RESEARCH DOSSIER
Long-Term Care Insurance Gender Inequality and Informal Care Burdens
Abstract. Long-term care insurance pools financial exposure associated with sustained functional dependency, while its realized effects reach beyond the insured person to family members who supply unpaid assistance, change paid employment, experience diminished pension accumulation, and absorb costs that do not enter the insurer's claims ledger. The distribution of these burdens is systematically associated with gender, age, household structure, income, health, disability, and institutional arrangements. This monograph undertakes a source-critical international investigation of long-term care financing and the gendered economics of informal care. It integrates original public time-use and household statistics, comparative OECD affordability estimates, ILO and European Union care research, documentary analysis of statutory and private insurance mechanisms, and identified empirical studies of Japanese and German reforms and private insurance spillovers. It separates care needs, assessed eligibility, insurance payment, actual formal care delivery, and informal family time as distinct objects of measurement. Women form majorities of people requiring care at older ages and of informal caregivers in the reported international datasets; however, insurance reforms can alter male and female labor supply differently because program design, employment margins and family choices differ. The analysis distinguishes cash transfers from service purchasing, interprets actuarial risk pooling alongside household resource costs, and assesses the strength and boundaries of observational and quasi-experimental evidence. Its original contribution is a comparative scholarly synthesis, systematic source-to-claim interpretation, and identification of testable research questions within established care economics. It claims no newly created theory, proprietary model, novel causal estimate, systematic meta-analysis, or original microdata.
Keywords: Long-term care insurance; gender inequality; informal care; unpaid eldercare; family caregivers; long-term services and supports; household production; care dependency; formal care; risk pooling; insurance affordability; cash benefits; in-kind benefits; occupational opportunity cost; labor supply; pension gaps; caregiving intensity; actuarial pricing; welfare incidence; social protection; caregiver health; quasi-experimental evaluation; gendered distribution; international comparisons.
Introduction: Insurance and the Allocation of Care
Long-term care insurance operates within a social system in which financial protection, service delivery and household labor are closely linked. A person dependent on assistance with washing, dressing, eating, mobility or supervision may receive help from professionals, family members, friends, paid domestic workers or combinations of these arrangements. The formal cost of care is only one part of the household's exposure. An adult daughter may reduce paid work to supervise a parent; an older spouse may perform physically demanding assistance; relatives may coordinate care across cities; and a person living alone may have no realistic source of family care. A payment can satisfy an insurer's contractual obligation while leaving these arrangements almost unchanged.
The scientific question is which costs are transferred, which remain with households, and who ultimately receives the economic value of insurance. A public contribution, private premium, insured cash payment and purchased nursing service represent different transactions. The recipient of insurance benefits need not be the person whose time or earnings change. If care is financed but no qualified worker is available, nominal coverage may coexist with unmet need. If cash is paid to a care recipient whose daughter supplies most assistance, the financial payment may improve the household's liquidity without releasing caregiving time. If formal services become available, the benefit may preserve employment, training, health or autonomy for family members as well as improve care for the recipient.
The OECD's 2024 international affordability study estimates that close to one in four older people in the examined countries have some long-term care needs. The OECD informal-carer indicators report that women are 61% of informal carers across 26 OECD countries; the sample primarily concerns older respondents. For the United States, the 2023–2024 American Time Use Survey estimates 38.192 million eldercare providers among persons aged at least 15, of whom 20.990 million were women and 17.201 million were men. Each statistic has its own population, time frame and definition. These measurements support investigating sex-differentiated care distributions; they cannot be added together to produce a global prevalence or causal effect.
The research investigates the distribution of dependency, unpaid care time, insurance access, contributions and benefits. It compares statutory programs, voluntary insurance and direct caregiver support, distinguishes economic mechanism from policy aspiration, and evaluates selected causal studies at the level of their design. The research method is a structured critical review and documentary comparative analysis with independent arithmetic reconstruction of published aggregates and transparent textbook accounting identities. It does not represent an independently collected survey, an insurer microdata study or a newly estimated econometric analysis.
Its unit of analysis changes with the question. For the probability of long-term care need, the denominator is a defined population at risk. For access to insurance, it is the eligible or potentially insurable population. For benefit adequacy, it is the insured person or household with a specified care need. For informal time, it is the caregiver or the household supplying care, with hours and activity definitions specified. For social incidence, it is the set of people financing and receiving protection over the relevant period. Carrying these denominators through the argument is essential to a responsible interpretation of gender inequalities.
Historical and Conceptual Foundations
Functional dependency, care and insured entitlement
Long-term care describes sustained assistance associated with limitations in functioning and capacity to manage daily life. Personal activities of daily living often encompass washing, dressing, eating, toileting, transfer and mobility. Instrumental tasks may include shopping, cooking, medication management, domestic activities and transport. Cognitive impairment can create supervision requirements even when some physical activities remain possible. The WHO's International Classification of Functioning offers a broader analytical vocabulary for bodily functions, activities, participation and environmental conditions. This classification supports measurement of functional need; particular statutory and contractual eligibility rules are separately determined.
An insurance entitlement exists only when a covered condition meets the relevant assessment standard and any contractual or administrative requirements. Singapore's national program, for instance, requires assessed inability to perform at least three of six specified activities for a CareShield Life claim, as set out in the CareShield Life eligibility guidance. Other schemes recognize graded dependency or a need for permanent supervision. A diagnosis, functional limitation, insurance eligibility and actual receipt of a reimbursed service describe different stages. A person may be dependent without qualifying for an insured claim, or meet a statutory threshold yet lack an appropriate local service provider.
Four related populations must therefore be distinguished throughout comparative research. The need population includes people requiring assistance under a defined functional measure. The entitled population includes those meeting a particular program's rules. The served population includes those receiving paid services or cash benefits. The caregiver population includes persons providing unpaid assistance, including support provided to people outside any formal insurance scheme. Their overlap changes with law, survey construction and household arrangements. Neither approval rates among claimants nor volumes of reimbursed home-care hours are direct measures of total social need.
A fifth distinction concerns intensity. Someone who drives a relative to an appointment once a month and someone who supervises dementia-related safety risks throughout every day can both report being an informal caregiver. Binary participation identifies the extensive margin of care, whereas hours, daily frequency, task type and duration measure intensive margins. Sleep interruption, travel and the responsibility to remain on call may be poorly captured by conventional time-use diaries. Service replacement estimates should therefore specify which components of unpaid care a hired worker could substitute and which relational or supervisory activities would continue.
Historical movement from household responsibility to collective finance
For much of history households, local communities, religious organizations and charities supplied care through varied reciprocal and hierarchical arrangements. Modern welfare-state programs increasingly developed pooled financing for dependency during the twentieth century as longevity, urbanization, labor-market participation and professional care changed the feasible allocation of responsibility. Germany introduced compulsory long-term care insurance in 1995, with a combination of care services and monetary support for family-based arrangements. The German federal insurance guidance describes its contributory basis and partial-benefit design.
Japan's public long-term care insurance began in 2000 and later reforms modified the availability of services for particular dependency categories. The 2006 policy change examined by Miyawaki and colleagues provides evidence about what can happen when formal support becomes less available. The Netherlands, Singapore, the United States and the United Kingdom demonstrate additional forms of financing and administration. These historical developments expanded the possible role of collective risk sharing while preserving distinct national choices about family obligations, income support, provider ownership and the boundary between health treatment and social care.
Institutional categories require precision. A voluntary long-term care policy pools a contractual risk among persons who purchase and maintain coverage, subject to underwriting, premium assumptions, benefit definitions and claims rules. Compulsory social insurance generally combines redistribution and pooling, with financing conditions tied to statutory participation. Tax-financed care support distributes costs across the public budget, frequently using income, residence or need criteria. Direct payments to unpaid caregivers may be administered as social benefits rather than as insurance claims on the recipient's policy. A common language of “coverage” does not imply common beneficiaries or obligations.
The gendered dual position: receiving and providing care
Gender appears on both sides of the dependency relationship. In the OECD's growing long-term care needs chapter, women account for approximately two-thirds of older persons with low, moderate and severe needs across analyzed countries. The greater number reflects a combination of longevity, exposure and household circumstances; it is a composition among people with need, not a gender-specific insured claims frequency. It must be conditioned on age, care severity, country and enrollment when used in actuarial comparisons.
Women also provide a disproportionate amount of unpaid care. The ILO's global care-work assessment reports that women provide 76.2% of unpaid care-work hours across its study population. This is an estimate for unpaid care broadly, including childcare and other household care, and must never be relabeled as the share of eldercare hours. The OECD's 61% women-carer figure instead describes the people identified as informal carers across 26 national surveys. Different denominators—hours versus people, all care versus eldercare, global versus selected OECD settings—explain why these numbers answer different questions.
Over a life course, an individual can contribute to a public scheme during employment, provide unpaid assistance to older relatives, lose earnings or pension credits during care-intensive periods and later require support herself. The empirical magnitude of such a sequence depends on family structure, wage and career history, social contributions and the available form of protection. It should not be inferred from women-only averages. Women differ materially by income, health, formal-care access and migration; men can also experience severe dependence or extensive caregiving. An intergenerational account must compare actual trajectories, not assume one household type is universal.
Scientific concepts and causal pathways
Several conventional economic mechanisms organize the analysis. Risk pooling converts uncertain individual expenditures into expected contributions and contingent payments. Care substitution occurs when a paid service replaces a task otherwise performed by an informal carer. Liquidity support allows a household to finance services or consumption without immediate asset liquidation. Labor-supply adjustment occurs when a caregiver changes market hours or participation in response to dependency, benefit rules and service costs. Caregiver health can respond through physical demands, exhaustion, time pressure or relief supplied by respite services.
These mechanisms are mediated by benefit design. Cash can increase the ability to support home arrangements yet strengthen incentives to supply unpaid family time. In-kind formal services may release time for paid work but can require coordination and may be difficult to access. An entitlement to reimbursement offers little practical substitution where the covered service is unavailable. Existing empirical studies identify particular parts of these pathways rather than a universal law governing all countries.
Caregiving Time and the Gendered Division of Labor
Reconstructing U.S. caregiver denominators
The U.S. Bureau of Labor Statistics (2025) defines an unpaid eldercare provider as a person at least 15 years old who helped a person aged 65 or older with a condition related to aging during the relevant retrospective period. Its sex-specific eldercare table estimates a civilian noninstitutional population of 272.126 million and 38.192 million providers for 2023–2024. The calculation 38.192 divided by 272.126 gives approximately 14.0% prevalence. The women-provider count, 20.990 million, divided by the total provider count gives about 55.0%; the male count is 17.201 million. The published sex counts sum to 38.191 million rather than the rounded total of 38.192 million, a one-thousand discrepancy expected from separately rounded estimates.
The same table shows that among women in the underlying population 15.1% provided eldercare, compared with 12.9% of men. The figures 55.0%, 15.1% and 12.9% have different denominators. One describes the female share of carers; the others describe the probability of caregiving within each sex-specific population. A manuscript that fails to distinguish these quantities can substantially exaggerate or understate a gender effect. None of these descriptive figures proves that insurance access causes women to enter or leave caregiving.
Among identified eldercare providers, 28% engaged in care on an average diary day and spent approximately 3.9 hours on days when care was undertaken, as documented in the BLS eldercare release. This conditional average cannot be interpreted as 3.9 hours every day for all 38.192 million providers. The total annual time burden requires activity-day probabilities, the distribution of intensity and relevant survey weights. It also requires explicit treatment of overlapping activities, travel, passive supervision, overnight interruptions and care provided to more than one recipient.
The Federal Reserve's household care-work survey uses a broader adult-care definition encompassing aging, illness or disability. Its reported 2024 proportions for unpaid adult care are approximately 17% overall, 18% for women and 16% for men. These figures cannot be treated as a later wave of the BLS eldercare statistic. Care recipients younger than 65, question wording, activity thresholds, recall periods and survey design differ. Harmonization requires reconstructing each original outcome before comparison.
European and international comparability
The OECD's 2025 informal-carer comparison reports that 13% of people aged 50 or over provide informal care in the surveyed countries. Approximately 61% of carers in 26 countries are women. The indicator draws on national and regional surveys, with variations in the definition of daily and weekly caregiving, self-reported roles and age composition. It is particularly useful for showing a broad structural imbalance among mature-age carers; it is not a direct count of all female carers at every age.
The European Institute for Gender Equality (EIGE) fielded the 2024 CARE Survey across the 27 European Union states with more than 65,000 participants aged 16–74. Its CARE Survey research report, published in 2026, measures unpaid care alongside household work, individual activities and interactions with formal services. Its scope reaches beyond eldercare and excludes some carers older than 74. This study contributes context about gendered responsibilities and care access, but its survey prevalence should not be combined numerically with the OECD 50-plus eldercare indicator as though identical populations were sampled.
The ILO's Care Work and Care Jobs builds a wider picture of unpaid labor, paid care work and gendered employment arrangements. The associated analytical distinction is crucial: unpaid household support may be invisible in monetary care spending while displacing time that could otherwise be used in market work, other family work, education or leisure. Neither a money-only claims ledger nor an hours-only care survey captures the whole welfare effect. The linkage requires carefully matched concepts and time horizons.
A cross-country dataset should preserve the exact reference period, age range, care-receiver definition and distinction between care to a household member and someone outside the household. Multi-recipient caregiving needs separate treatment. Definitions of formal care vary: a professional visiting a home may be counted as formal care even when paid privately, while a state-funded family allowance may be classified as cash assistance. Comparison should report uncertainty and source-level limitations rather than present harmonization as complete when the underlying surveys differ.
Care intensity and replacement valuation
Informal care time carries an economic opportunity cost, but methods for valuation have distinct interpretations. A replacement-cost method multiplies hours of substitutable care by the wage and employment cost of suitable paid workers. An opportunity-cost method values the actual labor, leisure or other activities forgone by the informal carer. Observed lost earnings estimate a market response to care; they do not include all changes in health, training, pension accrual or emotional well-being. A willingness-to-pay approach may capture preferences for the care arrangement under explicit conditions. These measures should appear as alternative lenses, rather than be added automatically.
For an annual caregiving episode, a transparent replacement-cost identity is V = h × w, where h denotes hours of substitutable activities and w the cost per appropriate hour of professional service. At an illustrative 10 hours per week over 52 weeks, h is 520 hours. At a hypothetical $15 per hour the annual replacement value is $7,800; at $25 it is $13,000. These are arithmetic examples, not estimates of a current national wage or typical eldercare burden. A valid application requires local price data, the relevant task mix, time spent on standby, transportation and administrative or coordination costs. If lost earnings are used as an alternative opportunity valuation of the same hours, adding both in full double-counts time.
Intensity may be more consequential for employment than the binary condition of being a carer. Repeated overnight supervision, emergency travel and unpredictable personal-care requirements can undermine regular work schedules even when average recorded daily hours are moderate. Conversely, periodic shopping, telephone coordination and transport may be compatible with employment. A defensible analysis records intensity categories, duration, primary responsibility and unpredictable demands. Changes in hours should be interpreted alongside the quality and safety of care received.
Employment, pensions and cumulative consequences
A caregiver's decision to reduce paid employment can produce a chain of economic consequences. Immediate lost wages and employer benefits can be followed by slower occupational progression, lower retirement contributions and reduced capacity to maintain private insurance. Time away from formal employment can become especially expensive when benefits vest or pension entitlements depend on contribution histories. These cumulative outcomes require longitudinal data; a snapshot of weekly hours cannot identify them.
Barszczewski, Milstein, Lee, and Llena-Nozal (2025) analyze women's responses to worsening parental health using SHARE observations from 24 European countries. Their economic-cost study of parental care reports that parental ill health is associated with higher probabilities of informal care and substantial reductions in labor supply among affected women in selected European regions. The abstract places some regional reductions in the 30%–70% range under its study-specific effect definitions and provides modeled GDP consequences of 0.37%–0.45% across certain regions. These findings cannot be simplified into a claim that 30%–70% of all women leave employment. The age window, sample construction, outcome specification and extrapolation assumptions are part of the result.
This research is informative because parental health transitions supply more temporal structure than a static comparison between caregivers and noncaregivers. Several identification challenges remain. A parent's worsening health can affect an adult child through stress, travel, financial transfers and other pathways apart from care hours. Survey attrition may be correlated with caregiving burden. Regional labor markets and formal-service entitlements differ. A causal interpretation requires assumptions about those channels; the published estimates are evidence about a specified European population and period, not universal parameters for actuarial pricing.
The German reform study by Geyer and Korfhage (2018) provides a useful counterpoint. Their labor-supply analysis of Germany's reform evaluates the 1995 introduction of public long-term care insurance using difference-in-differences. The reported negative labor-supply response was pronounced for men rather than women in the examined setting. This does not contradict female concentration in caregiving. A reform's effect is conditional on baseline work participation, retirement margins, benefit incentives and the form of family care. Someone already outside paid employment cannot reduce paid work on the measured margin, even if care hours or stress increase.
The question of pension effects requires a distinct empirical dataset. It cannot be answered by annual employment correlations alone. Proper measurement follows years of reduced work, retirement contribution records, any credited caregiving periods, survivor benefits and the conditions of a future dependency episode. Gender gaps in care and pension outcomes may share antecedents such as unequal paid wages and working-time arrangements; causal models should not attribute the entire retirement disparity to caregiving without identifying these other channels.
Caregiver health, well-being and choice
Unpaid care can express reciprocity, affection and personal preference, and can also involve fatigue, injuries, anxiety, disrupted sleep and social isolation. A good empirical welfare analysis preserves the caregiver's own valuation of care, the recipient's autonomy and dignity, and the difference between chosen family involvement and constrained provision forced by service scarcity. Treating every unpaid hour as a welfare loss would ignore valued relationships. Treating every unpaid hour as free would omit a large resource commitment and the consequences of limited choice.
Miyawaki and colleagues (2020) examine a Japanese reform that reduced formal care availability for selected lower-needs beneficiaries. Their quasi-experimental reform evaluation uses repeated cross-sections of 12,764 co-resident primary caregivers across 2001–2016 and reports a relative increase of 7.4 percentage points in caregiving exceeding three hours a day, along with declines in formal home-help and day-service use and deterioration in selected health indicators. These are treated-versus-comparison contrasts under a defined reform, not a prediction that every increase in insurance generosity reverses these exact outcomes.
The Japanese finding supports a plausible service-substitution mechanism: as formal provision contracted, household work increased in the affected group. The authors' matching and difference-in-differences approach addresses observed confounding and time changes but still depends on pre-reform comparability, accurate severity classification and stable measurement. It studies co-resident caregivers, not all caregivers in Japan. Health measures include self-reports that can vary with expectations and response patterns. These boundaries matter to any proposal to transfer the numerical result to another country's insurance market.
A further mechanism is complementarity. A professional caregiver may perform specialized tasks while the family continues supervision, scheduling and emotional support. Respite services can preserve the capacity of informal carers and improve quality even if recorded household care hours fall only slightly. Where formal services have poor continuity, an insured household may provide additional unpaid monitoring. The effect of formal coverage should therefore be measured across task composition, quality of care, caregiver health and employment, not simply by counting paid versus unpaid hours.
Risk Pooling, Insurance Affordability, and Benefit Design
Financial risk transferred by different insurance contracts
Long-term care protection can be delivered as expense reimbursement, a fixed daily or monthly cash indemnity, a service entitlement, or a combination. Each arrangement transfers a different portion of the uncertain loss. Expense reimbursement protects against eligible paid-service expenditure subject to caps and exclusions. A cash indemnity gives liquidity after a qualifying event and may finance paid help, adaptations or family support, but retains the problem of procuring care. An in-kind service entitlement promises access to specific assistance, although real performance depends on capacity and administrative delivery.
The American NAIC long-term care insurance overview describes benefit areas that may include home care, assisted living, respite and residential care. The contract's risk profile is shaped by waiting periods, daily maximums, elimination periods, benefit duration, inflation protection, lapses, morbidity, recoveries and claims adjudication. Private coverage may involve health underwriting and premiums that change over long horizons. For a gender-distribution analysis, it is necessary to observe who can buy the contract, who retains it, who meets the trigger and who can convert the payout into services.
A conventional actuarial identity expresses the expected present value of benefits as E[Σ v^t I_t B_t], where I_t indicates a covered payable state, B_t its amount and v^t the discount factor for payment time t. This is a textbook representation, not a newly proposed model. Full valuation additionally requires state-transition probabilities, mortality, recovery, lapses, inflation, expenses, premium income and dependencies among these variables. Insurer experience may reflect historical family care availability; changing household composition and employment can therefore change formal-care utilization even when biological dependency trends remain similar.
Sex-related differences in expected claim duration may arise in data because of longevity and care needs, but observed differences can also reflect marriage, household resources, institutional arrangements and product selection. Rate making and fairness involve separate questions. A statistical predictor of claims cannot by itself determine the ethical or legal appropriateness of charging by sex. Compulsory schemes may pool across groups, while voluntary schemes can experience risk selection. Estimating claims costs, identifying who pays and evaluating distributive rules are distinct stages of scientific analysis.
Coverage, affordability and the protection gap
A low premium does not necessarily imply affordable protection. It may purchase a narrow trigger, short duration or low benefit relative to prevailing service costs. For a person with irregular earnings, premium affordability at issue can deteriorate after job loss or periods of unpaid care. A policy lapse may erase protection at the age when dependency risk becomes more salient. A public scheme may be broadly accessible yet retain substantial co-payments or restrict benefits to severe dependency.
The OECD's affordability policy assessment estimates that without public support, formal long-term care can cost between about one and almost seven times the median income of older people, depending on care needs and country. Even limited weekly home care can exceed a large fraction of low-income older people's disposable resources. These are modeled comparisons using specified care packages and incomes; they are not observed individual insurance losses. An appropriate empirical analysis reports whether a ratio uses personal income, household disposable income or a national median, and whether costs are net of public transfers.
Women's higher representation among care recipients interacts with other circumstances. Older women may live alone more often in some cohorts and have lower pension income, increasing their dependence on public protection or family care. At working age, reduced earnings from caregiving can simultaneously make private long-term care insurance harder to purchase. Such a mismatch between expected need and purchasing power cannot be measured by claim frequency alone. A subsidy or compulsory pool might reduce it; the magnitude depends on eligibility, financing, access to services and whether protection persists.
Cash benefits, in-kind benefits and household bargaining
The payment instrument influences the distribution of control and labor. Cash gives the beneficiary flexibility but may leave most unpaid tasks unchanged, especially when paid caregivers are unavailable. In-kind services can directly substitute professional time for family labor, though the recipient may prefer a known caregiver and the service may be limited to defined tasks. Reimbursement sits between these forms: the household purchases qualifying services and obtains compensation, subject to liquidity, provider billing and claim terms.
The German statutory care-insurance system combines care grades with choices involving professional services and cash benefits that can support family care. It is structured as partial coverage; recipients may still pay substantial sums or rely on relatives. The distinction matters for economic evaluation. A benefit that encourages relatives to remain caregivers may improve recipients' household finances while reducing some carers' attachment to employment. A service benefit may support labor-force attachment but require an adequately staffed care workforce.
Household bargaining cannot be inferred from the beneficiary's name on an insurance check. A person receiving a cash allowance may pay a daughter for care, retain the money for other expenses or jointly manage it with relatives. Different arrangements have different consequences for caregiver autonomy, household consumption and future pension protection. To study them, administrative payments must be linked to household-level transfers and the allocation of unpaid hours. The mere existence of a cash benefit proves none of these distributions.
Coe, Goda, and Van Houtven (2015) investigate family effects of private coverage using instrumental variables. Their original working-paper analysis reported reduced informal care and greater adult-child full-time employment. Crucially, their later peer-reviewed 2023 journal article presents a more qualified result: it finds no evidence of reduced actual informal-care use over approximately eight years, while finding changed expectations about children's future care, less adult-child co-residence, and stronger labor-market attachment. The working paper and subsequent journal article must be identified by year and version rather than represented as one unchanged empirical finding. The 2023 article carries greater weight for the final interpretation. Instrument validity, the identified population, and generalization beyond U.S. private insurance remain separate issues.
Providers, workforce supply and real benefit effectiveness
A reimbursable service is useful only when the household can obtain it. Qualified caregiver supply, regional wages, transportation, waiting lists, quality regulation and rural accessibility shape the conversion of financial benefits into assistance. An insurer can honor a monetary limit while the family continues care because eligible workers are unavailable. A person who receives cash may have more flexibility but still face a local provider shortage. Practical adequacy depends on the care hours actually available, not simply a nominal benefit ceiling.
The European Commission's European Care Strategy identifies affordability, access, workforce quality and informal-carer support as connected policy objectives. EIGE's November 2026 long-term care brief relates service shortages to women's work and well-being. These policy documents demonstrate a recognized institutional agenda; they do not by themselves identify the causal effect of a specific rise in spending or service capacity.
A transparent purchasing-power calculation writes available professional hours as H = B/p, with B the cash benefit available for services and p the price per qualified hour. If p rises by 20% while B remains fixed, the maximum hypothetical hours fall to 1/1.2, or about 83.3%, of the former value. This is arithmetic, not a forecast of actual inflation in any jurisdiction. Actual hours also depend on minimum visit length, eligibility and administration, and whether care is available at the quoted price. For indexed contracts, the rate and period of benefit escalation must be matched to actual care-service cost growth.
Workforce shortages can also change long-run expenditure and insurance experience. A higher benefit may bid up wages when service supply is constrained; over time, predictable reimbursement may support training and workforce entry. The sign and magnitude are empirical questions. Reliable modeling requires local provider counts, pay rates, migration and training rules, claim utilization and geographic access. A published national increase in formal-care spending cannot alone establish that unpaid care hours fell.
Distributional Welfare Consequences and International Institutions
The economic incidence of a care episode
Long-term care risk is shared among households and collective financing institutions, and the same episode can impose costs on multiple people. A recipient experiences functional limitations, financial expenditure and potentially lost autonomy. A caregiver supplies time, may alter employment, can face health risks and may receive monetary transfers. Public agencies or insurers collect contributions, adjudicate eligibility and fund specified payments. The formal service provider allocates paid labor and capital. Descriptive expenditure accounts trace transfers between these actors; welfare analysis additionally asks what resources were actually used, what activities were displaced and what outcomes resulted.
A payment from an insurance fund to a recipient is a transfer. The financed professional hours, facility resources, coordination and claims administration represent real resource use. Family time is also a resource commitment even if its value is absent from formal-sector output accounts. Counting the full monetary transfer as a social resource gain and then separately adding the value of care purchased with it would count the same financing twice. Similarly, counting foregone wages and a complete replacement-cost valuation of identical care hours as independent resources can exaggerate economic cost. Distributional gains and losses should be reported alongside net resource costs but kept conceptually separate.
The insured person's welfare need not move with the caregiver's welfare. Formal care might improve a recipient's independence while giving a spouse time to rest or maintain employment. A recipient could prefer care from a relative, while the relative experiences an opportunity cost. Both preferences matter. There is therefore no defensible welfare interpretation based solely on a household's total income without also examining who controls resources, who supplies time and how care needs are met.
Across generations, care can connect the present earning capacity of middle-aged adults to the future dependency of elderly parents. If a daughter reduces contributions to retirement insurance while providing unpaid care, her own later income protection can deteriorate. This possibility creates a long-horizon exposure, but estimating its magnitude requires individual histories rather than one-year aggregates. Programs that credit caregiving periods for pension purposes address a different part of the exposure than programs financing immediate home services.
Germany: partial social insurance and family care
German compulsory long-term care insurance offers partial protection based on dependency assessments, with choices among professional benefits and family-oriented cash support. The German federal scheme explanation makes clear that its payments ordinarily do not cover every possible expense. Accordingly, a recipient classified as entitled may still require household money or unpaid work. A valid assessment of gender equity must measure the distribution of care grades, cash-versus-service selection, financial residuals and subsequent caregiver outcomes.
This system is relevant because it can support informal arrangements deliberately. Paying cash may be consistent with recipients' preferences and can give the household more resources. It may also change the incentive to reduce market work. Geyer and Korfhage's 1995 reform evaluation found a negative labor-supply response among men rather than women in its defined comparison. That result should be read in relation to the specific reform and baseline employment positions. It does not imply that men perform most unpaid care, or that women have negligible caregiving costs.
A modern evaluation would require longitudinal linkage of assessed grade, program choice, caregiver relationship, care hours, labor histories and earnings, with regional home-care workforce information. The propensity to choose a cash benefit may reflect family preference or the absence of an acceptable professional service. Treating benefit choice as random would confound these explanations. The rules described by a current government website also should not be projected backward uncritically to 1995 benefit schedules or later reform dates.
Japan: a reduction in services and household substitution
Japan provides a particularly informative contrast because the policy change studied by Miyawaki and colleagues affected the availability of formal services for groups differentiated by initial care-need severity. The authors' 2006 reform study reports reduced use of selected formal services and increased intensive care among affected co-resident family caregivers, with deteriorations in some measured caregiver health outcomes. The results show that a reduction in accessible professional care can be absorbed partly through increased unpaid time.
The studied effect is conditional. People living alone, households with paid domestic assistance and carers who do not share a residence are outside the principal sample. A less-dependent comparison group can differ from a more-dependent group in progression of health, household expectations and opportunities for prevention. Even after matching on observables, unmeasured shifts may affect the estimated differential changes. The reported evidence is stronger than a simple observation of correlation, but the causal meaning depends on the stated difference-in-differences conditions.
The Japanese case also cautions against equating a change in insurer spending with the entire cost of policy reform. Public expenditure may fall while care is shifted into unpaid households and health strain rises. Conversely, increasing formal service availability may have a nonbudgetary return in caregivers' time or employment that a narrow health-insurance accounting exercise omits. The effect must be valued through observed service use, family time, recipient outcomes, program financing and appropriate counterfactuals.
The Netherlands: intensive need and differentiated cost sharing
The Dutch Long-term Care Act (Wlz) addresses people with substantial continuing care or supervision needs. The Dutch Wlz program guidance describes the assessment of entitlement and the institutional organization of care, which may be delivered in an institution or under authorized home arrangements. Its population is selected by a specific need threshold, so direct comparisons with schemes that cover lower levels of dependency require careful severity adjustment.
The Netherlands also distinguishes entitlement from the beneficiary's financial contribution. Its 2026 care-cost contribution rules explain that personal circumstances and resources can influence cost-sharing liability. A scheme can provide a meaningful service entitlement while still exposing some households to substantial residual payments. Gendered incidence requires seeing who pays the contribution, what income remains, whether a spouse provides complementary care and whether the service is actually delivered.
Comparative analysis should resist ranking countries solely by spending per covered person. The Dutch Wlz's emphasis on high-intensity needs changes the composition of its recipient denominator. Higher average expenditure may reflect more severe eligibility rather than more generous benefits at an equivalent level of need. An international normalization would stratify by dependency intensity, care setting, price level and recipient characteristics.
Singapore: severe-dependency insurance as cash protection
Singapore's CareShield Life is a national insurance arrangement delivering cash once a specified severe-dependency trigger is met. The official CareShield Life overview describes lifetime payments while a qualifying severe disability persists and government support for premiums. An official claims assessment guide defines the inability to perform at least three of six specified daily activities. The program's assessment condition is narrower than the presence of any chronic condition or need for occasional assistance.
As of the 2026 guidance, the scheme cites a 2026 monthly payout of S$689 for the specified example and a schedule of increases until a qualifying stopping point. The CareShield payout information describes flexibility in the use of cash. The value of this form of benefit is financial: it can support paid services, family arrangements or other needs. The cash instrument alone does not tell us whether female relatives reduce care or increase their employment. Any claim about labor effects must rely on separate outcome observations.
The scheme also illustrates a measurement problem. A benefit requiring three or more ADL failures is not directly comparable with a graded dependency system covering moderate needs. A greater female prevalence of older-age disability does not translate automatically into a higher incidence of paid CareShield claims among enrollees, because claim onset depends on a severe threshold, cohort coverage and assessment. International comparison requires eligibility harmonization and calendar-year details before benefit adequacy is interpreted.
United States: Medicaid, private insurance and family exposure
In the United States, a significant part of publicly financed long-term support is arranged through Medicaid, with state-level differences in eligibility, financing and availability. The federal Medicaid LTSS overview describes support in institutional and home and community settings. The managed LTSS program guidance explains particular managed arrangements with varying state design. These institutional choices influence where a person may receive care, the extent of public protection and how provider networks determine actual access.
Private long-term care insurance adds another layer. Policyholders pay for contracts with defined triggers, benefit amounts, covered facilities, waiting periods and duration. The NAIC's consumer insurance guidance establishes broad categories but does not substitute for the text of an individual issued policy. Some privately insured people may be able to purchase formal services without relying on relatives; others may still face insufficient benefits or staffing limitations. People excluded from private coverage by price or underwriting may have greater need for public services.
The U.S. family-spillover research is particularly important because it examines adult children as well as the insured parent. The 2023 published study reports changes in care expectations, living arrangements and employment, while finding no evidence of reduced actual informal-care use over the studied horizon. The earlier 2015 working paper had suggested reduced informal care. This version difference is consequential for claims about service substitution; neither estimate can be generalized to Medicaid or all private insurance contracts. The United States is institutionally heterogeneous enough that broad national averages can conceal substantial differences among states and service markets.
The American Time Use Survey provides descriptive evidence about how widespread unpaid eldercare is, but does not indicate which recipients have a private policy, Medicaid-funded LTSS, or no formal coverage. Likewise, a private insurer's claims dataset may not record whether a claimant has one or several adult children providing care. To identify the intersection of insurance and caregiver labor, household surveys need administrative or validated self-reported benefit information as well as employment outcomes and detailed care tasks.
United Kingdom: caregiver-specific social protection
Direct support to carers is an important complement to recipient-centered insurance. The United Kingdom's Carer's Allowance eligibility guidance specifies a minimum of 35 hours of weekly care, a qualifying benefit for the care recipient, and employment-related conditions; relevant arrangements differ across parts of the UK, particularly Scotland. Such rules acknowledge unpaid work as a social-protection concern and can help support carers who supply intensive assistance.
A caregiver allowance is a different legal instrument from an insured care-recipient claim. Its entitlement follows the carer's circumstances and the status of the person receiving assistance. Thresholds may alter employment incentives or exclude some carers with substantial but differently structured commitments. The full distributional analysis should compare recipients, nonrecipients, hours, earnings, unmet care needs and transitions into or out of employment while taking account of the earnings conditions in force during the observed year.
The distinction between gender equality in entitlement and gender equality in outcome remains central. A formally gender-neutral benefit can be predominantly received by women if women are more likely to satisfy the care-hours condition. It can also have different welfare consequences across households with different capacities to hire substitute care. Evaluating only the legal text would miss those effects; evaluating only claimant counts would omit eligible nonclaimants and people caring outside the rule.
Europe: services and the paid care workforce
The European Care Strategy aims to improve accessible, affordable and high-quality services and to strengthen assistance for informal caregivers. The EIGE's 2026 investment policy brief identifies limited care provision and gendered consequences for employment and well-being. Such documents are authoritative evidence of institutional goals and analysis, while reported causal magnitudes must come from appropriately designed evaluations.
Gender inequality also occurs within paid care work, which frequently employs women and sometimes migrant workers. Replacing unpaid family labor with underpaid, insecure formal labor can move the burden rather than eliminate it. The ILO's care-work and employment study provides a wider context for pay, job quality and the institutional organization of care. A serious welfare evaluation considers the wages and conditions under which formal service capacity expands, alongside care-recipient quality and family outcomes.
Regional disparities matter. Nominally identical benefits can yield different effective access where rural travel times, workforce supply, provider quality and housing constraints vary. Price indexes alone cannot correct for absence of available workers. The relationship between market wages and family opportunity costs also differs between cities and low-income regions. A benefit design that reduces informal hours in one area may have little effect elsewhere unless service supply is expanded.
Joint distributions and the danger of average-only conclusions
Country comparisons should distinguish the distributions of needs, eligible claims, delivered care, family time and income. A woman living alone with severe dependency may have no available informal support and therefore face a higher risk of unmet need. A woman with several relatives nearby may obtain extensive care but place costs on those relatives' employment. A working-age man caring for a spouse may experience a sharp retirement response. A low-income family may obtain a legal benefit but lack the liquidity to pay fees before reimbursement. The average program payment cannot capture these distinct experiences.
The meaningful outcome vector includes need severity, assessed entitlement, formal hours delivered, care quality, unpaid hours, caregiver employment, caregiver health, household out-of-pocket spending and fund expenditures. This is a reporting set based on conventional welfare and program-evaluation practices, not a newly invented theoretical system. Where data are unavailable, the correct result is an identified evidence gap rather than an estimated gender distribution without a denominator.
Primary Source, Legal, and Model Validation
Evidence hierarchy and reproducibility
This monograph draws on four types of evidence: original survey and time-use statistics; official descriptions of benefit rules; comparative institutional reports; and peer-reviewed or institutional working-paper studies evaluating behavior under identifiable policy changes. Each answers a different research question. Official guidance documents establish how a government describes an entitlement at the access date. Survey releases establish observed rates under their own definitions. A quasi-experimental study estimates an effect under identification assumptions. A normative policy document sets goals or interprets evidence; it does not, by itself, estimate the causal impact of implementation.
The OECD's 2024 care affordability analysis models care needs and residual financial burdens across selected countries. The OECD's 2025 informal-carer indicator reports survey-based prevalence and sex composition. These two reports are related in subject but not interchangeable in denominator. The BLS uses a U.S. person sample and a diary method, while the EIGE CARE Survey uses a different age range and thematic scope. A coherent synthesis preserves these design details instead of averaging all published gender percentages.
A complete evidence ledger must specify each document's issuer, date, title, stable URL, verification scope and precise proposition it supports. The research retains this linkage in EVIDENCE.json. Where a publication is accessible only as an abstract, the supported proposition must be limited to what that abstract actually says. Independent follow-up studies should obtain original microdata and code where permitted; no replication of those microdata is claimed here. Statistical uncertainty, missingness, attrition and sampling weights remain part of source-specific interpretation.
Quantitative denominator audit
The BLS total of 38.192 million divided by 272.126 million produces 0.14035, or approximately 14.0%. Female caregiver prevalence is 20.990 divided by 139.137 million, approximately 15.1%. The male equivalent is 17.201 divided by 132.989 million, approximately 12.9%. Women's share among all caregivers is approximately 55.0%. These independently recomputable calculations reproduce the proportions in the original U.S. eldercare tables. The difference between the separately rounded male-plus-female count and the rounded total is 0.001 million; it is a display artifact rather than evidence of another category.
The OECD 61% women-carer statistic is a descriptive distribution across 26 countries, not a 61% individual probability of being a caregiver. Its associated 13% informal-care prevalence refers to people aged 50 or more in the reported source. The ILO 76.2% figure is the share of unpaid care-work hours performed by women in the research's broad care universe. The OECD two-thirds female share among older people requiring care refers to the recipients of support, not carers. These four claims should never appear on one chart with a single unnamed denominator.
The EIGE survey's sampling age range of 16–74 also affects interpretation. Older spouses above 74 may supply significant care but be absent from this target population, while OECD 50-plus measures include many of them. Comparisons should be presented as complementary observations rather than changes over time in one global series. If microdata were later harmonized, the researcher would have to report resulting sample loss and exclusion bias, especially for older spouses and younger disabled recipients.
Causal interpretation of policy reforms
The causal question for an increase in service coverage is how the same household's informal-care hours would differ under the revised arrangement compared with a feasible alternative. With conventional potential outcomes Y(1) and Y(0), the average effect E[Y(1) − Y(0)] is defined for an identified population. It cannot be measured by observing households under only one treatment condition and comparing them without further assumptions.
Difference-in-differences uses before-and-after changes for exposed and comparison groups. For a causal estimate it requires the untreated trend to be sufficiently comparable across groups, stable composition or corrections, coherent treatment timing, and attention to policy spillovers. Japan's service-reduction evaluation provides a studied treatment based on differential impact among need categories. However, if those groups had different preexisting trends in severity or family capacity, the estimated contrast could combine the reform with other changes. The reported association should remain tied to its design.
Germany's long-term care reform study also uses a policy transition to identify labor adjustments. The striking male response shows that an effect estimate describes behavioral adjustment at the relevant margin rather than the unconditional distribution of caregiving duties. Family composition, older workers' retirement options, and access to cash benefits all shape where the adjustment appears. Cross-country comparisons should examine these mediators rather than classify one result as anomalous because it departs from the female-majority descriptive data.
Coe, Goda, and Van Houtven's 2023 peer-reviewed article employs an instrument based on state tax treatment of insurance to address selection into coverage. Unlike their earlier 2015 working-paper estimate, its abstract reports no detected decline in informal care use while documenting changes in children's perceived care willingness, co-residence and labor-market attachment. Interpretation requires instrument relevance, a defensible exclusion restriction and careful identification of the affected households. Even with valid assumptions, the estimates do not automatically describe compulsory insurance or Medicaid.
The parental-health analysis by Barszczewski and colleagues is valuable for identifying transitions in family need, but parental deterioration can affect adult children directly through anxiety, travel, income transfers and family decisions independently of hours spent caregiving. The SHARE-based economic cost research should be interpreted in the context of its identification strategy, country grouping and modeling choices. Its modeled output effect is not a direct national-account observation of the precise GDP loss attributed to caregiving.
Confounding, selection and measurement error
Several alternative explanations require systematic attention. Wealth and education affect both the probability of purchasing private insurance and the ability to pay for formal care. Dependency severity affects both informal hours and the probability of receiving a formal claim. A strong local provider market may encourage formal-care use while also influencing public investment. Caregivers' health can influence whether they accept caregiving duties, while care duties may in turn influence their health. These bidirectional relationships create selection and reverse causality in simple correlations.
A credible empirical study would record pre-care employment, income, health, household relationship, number of potential carers, recipient impairment, geographic location, care-service access, program eligibility and contemporaneous policy changes. Attrition should be assessed because heavily burdened carers may drop out of longitudinal surveys. Researchers should distinguish an intermediate mediator, such as formal service use, from a pre-treatment confounder. Conditioning on care utilization when estimating the total effect of insurance may remove part of the effect of interest.
Care hours are particularly sensitive to measurement. Passive supervision can coincide with domestic activity, generating overlapping diary time. Care to multiple recipients can be allocated imprecisely. Survey questions may emphasize personal care while omitting administration, transport, cognitive support and the obligation to remain available. An analysis can be made robust by reporting several operational thresholds, including frequency, weekly hours, personal-care tasks and supervision measures. It must also avoid counting the same hour twice when multiple activities occur simultaneously.
The distinction between unmet need and covered expenditure is equally important. Low observed spending could signify effective prevention, strong unpaid care, limited access, insufficient household income or denied claims. High spending could reflect severe dependency rather than better access. A valid comparative model needs functional assessments and care quality, not financial measures alone. Claims acceptance rates may be influenced by insurer case mix and contract triggers, while recipient experience can be affected by provider continuity after a claim is approved.
Legal versions and contract integrity
Statutory rules change, as do private policy forms and claims procedures. The Netherlands' 2026 cost-sharing guidance cannot be assigned retrospectively to a 2010 Dutch care episode. Singapore's 2026 payout figures and cohort-specific coverage rules must be tied to the actual claim period. UK Carer's Allowance earnings conditions should be checked at the decision date. A historical study of Germany's 1995 reform must use the legal rules effective in 1995 rather than current benefit tables.
Private insurance requires particular attention to policy wording. The NAIC's long-term care topic guidance describes the product category, yet does not establish what any particular insurer must pay under a specific issued policy. An individual-level claims study would need the complete specimen policy, dated endorsements, eligibility assessment form, elimination-period terms, inflation options, lapse history, claims documents and relevant jurisdictional law. This monograph makes comparative institutional claims rather than individual contractual determinations.
International legal comparison should also avoid treating “social insurance,” “Medicaid,” “cash allowance” and “private policy” as interchangeable legal forms. Their beneficiaries, financing institutions, judicial remedies and administrative discretion can differ. Claims for high-intensity residential support under the Dutch Wlz are not necessarily comparable with moderate home support or a general caregiver allowance. The proper comparative unit is a defined need-and-entitlement scenario, followed by actual delivery and residual household exposure.
Actuarial and economic validation
Insurance pricing requires expected frequency, severity and duration of covered dependency states. A longitudinal care model must consider entry into dependency, recovery, progression, death, benefit cessation, policy lapse, inflation and possible changes in family support. Gender is a possible observed covariate, but its interpretation is constrained by cohort, social conditions, policy selection and legal rules about classification. Care workforce shortages may alter paid-service utilization without changing the underlying medical progression of dependency.
Household welfare valuation also requires an explicit accounting choice. Replacement-cost estimates, lost earnings and subjective welfare assessments can each be useful, but they answer different questions. Lost wages should not be automatically added to full replacement cost for the same hours, and the face value of insurance cash cannot be counted as newly produced services. A complete evaluation would report resource cost, financial transfer and distributional incidence in separate columns. No original actuarial transition matrix or econometric coefficient is calculated here.
Caregiver time and insurance protection may be complements rather than simple substitutes. Professional clinical care can stabilize a recipient and allow family care to be provided safely; day care can release hours for employment; respite can maintain caregiver well-being; a cash benefit may finance a combination. The choice of a single substitutability coefficient without task-level evidence could introduce severe model risk. Scenario analysis should represent multiple possible behavioral responses and disclose how results vary with household composition and local provider supply.
Research limits and evidence governance
The strongest evidence in this manuscript concerns the descriptive concentration of unpaid care among women, the distribution of care need among older women, the heterogeneity of statutory benefit forms, and observable responses to selected reforms. The evidence is less complete for the lifetime causal effect of insurance on gender pension gaps, the welfare value of household bargaining, and program effects in low-income countries where formal service capacity and reliable insurance data are sparse. These limitations reflect the available data and the research's scope, not a claim that these effects are absent.
This synthesis contains no original household microdata, no new econometric estimation, and no systematic meta-analytic effect size. The European analysis cited is based on SHARE data and a published article; the Japanese and German estimates are attributed to their respective researchers. The private LTC spillover interpretation compares the 2015 NBER working paper with the revised 2023 peer-reviewed journal record. Secondary interpretations are clearly distinguished from the directly supported claims. No new authorial theory or universal premium doctrine is advanced.
Integrated Scientific Results and Further Research
Integrated findings
The evidence establishes that gender inequality in long-term care encompasses two connected population distributions: who experiences dependency and who supplies care. Women comprise approximately two-thirds of older people with care needs in the OECD comparison and majorities of carers in the other surveyed populations. These are different quantities and should be tracked independently. The overlap of later-life need and middle-age care labor motivates a longitudinal research agenda concerning earnings, retirement and eventual protection.
Formal care access is a distinct insurance outcome. A benefit's financial adequacy must be measured along with the service hours that can actually be obtained. Cash, reimbursement and in-kind arrangements differ in the risks they transfer, the recipient's control and their likely effects on family behavior. A universal prediction that more coverage necessarily reduces unpaid care does not follow from the evidence. Japan's service-cut experience indicates substitution in a specified setting, while Germany's cash-oriented reform demonstrates a labor incentive that appears on a different gender and employment margin.
Family spillovers can be material even when insurer payment is legally directed to an older person. The later U.S. private-insurance study identifies changes in adult children's labor-market attachment and residential arrangements, while reporting no detected reduction in actual informal-care use over its studied horizon. More broadly, the caregiver's time, health and earnings are economically relevant outcomes of protection. These effects require separate measurement; their social value cannot be read from insurer claims ratios or the beneficiary's expenditure alone.
Gender fairness also depends on financing and affordability. Women can face higher later-life need yet limited ability to buy voluntary coverage because of earnings interruptions. Compulsory insurance may redistribute these risks but can retain co-payments, eligibility exclusions or geographic service disparities. Private policies may transfer specified expense risk while underwriting and affordability restrict access. The relevant evaluation includes who contributes, who enters coverage, who qualifies, who can obtain services and who supplies residual unpaid assistance.
The conjunction of demographic aging and uncertain formal-care capacity creates model risk for actuarial and fiscal projections. Historical experience obtained in families with ample unpaid support may fail to describe future cohorts with smaller households, different employment patterns or fewer nearby relatives. Dependency incidence and paid-service utilization must be projected separately, with variation in care prices and provider supply. These conclusions are derived from conventional actuarial and economic reasoning applied to the reviewed evidence.
Research priorities and feasible identification
A priority empirical question is whether an exogenous expansion of reimbursable home-care hours changes the employment of daughters, sons, spouses and other carers differently. The treatment could be a policy threshold, local rollout or changes in reimbursement, but a credible design would check pre-trends, enrollment, provider capacity and family composition. The outcomes should include both participation and hours worked, with separate estimation for already employed and previously nonemployed people.
A second study should compare the effects of cash and in-kind benefits at similar levels of functional need. Observed choice between these options is not random; preferences, local wages, service scarcity and household liquidity select households into different packages. Identification may require rule changes, eligibility discontinuities or within-person transitions. Both caregiver and recipient preferences need to be observed to distinguish income support from genuine service substitution.
A third research line should link caregiving spells to pension accumulation over time. Data would need actual contribution records, durations of care, formal employment histories, retirement age, applicable pension credits and partner income. Gender comparisons must account for secular wage gaps and labor-market changes. A one-time retrospective question about caregiving is an inadequate basis for a lifetime causal pension estimate.
Geographic service supply offers a further test. Researchers could compare the real value of otherwise similar benefits across areas with different trained-careworker capacity, travel times and reimbursement rates, controlling for selection into location. A change in nominal entitlement might have a greater effect in regions where providers can expand services. This would identify the circumstances in which workforce investment and insurance benefit expansion are economic complements.
Finally, international work should incorporate people routinely missed by older-age survey definitions: younger adults with severe disability, recipients with cognitive supervision needs, informal carers above survey age limits, migrants in formal care occupations, and people without close family networks. Inclusion requires appropriate ethical data governance, disability-sensitive measures and transparent denominators. Comparative evidence should be extended beyond wealthy countries only when reliable definitions and source provenance permit defensible claims.
Scientific and editorial disposition
This research provides a complete comparative interpretation of the mechanisms connecting long-term care insurance, gender inequality and unpaid labor. It identifies where official original-source statistics are descriptive, where program texts specify legal entitlements, and where research offers conditional causal evidence. Its conclusions have clear limits: no global pooled effect estimate, no new premium pricing model, no claim of original econometric identification and no newly established doctrine. The scientific contribution lies in accurate integration, interpretation and research design.
The central analytical standard is to measure insured financial protection, realized formal care, residual informal time and caregiver opportunity costs together, with consistent time frames and denominators. When some of these outcomes are unobserved, the evaluation should identify precisely what is missing. This approach creates a more informative description of care insurance than benefit totals alone and follows from the established principles of welfare economics, household production, program evaluation and actuarial science.
Conclusion
Long-term care insurance finances uncertain dependency within an economy of care that includes paid providers and unpaid households. Insurance contracts and public entitlements determine formal contributions and payments, while care need, provider access, household decisions and labor-market conditions determine much of the actual allocation of assistance. A scientific assessment of gender inequality must examine each of these stages and their distribution across people.
The reviewed OECD data establish that women comprise a large majority of older persons with long-term care needs, while OECD, ILO, EIGE and U.S. survey sources show large gender asymmetries in informal work under different definitions. Source-specific denominators are indispensable: share of carers, share of hours and share of people requiring care are separate indicators. The research on Japan, Germany, adult children in the United States and middle-aged women across Europe shows that coverage and services can affect family behavior, labor supply and well-being, but in reform- and population-specific ways.
Insurance benefits are economically valuable through both money and the capacity to obtain appropriate care. Their social consequences include protection against out-of-pocket expenditure, substitution or support of unpaid labor, and changes in caregiver health, earnings and future pensions. Cash, in-kind services and direct caregiver allowances yield different distributions. The payer's financial cost is one component of these outcomes; the household's time and constrained choices require explicit accounting.
The concluding scientific judgment is that the effectiveness and distributional fairness of long-term care insurance can be evaluated only by connecting legal entitlement to actual care delivery and to the burdens carried by those who provide support. This is an evidence-grounded analytical conclusion within existing insurance and care economics. It supplies a basis for further causal research and actuarial sensitivity analysis without asserting a new theoretical system.
References
Barszczewski, Jacek, Ricarda Milstein, Jinkook Lee, and Ana Llena-Nozal. 2025. “How Do Women Respond to Increased Care Needs of Their Parents? The Economic Costs of Informal Caregiving.” The Journal of the Economics of Ageing 32: 100589. Published online October 25, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC13344410/
Board of Governors of the Federal Reserve System. 2025. “Care Work and Living Arrangements.” Economic Well-Being of U.S. Households in 2024. Washington, DC: Federal Reserve. https://www.federalreserve.gov/publications/2025-economic-well-being-of-us-households-in-2024-care-work-and-living-arrangements.htm
Central Provident Fund Board (Singapore). 2026. “What Is CareShield Life?” Official program guidance. https://www.cpf.gov.sg/service/article/what-is-careshield-life
Central Provident Fund Board (Singapore). 2026. “When Am I Eligible to Make a CareShield Life/ElderShield Claim?” Official claims guidance. https://www.cpf.gov.sg/service/article/when-am-i-eligible-to-make-a-careshield-life-eldershield-claim
Central Provident Fund Board (Singapore). 2026. “CareShield Life.” Official program and payout information. https://www.cpf.gov.sg/member/healthcare-financing/careshield-life
Centers for Medicare & Medicaid Services (CMS). 2026. “Long-Term Services and Supports.” Medicaid.gov, federal program guidance. https://www.medicaid.gov/medicaid/long-term-services-supports
Centers for Medicare & Medicaid Services (CMS). 2026. “Managed Long-Term Services and Supports.” Medicaid.gov, federal program guidance. https://www.medicaid.gov/medicaid/managed-care/managed-long-term-services-and-supports
Coe, Norma B., Gopi Shah Goda, and Courtney Harold Van Houtven. 2015. “Family Spillovers of Long-Term Care Insurance.” NBER Working Paper 21483. Cambridge, MA: National Bureau of Economic Research. https://www.nber.org/papers/w21483
Coe, Norma B., Gopi Shah Goda, and Courtney Harold Van Houtven. 2023. “Family Spillovers and Long-term Care Insurance.” Journal of Health Economics 90: 102781. DOI: 10.1016/j.jhealeco.2023.102781. https://pmc.ncbi.nlm.nih.gov/articles/PMC10533212/
European Commission. 2022. “A European Care Strategy for Caregivers and Care Receivers.” Employment, Social Affairs and Inclusion, September 7. https://employment-social-affairs.ec.europa.eu/news/european-care-strategy-caregivers-and-care-receivers-2022-09-07_en
European Institute for Gender Equality (EIGE). 2026. Sharing Care, Closing Gender Gaps: CARE Survey 2024. Luxembourg: Publications Office of the European Union. https://eige.europa.eu/publications-resources/publications/sharing-care-closing-gender-gaps-care-survey-2024
European Institute for Gender Equality (EIGE). 2026. Investing in Long-Term Care for a More Competitive and Inclusive EU. Luxembourg: Publications Office of the European Union. Published November 11. https://eige.europa.eu/publications-resources/publications/investing-long-term-care-more-competitive-and-inclusive-eu
Geyer, Johannes, and Thorben Korfhage. 2018. “Labor Supply Effects of Long-Term Care Reform in Germany.” Health Economics 27: 1328–1339. https://pubmed.ncbi.nlm.nih.gov/29700888/
German Federal Ministry of Health (Bundesministerium für Gesundheit). 2025. “Long-Term Care Insurance.” Federal official guidance on compulsory partial-benefit financing and care grades. Updated February 25, 2025; accessed November 14, 2026. https://www.bundesgesundheitsministerium.de/en/themen/pflege/online-ratgeber-pflege/long-term-care-insurance/page
Government of the Netherlands. 2026. “Long-Term Care Act (Wlz).” Official long-term care program guidance. https://www.government.nl/themes/family-health-and-care/nursing-homes-and-residential-care/long-term-care-act-wlz
Government of the Netherlands. 2026. “I Have a Wlz Care Needs Assessment. Do I Have to Pay Towards the Costs of My Care?” Official 2026 contribution guidance. https://www.government.nl/documents/question-and-answer/i-have-a-wlz-care-needs-assessment.-do-i-have-to-pay-towards-the-costs-of-my-care
Government of the United Kingdom. 2026. “Carer's Allowance: Eligibility.” GOV.UK, current entitlement guidance. https://www.gov.uk/carers-allowance/eligibility
International Labour Organization (ILO). 2018. Care Work and Care Jobs for the Future of Decent Work. Geneva: International Labour Office. https://www.ilo.org/publications/major-publications/care-work-and-care-jobs-future-decent-work
Miyawaki, Atsushi, Yasuki Kobayashi, Haruko Noguchi, Taeko Watanabe, Hideto Takahashi, and Nanako Tamiya. 2020. “Effect of Reduced Formal Care Availability on Formal/Informal Care Patterns and Caregiver Health: A Quasi-Experimental Study Using the Japanese Long-Term Care Insurance Reform.” BMC Geriatrics 20: 207. https://link.springer.com/article/10.1186/s12877-020-01588-7
National Association of Insurance Commissioners (NAIC). 2026. “Long-Term Care Insurance.” Insurance Topics, supervisory and consumer guidance. https://content.naic.org/insurance-topics/long-term-care-insurance
Organisation for Economic Co-operation and Development (OECD). 2024. Is Care Affordable for Older People? OECD Health Policy Studies. Paris: OECD Publishing. https://www.oecd.org/en/publications/is-care-affordable-for-older-people_450ea778-en/full-report.html
Organisation for Economic Co-operation and Development (OECD). 2024. “Growing Long-Term Care Needs Across OECD Countries.” In Is Care Affordable for Older People? Paris: OECD Publishing. https://www.oecd.org/en/publications/is-care-affordable-for-older-people_450ea778-en/full-report/component-5.html
Organisation for Economic Co-operation and Development (OECD). 2024. “Assessment and Policy Solutions for Affordable Long-Term Care.” In Is Care Affordable for Older People? Paris: OECD Publishing. https://www.oecd.org/en/publications/is-care-affordable-for-older-people_450ea778-en/full-report/component-4.html
Organisation for Economic Co-operation and Development (OECD). 2025. “Informal Carers.” In Health at a Glance 2025: OECD Indicators. Paris: OECD Publishing. https://www.oecd.org/en/publications/health-at-a-glance-2025_8f9e3f98-en/full-report/informal-carers_6f243c4d.html
U.S. Bureau of Labor Statistics (BLS). 2025. “Unpaid Eldercare in the United States—2023–2024.” American Time Use Survey release USDL-25-1462, September 25. https://www.bls.gov/news.release/elcare.htm
U.S. Bureau of Labor Statistics (BLS). 2025. “Table 1. Number and Percent of the U.S. Population Who Were Eldercare Providers by Sex and Selected Characteristics, Averages for 2023–2024.” American Time Use Survey. https://www.bls.gov/news.release/elcare.t01.htm
World Health Organization (WHO). 2001. International Classification of Functioning, Disability and Health (ICF). Geneva: WHO; living classification page. https://www.who.int/classifications/international-classification-of-functioning-disability-and-health