discounted break-even later, all else fixed; mortality weighting can change the relative expected value further. The result is a sensitivity direction, not a numerical actuarial neutrality assertion. One can calculate those valuations only after selecting and disclosing a mortality model, a discount curve, cost-of-living adjustment treatment, and any relevant survivor benefit structure.
Distinguishing estimates, projections, and claims of law
The OECD's current-age statistics describe rules used for a modeled 2024 full-career case. Its future age projections apply then-current legislated parameters to younger careers and are conditional on future policy continuity. The effective labor-market exit series estimates behavior under a specific synthetic-cohort methodology. These three quantities should never be placed in one table without labeling their observation period, cohort assumptions and estimand.
Similarly, an official pension-agency website is strong evidence for the currently administered benefit scheme but can simplify exceptions and be updated as policy changes. A peer-reviewed causal paper can be strong evidence for the effect of a historical reform on its defined sample, yet weak evidence for a current legal rule. An ILO instrument provides normative text, not a count of insured persons in a given jurisdiction. A global statistical report provides descriptive breadth, not the individualized eligibility detail needed to adjudicate a claim. Good research matches the source type to the assertion rather than ranking every source on one generic scale.
Evidence integration without false precision
The central comparative result can be stated with a clear evidentiary division. Authoritative legal and administrative sources establish the existence of early and normal ages, contribution conditions, reductions and cohort schedules in the selected systems. OECD and ILO statistics establish the scale of age differences, labor-market withdrawal and coverage or poverty concerns at published reference dates. Published quasi-experimental studies establish local behavioral responses to specified historic reforms. Formal actuarial equations and simple arithmetic clarify how those rules translate into cash flows under explicitly chosen assumptions. No source in this dossier supports a universal estimate of the welfare-optimal global retirement age, and none is presented as doing so.
A limitation of the present research is that it does not assemble a new microdata sample or run a new econometric identification strategy. It does not independently adjudicate every subsequent 2026 amendment within every OECD country, inspect confidential administrative records, or model population-specific mortality and household welfare with empirical estimates. Those would require a defined country panel, versioned legislation, and auditable data. The absence of those tasks is disclosed as study scope, while the questions examined with the available official texts and published research are resolved in the analytical synthesis below.
Integrated Scientific Synthesis and Research Agenda
The legal threshold and the economic transition
The comparative evidence supports an institutional interpretation of pension-age policy. There is a legal transition from non-eligibility to eligibility; a financial transition from earnings or other support to public pension income; and a labor-market transition from continued work to reduced or ceased participation. These transitions may occur in different months or years. They are also affected by legal changes through different causal mechanisms. Where early claiming is legally prohibited before a threshold, the reform modifies the feasible time of receipt; where claiming remains allowed but the full-benefit age rises, the reform changes the cash value of early receipt; where partial pensions are allowed, the timing and quantity of pension receipt are separately selectable. These are descriptive analytical distinctions supported by the examined institutions, not a new philosophical doctrine or scientific model.
The same observation explains why a policy judged by headline pension age can look more coherent than it is. A person with 45 recognized years of insured work may access a pension at a different age from a person with 35 years, as demonstrated by the German multiple old-age pension routes. An early recipient in Finland may still work full-time under the partial pension option. A worker in the United Kingdom may stop work before State Pension age but cannot ordinarily convert that event into an early new State Pension claim under the eligibility rules. These cases validate the need for separate legal, payment and employment records.
A government-wide interpretation of reform effects
The Austrian early-age reform findings demonstrate that a statutory delay can raise older-worker employment and still increase unemployment-insurance reliance. This has a clear institutional consequence: gross public pension savings do not establish net government savings. Pension authorities, employment agencies, disability administration, social-assistance programs and tax authorities may all experience correlated changes. Evaluation should use a common counterfactual and an explicit consolidated fiscal boundary. A savings estimate restricted to the pension account is a partial expenditure estimate, not a government-wide reform appraisal.
The same evidence indicates that changes in employment are heterogeneous. A worker who can remain employed under a higher age rule gains more opportunity to accrue earnings and contributions than a worker who is healthy enough to work but cannot obtain a job. Health and wage differences matter to the effect of the policy. An international assessment should therefore report employment, unemployment, disability transfers, earnings, and poverty by relevant socioeconomic groups. Such disaggregation is essential to distinguish aggregate improvement from a shifted burden borne by less employable or less healthy groups.
Distributional adequacy across early and late retirement
The OECD old-age poverty findings and the ILO's older-age pension coverage estimates show why a retirement-age reform should not be scored solely on the average effective age or the pension scheme's funding balance. Older-age income adequacy includes benefit amount, effective access, household resources, and continued exposure to health and longevity risks. An early reduced pension may protect against a short-run bridge-income crisis while depressing benefit income much later. A later minimum access age may protect the annuity amount but leave people with an unmet cash-income need before age eligibility. The relative welfare consequences depend on the circumstances of the affected person and the compensating design of other benefits.
Gender and career histories amplify this issue. Lower wages, interrupted formal employment, and unpaid care can affect contribution qualification and the size of an earnings-related pension. The OECD's gender pension evidence also demonstrates that equal legal ages and equal pension outcomes are different questions. Cohort schedules, caring credits, survivor benefits, disability paths, and guaranteed minima should be examined together where the goal is equitable old-age protection.
The proper policy questions
The study yields six substantive evaluative questions for policymakers and researchers. Which specific benefit component becomes legally payable at the proposed age? How does the age change the expected payment schedule for each affected cohort? Which workers can extend earnings and which instead enter other transfer programs? How does the change affect poverty and income security before and after the threshold? How does the reform distribute costs across career histories, health groups, genders, and generations? What does the change cost or save across the entire public budget after taking behavioral and administrative spillovers into account? These questions translate the existing evidence into a transparent evaluation agenda without creating an unverified new model.
An immediate research priority is a version-controlled, component-level comparative archive recording enacted changes and affected birth cohorts, validated against official legal instruments. A second priority is linked administrative study of employment, job loss, earnings, contributions, sickness or disability benefits, unemployment support, and pension claims around early eligibility thresholds. A third is household-level lifetime income analysis incorporating differences in longevity, wealth and caregiving histories. A fourth is the study of partial and phased pensions, where early claiming may preserve employment rather than end it. Each topic is a research task requiring its own dataset and identification design, not a claim that the present dossier has already estimated those effects.
Boundary of the findings
The strongest findings here are definitional and comparative: official schemes use materially different early-eligibility mechanisms; the available OECD indicators are not interchangeable; and credible historical research detects both employment gains and transfer substitution after eligibility-age reforms. The normative implication is that reforms should preserve clear legal entitlements and assess their social-protection consequences. The empirical implication is that effect sizes are specific to cohorts, schemes and labor markets. The actuarial implication is that benefit timing, reduction factors and survival assumptions must be modeled together. These conclusions are supported by reviewed source content and transparent derivation rather than by asserted originality of a new doctrine.
Conclusion
Public pension minimum retirement ages determine access to a legally specified income stream within a wider system of old-age protection. They are best understood alongside the full or unreduced age, required contribution history, benefit formulas, special entitlement paths and the actual options available to an older worker. Treating one reported retirement age as a national summary obscures critical differences between a prohibited early claim, an early claim with a permanent reduction, an unreduced long-career pension, a partially drawn pension, and a means-tested income guarantee.
The comparative legal evidence establishes that national systems structure these margins differently. U.S. Social Security makes early claiming from age 62 possible with substantial cohort-dependent reductions; the British new State Pension normally begins at the legislated eligibility age; Germany uses differentiated credited-service routes; Sweden operates with linked recommended ages and benefit layers; Finland permits partial pension commencement; and Australia separates the public Age Pension age from the private-superannuation access system. OECD statistics and ILO evidence show why the choice of population, component and cohort matters to interpretation. Credible Austrian reform research establishes that delaying early eligibility can increase employment while expanding unemployment-benefit use, demonstrating the need for government-wide fiscal and distributional analysis.
The resulting scientific conclusion is precise: the effect of a retirement-age rule is mediated by legal eligibility, the lifetime payment schedule, the economic ability to continue working, and the availability of alternative protection. An effective reform appraisal must preserve the distinction between statutory age, first pension receipt and labor-market exit; verify rules for the relevant birth cohorts; assess early and later-life income adequacy; trace spillovers through other social programs; and report uncertainty about mortality, discounting and counterfactual behavior. The analysis supplies an internally edited evidence-based account and a reproducible audit logic, while leaving future causal estimation to studies with the necessary original datasets.
References
Coile, Courtney, and Jonathan Gruber. 2007. “Future Social Security Entitlements and the Retirement Decision.” Review of Economics and Statistics 89 (2): 234–246. Earlier openly accessible version: “Social Security and Retirement,” NBER Working Paper 7830 (2000). https://www.nber.org/papers/w7830
Deutsche Rentenversicherung. n.d. “Altersrenten für langjährig und besonders langjährig Versicherte.” Official German pension administration; accessed November 17, 2026. https://www.deutsche-rentenversicherung.de/DRV/DE/Rente/Allgemeine-Informationen/Rentenarten-und-Leistungen/Altersrente-fuer-langjaehrig-Versicherte/Altersrente_fuer_langjaehrig_Versicherte.html
Deutsche Rentenversicherung. n.d. “Benefits: Old-age Pension for the Especially Long-term Insured.” Official English-language program description; accessed November 17, 2026. https://www.deutsche-rentenversicherung.de/DRV/EN/Leistungen/leistungen?nn=6a097e0fca84ce5471c1c25e
Finnish Centre for Pensions. n.d. “Partial Old-age Pension.” Official Finnish pension system description; accessed November 17, 2026. https://www.etk.fi/en/finnish-pension-system/pensions/earnings-related-pension-benefits/partial-old-age-pension/
Government of the United Kingdom. 2014. “State Pension Age Timetable.” Department for Work and Pensions; accessed November 17, 2026. https://www.gov.uk/government/publications/state-pension-age-timetable/state-pension-age-timetable
Government of the United Kingdom. n.d. “The New State Pension: Eligibility.” GOV.UK; accessed November 17, 2026. https://www.gov.uk/new-state-pension
International Labour Organization (ILO). 1952. Social Security (Minimum Standards) Convention, No. 102, especially Part V, Articles 25–30. NORMLEX official instrument. https://normlex.ilo.org/dyn/nrmlx_en/f?p=NORMLEXPUB%3A12100%3A0%3A%3ANO%3A%3AP12100_ilo_code%3AC102
International Labour Organization (ILO). 2012. Social Protection Floors Recommendation, No. 202. NORMLEX official instrument. https://normlex.ilo.org/dyn/nrmlx_en/f?p=NORMLEXPUB%3A12100%3A0%3A%3ANO%3A%3AP12100_ILO_CODE%3AR202
International Labour Organization (ILO). 2024. World Social Protection Report 2024–26: Universal Social Protection for Climate Action and a Just Transition. Executive Summary. https://webapps.ilo.org/static/english/reports/flagship/world_social_protection_report_2024-26/Executive%20Summary.html
International Labour Organization (ILO). 2024. World Social Protection Report 2024–26: Universal Social Protection for Climate Action and a Just Transition. Chapter 4, section 4.3, “Social Protection for Older Women and Men.” https://webapps.ilo.org/static/english/reports/flagship/world_social_protection_report_2024-26/Chapter%2004.html
Manoli, Day, and Andrea Weber. 2016. “Nonparametric Evidence on the Effects of Financial Incentives on Retirement Decisions.” American Economic Journal: Economic Policy 8 (4): 160–182. https://swlb2.aeaweb.org/articles?id=10.1257%2Fpol.20140209
Mastrobuoni, Giovanni. 2009. “Labor Supply Effects of the Recent Social Security Benefit Cuts: Empirical Estimates Using Cohort Discontinuities.” Journal of Public Economics 93 (11–12): 1224–1233. https://www.sciencedirect.com/science/article/pii/S0047272709000875
OECD. 2025. Pensions at a Glance 2025: OECD and G20 Indicators. Paris: OECD Publishing. Chapter “Current Retirement Ages.” https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/current-retirement-ages_0f63b747.html
OECD. 2025. Pensions at a Glance 2025: OECD and G20 Indicators. Chapter “Future Retirement Ages.” https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/future-retirement-ages_23752280.html
OECD. 2025. Pensions at a Glance 2025: OECD and G20 Indicators. Chapter “Effective Age of Labour Market Exit.” https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/effective-age-of-labour-market-exit_fd1212b7.html
OECD. 2025. Pensions at a Glance 2025: OECD and G20 Indicators. Chapter “Old-Age Income Poverty.” https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/old-age-income-poverty_3cd946d7.html
OECD. 2025. Pensions at a Glance 2025: OECD and G20 Indicators. Chapter “Gender Pension Gap.” https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/gender-pension-gap_90ed13b5.html
OECD. 2025. Pensions at a Glance 2025: OECD and G20 Indicators. Chapter “Recent Pension Reforms.” https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/recent-pension-reforms_146d2687.html
Pensionsmyndigheten (Swedish Pensions Agency). n.d. “Recommended Retirement Age: When Can I Start Drawing My National Public Pension?” Official agency guidance; accessed November 17, 2026. https://www.pensionsmyndigheten.se/other-languages/english-engelska/english-engelska/recommended-retirement-age
Services Australia. n.d. “Who Can Get Age Pension.” Australian Government; accessed November 17, 2026. https://www.servicesaustralia.gov.au/who-can-get-age-pension?context=22526
Australian Securities and Investments Commission (ASIC). 2026. “Age Pension and Government Benefits.” Moneysmart, updated September 28, 2026. https://moneysmart.gov.au/retirement-income-sources/age-pension-and-government-benefits
Suomi.fi. 2026. “Claiming Your Old-Age Pension.” Suomi.fi, guidance updated March 2026; accessed November 17, 2026. https://www.suomi.fi/guides/old-age-pension/claiming-your-pension/claiming-your-old-age-pension
Social Security Administration (SSA). n.d. “Retirement Age and Benefit Reduction.” U.S. federal agency guidance; accessed November 17, 2026. https://www.ssa.gov/benefits/retirement/planner/agereduction.html
Social Security Administration (SSA). n.d. “Early or Late Retirement?” Office of the Chief Actuary, U.S. federal agency; accessed November 17, 2026. https://www.ssa.gov/oact/quickcalc/early_late.html
Social Security Administration (SSA). n.d. “You Can Receive Benefits Before Your Full Retirement Age.” U.S. federal agency guidance; accessed November 17, 2026. https://www.ssa.gov/benefits/retirement/planner/applying2.html
Social Security Administration (SSA). n.d. “The Age You Start Receiving Benefits and the Age You Stop Working.” U.S. federal agency guidance; accessed November 17, 2026. https://www.ssa.gov/benefits/retirement/planner/stopwork.html
Social Security Administration (SSA). n.d. “Retirement Age Calculator: Why Did the Full Retirement Age Change?” U.S. federal agency guidance; accessed November 17, 2026. https://www.ssa.gov/benefits/retirement/planner/ageincrease.html
Social Security Administration (SSA). 2019. “Program Explainer: Benefit Claiming Age.” Office of Retirement and Disability Policy, November 2019. https://www.ssa.gov/policy/docs/program-explainers/benefit-claiming-age.html
Staubli, Stefan, and Josef Zweimüller. 2013. “Does Raising the Early Retirement Age Increase Employment of Older Workers?” Journal of Public Economics 108: 17–32. doi:10.1016/j.jpubeco.2013.09.003. Open full-text manuscript: https://pmc.ncbi.nlm.nih.gov/articles/PMC3851570/
Staubli, Stefan, and Josef Zweimüller. 2013. “Does Raising the Early Retirement Age Increase Employment of Older Workers?” Journal of Public Economics 108: 17–32. Publisher record. https://www.sciencedirect.com/science/article/pii/S0047272713001734