RESEARCH DOSSIER
Home and Community-Based Care Insurance Benefit Design
Abstract. Home and community-based care (HCBC) insurance concerns the financing of sustained personal assistance and support services where people live, including services delivered in private homes and organized community settings. Benefit design converts functional dependency and care needs into covered service categories, entitlement conditions, allowable providers, payment obligations, and eventual patterns of residence. This monograph evaluates the legal, actuarial, economic, and service-system mechanisms that govern those conversions. It distinguishes four often-conflated quantities: functional need, qualifying insurance eligibility, authorized covered care, and delivered paid care. Using primary public-law and supervisory materials, official international comparisons, and selected empirical studies, it examines private long-term care insurance, U.S. Medicaid home- and community-based services, German social long-term care insurance, Singapore's cash-based CareShield Life, and Australia's Support at Home arrangements. The methodological contribution is an integrated source-critical comparison of benefit architecture, state transitions, and budget exposure, accompanied by explicitly hypothetical payment and sensitivity illustrations. Coverage generosity affects financial exposure through service prices, provider availability, utilization, family labor, cost sharing, and continuity rules, so headline benefit limits alone cannot measure protection. Empirical evidence associates some community-service expansions with improved community residence, while causal identification and transfers across national systems require caution. The analysis derives actionable contract-valuation and research implications without claiming a new theory, a newly estimated causal effect, or an exhaustive international legal survey.
Keywords: Home and community-based services; long-term care insurance; personal assistance; activities of daily living; severe cognitive impairment; benefit eligibility; indemnity benefits; expense reimbursement; Medicaid; social long-term care insurance; care continuity; informal caregivers; community living; multistate actuarial models; insurance affordability; disability rights; comparative insurance law.
Introduction
Long-term care is an extended response to limitations in everyday functioning. Its financing distributes spending over time and across households, employers, insurers, taxpayers, and unpaid caregivers. Within that larger field, home and community-based care has a distinctive insurance problem: the resources that sustain residence in a community are heterogeneous, repeatedly delivered, sensitive to the living environment, and often produced jointly by paid workers and family members. A single benefit limit may cover several different kinds of assistance, yet it provides little information about whether eligible care will be available when needed.
The WHO classification of functioning establishes an appropriate point of departure: disability and functioning are understood through the interaction of health conditions, activities, participation, and environmental factors. Functional limitation is relevant to insurance, but an insurance policy adds contractual triggers, evidence requirements, covered services, exclusions, monetary constraints, and claims administration. A public program adds eligibility law, public funding rules, provider regulation, and, sometimes, enrollment or service capacity constraints. A person may therefore have substantial care needs while receiving a smaller covered package, and may receive an authorization without obtaining every scheduled hour of actual assistance.
The human-rights dimension is equally concrete. The Convention's Article 19 recognizes equal rights to living independently and being included in the community, including access to in-home, residential, and other community support services. This provision informs the analysis of individual choice and the social consequences of care financing. It does not itself specify a uniform insurance contract, reimbursement rate, or national entitlement to a particular hour of service. The distinction between the legal right, its implementing institutions, and a specific insured promise matters throughout.
The study asks how home-based services become qualifying benefits; how cash, reimbursement, provider-based, and blended designs distribute cost and risk; and under what conditions a formal benefit preserves the possibility of continued community residence. The comparative scope includes private insurance and selected publicly funded or compulsory programs, with the latter examined as institutional comparators rather than mislabeled private contracts. The primary evidence consists of official program rules, legal texts, statutory descriptions, supervisory guides, and international statistical and policy analyses. Selected research literature supplies theoretical mechanisms and empirical findings. Examples and numerical illustrations are labeled hypothetical and do not estimate real-world claim frequencies or recommend a particular product.
A central organizing distinction is between a covered service and an achieved outcome. An insurer may promise reimbursement for home health aides, respite, adult day services, or supportive technologies; a policyholder may seek stable living arrangements, autonomy, and reduced family strain. Spending rules can be stated precisely. Achieved outcomes depend on service capacity, need trajectories, choices, housing accessibility, the health system, and the counterfactual alternative. Scientific evaluation must trace the relationship without assuming that every paid service prevents institutional placement.
Conceptual and Historical Foundations
What the insured object includes
Home and community-based care denotes a set of supports furnished in or around ordinary living arrangements to address sustained needs for assistance. It can include help with bathing, dressing, transferring, toileting, continence, and eating; supervision associated with cognitive impairment; instrumental help with meals, shopping, and household tasks; adult day care; respite; personal care; assistive equipment; transportation; home modifications; and care coordination. Not every arrangement includes every item. The Medicaid 1915(c) authority illustrates a recognized range of community services, while the actual approved waiver and person-centered service plan determine whether a given participant receives a particular service.
Three boundaries require explicit treatment. First, medical treatment and ongoing custodial or personal assistance overlap operationally but can be financed through different rules. Second, care at home is a setting, while community-based care may be delivered at a day center or within supported community housing. Third, long-term care need is a functional and temporal condition; eligibility for a payment can require a different test and may be triggered only when a specified number of activities of daily living are impaired or substantial supervision is necessary.
The distinction is visible in Medicare. The program's long-term care coverage explanation states that Medicare generally does not pay for long-term custodial care when that is the only care required. Medicare may finance other medically necessary home health services under their own rules. The insured object in a comprehensive long-term care contract thus cannot be inferred from a general health coverage label.
Care need, eligible claim, and paid service
Four quantities must be kept analytically separate. Need is an assessed requirement for support given a person's functioning, health, environment, and goals. Eligibility is satisfaction of contractual or statutory conditions. Authorization identifies approved types, quantities, durations, and providers. Delivery is actual care furnished, potentially followed by a payable claim. The transitions from need to eligibility, eligibility to authorization, and authorization to delivery can each involve delay, discretion, shortages, or disagreement.
In U.S. tax-qualified long-term care insurance, Internal Revenue Code section 7702B supplies conditions for qualified long-term care services and chronically ill status, including substantial assistance with at least two specified activities of daily living expected for at least 90 days or substantial supervision due to severe cognitive impairment, subject to prescribed certification and plan requirements. This federal tax definition does not turn every public home-care program into a tax-qualified policy. It also does not mean that any observed limitation automatically establishes a specific insurer's liability. Contract language, qualified service requirements, and applicable law still control.
An observed payment is the intersection of several conditions. Let N represent need, E a satisfied benefit trigger, A authorization, D actual delivery, and C compliance with covered-service conditions. A reimbursement payment can occur only for qualifying delivered expenses, conceptually E ∩ A ∩ D ∩ C, subject to limits and claims procedures. In an indemnity or disability-style design, a contractual payment may depend on E and specified duration or amount even when the beneficiary's exact spending is not submitted. This distinction explains why cash payment totals and hours of formal care cannot be interpreted as interchangeable utilization measures.
Historical and institutional pathways
The emergence of home and community-based care financing reflects overlapping trajectories: the growth of private long-term care insurance, social insurance for care dependency, the expansion of publicly financed community services, and disability-rights commitments emphasizing the ability to choose one's living arrangements. Different jurisdictions institutionalized those trajectories through different legal instruments.
U.S. Medicaid expanded opportunities for community services through multiple statutory authorities, including the 1915(c) waiver framework. Those waivers permit states to provide specified services to eligible populations who would otherwise require an institutional level of care, within federal approval and financial conditions. This produced service-specific and state-specific variation rather than a single national personal-care contract.
Germany introduced compulsory social long-term care insurance in the 1990s and provides structured benefits based on assessed care grades, including cash, professional home services, and combinations. The federal long-term care overview identifies this as a partial-benefit system. Care arrangements and household contributions must be assessed alongside statutory amounts.
Singapore's CareShield Life program takes a contrasting path: eligible severe disability gives access to ongoing cash payouts rather than reimbursement of each home-care invoice. Its three-activity-of-daily-living threshold and lifetime payout structure illustrate how an impairment-defined contingency can finance flexible assistance without establishing that the insured payment purchases a complete formal-care package.
Australia's Support at Home program, implemented from November 2025, provides another comparison with assessment-linked categories and participant contributions. Its service-contribution rules differentiate classes of support and applicable out-of-pocket contributions. Programmatic funding and assessed public-service access must again be kept distinct from the terms of a private insurer's indemnity or reimbursement contract.
These paths reveal a substantive economic distinction. Some schemes pool the risk of a chronic functional state and pay money; others finance a specified service basket; still others limit the insured event through reimbursement ceilings. None can be compared solely by their nominal advertised monthly amounts.
Research design and source hierarchy
This monograph applies documentary comparative analysis to verified primary institutional descriptions, selected statutory provisions, supervisory guidance, independent international reports, and peer-reviewed research. Claims of law are attributed to the jurisdiction and instrument that establish them. Where a program's own public explanation supplies a benefit figure, it is treated as an official description at the checked date, with the operative legal instruments remaining controlling. Model laws are identified as models rather than automatic law. Academic studies are read within their population, estimator, and observation window.
The OECD care-affordability study provides cross-country analysis of long-term care costs and household exposure, including the role of country-specific institutional arrangements. It demonstrates why benefit adequacy must be measured against needs-adjusted care expenditures and available household resources. The WHO integrated-care package contributes a services-and-system perspective, but it is not itself a validated universal insurance benefit formula.
The scientific procedure is to identify each claim's source, legal or empirical scope, effective period, and denominator; specify the comparison being made; and distinguish observed evidence from inference. The research uses conceptual accounting identities and hypothetical financial scenarios to reveal tradeoffs. It does not estimate new claims experience or attribute observed international differences to a single policy feature. Full source-to-claim records accompany the manuscript in EVIDENCE.json.
Home-Based Qualifying Services
Functional assessment and eligibility
A home-care benefit needs an operational definition of the state that makes assistance necessary. Assessment commonly combines reported daily functioning, a structured evaluation, clinical and cognitive evidence, and an account of assistance available in the person's environment. Activities of daily living measure basic self-care. Instrumental activities of daily living, such as preparing food, using transportation, managing medication, and shopping, help explain dependence in community life but may have different legal weight. The WHO ICF framework makes clear that functional performance and participation depend partly on the context in which an individual acts. This matters when assessment is performed under controlled conditions rather than in the person's actual residence.
Insurance assessment introduces a second layer. A long-term care policy can use a specified inability to perform activities of daily living, supervision for cognitive impairment, or particular certification criteria. An assessor must determine which activities count, what degree and frequency of assistance qualify, whether aids and adaptations are considered, who may certify, and whether the impairment must persist for a specified period. The federal tax-qualified definition requires at least two qualifying activities for an expected 90-day period or severe cognitive impairment involving substantial supervision, with additional certification and plan-of-care conditions. Its function is to define a category of federally tax-qualified care, not to replace every applicable state insurance rule or a product's full coverage terms.
Three measurement errors are important. Classification error arises when different assessors evaluate the same functional capacity differently. Temporal error arises when episodic or progressive needs are measured during an unusually good or bad day. Setting error arises when a person is assessed in an environment with assistance or equipment not available at home. These errors affect eligibility, claim incidence, benefit duration, and the difference between authorized and delivered care. Reviews of eligibility decisions should examine inter-assessor agreement, repeated observations, and administrative appeals, not simply whether a threshold exists in a policy.
Cognition has an independent influence. A person may walk and feed themself but require supervision to avoid dangers associated with severe impairment. Long-term care products that treat supervision as a separate qualifying path recognize this pattern. Nevertheless, cognitive diagnosis, supervision intensity, evidence of unsafe behavior, and the contract's definition of severe impairment must be related carefully. A diagnostic label alone does not establish every monetary entitlement.
Service taxonomy and qualifying-provider rules
A useful research taxonomy divides services into personal assistance, skilled and nonskilled health support, practical household services, participation and respite, assistive technologies and home modifications, and navigation or coordination. The classifications intersect in practice. A worker may assist with bathing, transfer, medication reminders, and meals during one visit. Whether the claim is payable may depend on licensing, professional status, permitted tasks, separate invoices, approved care plans, and the policy's distinction between medical services and ordinary domestic work.
The Medicaid waiver service framework permits a wide variety of supports, but the precise state waiver identifies the approved population, covered categories, settings, utilization controls, and operational obligations. An insurance analyst should avoid assuming a federally listed service is necessarily available to every Medicaid participant. Waiver service labels are enabling categories, and actual access is mediated by state choices and participant-level plans.
Private long-term care insurance adds several coverage distinctions. Some contracts cover only services from approved agencies or qualified licensed providers; others offer more flexible eligibility and home-health or informal-care riders. Some reimburse eligible incurred charges up to a daily or monthly ceiling. Some promise a specified cash amount once an eligible claim is established. The NAIC consumer overview distinguishes different long-term care financing arrangements and underscores the importance of policy terms. It is explanatory institutional guidance, not proof that all contracts cover particular relatives, alternative therapies, or supplies.
Provider rules serve several functions. They can support verification of delivered work, minimum competence, claims auditing, and protection from abuse. They also influence provider supply and price. A highly restrictive provider definition can make a contractual benefit difficult to use in rural markets or during labor shortages. A permissive rule can improve flexibility but require different verification and safeguarding mechanisms. The relevant optimization problem involves both the expected value of eligible care and the probability that a qualified provider can actually be engaged.
An effective service schedule therefore needs explicit answers about service categories; occupational qualifications; permitted relatives or friends; geographic coverage; telehealth or remote monitoring; hospitalization and temporary facility stays; hours per day and days per month; transportation; and what documentation must accompany a reimbursement claim. Those elements should be evaluated as interacting contract terms. Broad categories with strict invoices can deliver less effective assistance than a narrower benefit with a usable provider network.
Public program authorities and individual plans
U.S. Medicaid illustrates why service entitlement must be traced to a particular legal authority. Under section 1915(c) waivers, states may obtain federal approval to provide alternatives to institutional services for eligible individuals who meet the relevant level-of-care requirements. Waivers have defined target groups, eligibility rules, service packages, cost-neutrality requirements, and operational conditions. Depending on the program, enrollment limits and waiting lists may matter. A prospective claimant's apparent functional suitability does not establish available waiver enrollment.
The section 1915(i) state-plan option allows home and community-based services subject to its own needs-based criteria and state-plan requirements. The section 1915(j) option supports self-directed personal assistance under specified conditions, including participant direction over some services or budgets. The Community First Choice authority provides a further statutory pathway for particular attendant services and supports. Each authority should be evaluated through its own conditions rather than merged into a single generalized 'Medicaid home-care policy.'
A person-centered service plan is analytically central because it links assessed needs to specified services, choice, providers, and timing. The federal HCBS settings regulation establishes standards intended to protect integration, autonomy, privacy, and choice in qualifying settings. The state implementation context and actual service agreement determine how those principles are realized. A compliant setting on paper may still experience missed visits or inadequate staffing; program standards and observed care delivery are different data.
The interaction of multiple financing streams is a frequent challenge. A beneficiary may use Medicaid-funded personal care, Medicare-covered medical home health, a private long-term care policy, and uncompensated family assistance. Payment priority, duplication restrictions, coordination-of-benefits rules, and claim documentation require case-specific examination. The Medicare coverage boundary is particularly important: medical coverage and custodial support have different eligibility logic. An economic analysis that classifies every medically related home visit as a long-term care insurance claim overstates the scope of the latter.
Environments, access, and continuity
Community living depends on home configuration, transport, reliable personal assistance, accessibility, and local service systems. An otherwise generous reimbursement design can fail if the qualifying service is unavailable at the required time. Environmental features thus act as effect modifiers for the practical value of insurance. The same nominal monthly allowance can finance an adequate schedule in a region with many workers and fall substantially short where wages and travel costs are high.
Coverage design can influence the ability to remain at home in several ways. Home modifications may reduce the need for repeated physical assistance; respite may sustain a family caregiver; day services may provide supervision and social participation; coordination can prevent a lapse when a provider leaves. These effects require measurement and cannot be inferred simply from inclusion on a service list. Each service also has potential substitution or complementarity with informal support and inpatient or residential care.
The UN community-inclusion obligation supplies a principled perspective on choice and equal opportunity. In insurance science, the related empirical question is whether coverage improves the feasible set of residence and support arrangements for individuals with similar needs. Relevant measures include the number of authorized and delivered hours, missed shifts, caregiver time, avoidable moves, participant-reported control, and time living in a chosen community setting. A single institutionalization rate captures only part of that picture.
Cash and Reimbursement Features
Contractual payment mechanisms
Payment structures determine how insurance transfers financial risk and how claimants can convert benefits into care. Expense-incurred reimbursement pays an eligible expense, subject to policy ceilings, proof, service qualifications, and other conditions. An indemnity arrangement pays a predetermined amount for a qualifying day or episode, potentially without requiring the same amount of actual expense. A disability-style cash benefit provides a specified periodic payment during a qualifying state. A hybrid design may combine reimbursable services with cash, care coordination, or special allowances.
These mechanisms can be described in terms of conditional payment functions. Let q_t denote the eligible functional state at period t; x_t eligible incurred expenditure; L_t a contractual cap; a_t an authorized service quantity; and r_t the benefit rate. A simple reimbursement form is B_t = 1(q_t) min(x_t, L_t), assuming all contractual and invoice conditions are satisfied. A fixed cash benefit is B_t = 1(q_t) K_t, where K_t is the stipulated cash payment. A service-unit benefit can be expressed B_t = 1(q_t) min(a_t, delivered units_t) r_t, subject to provider conditions. These are stylized accounting identities, not descriptions of every actual policy.
The NAIC long-term care materials explain that insurers use different policy terms, benefit amounts, and protections. These should be read as a warning against universalizing a marketing category. In reimbursement insurance, unspent eligibility does not generally create an automatically payable balance. In cash insurance, payment may be available despite low formal-service utilization, but the amount need not equal the cost of maintaining an appropriate care plan.
Payment method creates distinct risks. Reimbursement requires accurate invoicing, contractor qualifications, and settlement; exposure is tied to eligible expenditure, which may rise with service prices and utilization. Cash reduces invoice dependence but creates uncertainty about how much of the money becomes formal care and may leave the beneficiary exposed to price inflation. Service-unit benefits establish a measurable deliverable but can become unavailable during workforce shortages. Each structure distributes flexibility, audit costs, basis risk, and exposure differently.
For a clearly hypothetical illustration, consider a claimant who incurs $4,800 in qualifying monthly home assistance, holds a reimbursement cap of $3,500, and satisfies all applicable conditions. The payment is $3,500 and the eligible uncovered expense is $1,300, before any additional contractual deductible or co-payment. A separate $2,000 monthly cash policy would pay $2,000 if its own benefit trigger and continuation rules were met, regardless of whether exactly $4,800 was invoiced. These numbers are deliberately invented for methodological explanation; they are not observed U.S. policy rates or current care prices.
Contract triggers, limits, and duration
Contract generosity is a vector of conditions, not a scalar amount. The relevant terms include issue eligibility, a qualifying event trigger, waiting or elimination periods, daily or monthly ceilings, pool-of-money limits, benefit-period maxima, inflation options, services covered, family-care arrangements, exclusions, and premium continuation or waiver rules. A policy's advertised maximum must therefore be interpreted with its eligibility and duration provisions.
Suppose a reimbursement contract has a monthly eligible benefit ceiling of $3,500 and a remaining aggregate pool of $105,000. At full utilization and with unchanged payment conditions, the arithmetic maximum is 30 months. Actual duration may be longer if covered expenses are below the ceiling, or shorter if exclusions, benefit adjustments, or exhaustion rules apply. The calculation is a hypothetical identity, not an actuarial estimate of expected duration, because claim termination can result from recovery, death, institutional transfer, exhaustion, cancellation, or other conditions.
Waiting periods shift early spending to households or public programs and may create access discontinuities. In some contracts the waiting period counts calendar days once a claim is established; in others the terms may require days of covered services. The distinction can materially change out-of-pocket financing when care begins intermittently. Benefit duration also interacts with the possibility of progressive dependency: exhaustion may occur precisely when household needs are increasing. The researcher should examine wording and actual claim experience before concluding that a shorter period improves efficiency.
Inflation protection raises additional design questions. A benefit linked to a general consumer price index, a fixed annual escalator, or a nominal ceiling may evolve differently from labor-intensive personal-care costs. The real service capacity of a benefit is approximately nominal payment divided by the local price per qualifying care hour. A nominal 3% annual benefit increase combined with 5% annual care-price growth reduces purchasable hours by roughly 1.9% a year [1.03/1.05 - 1], ignoring changes in need and service supply. These are assumed rates used to demonstrate the mechanism; the study makes no claim that 3% and 5% are observed market forecasts.
Public cash and service schemes provide useful contrasts. Under Germany's home care benefit arrangements, different cash and professional-care benefit amounts apply by care grade, and combining arrangements is possible within statutory rules. The 2026 federal benefits table lists, for example, monthly care allowance of EUR 347 and professional home-care benefit up to EUR 796 at care grade 2. Those figures are ceilings in distinct statutory categories, not evidence that all costs of home support are paid. National program design, need assessment, and household contributions remain central.
In Singapore, the CareShield Life cash design uses a severe-disability trigger and periodic payouts that can support a mixture of family and paid care. Its benefit adequacy must be judged against the beneficiary's local care costs and other support sources. The comparison reveals the importance of keeping insured cash, covered service expenditure, and total social cost distinct.
Actuarial cash flows and adverse selection
The actuarial problem is dynamic because functional dependency, service setting, survival, prices, and contract eligibility evolve over time. Let S_t represent the insured state at time t, with categories such as independent living without payable care, qualifying care at home, qualifying care in another community setting, institutional care, and death. Let P_ij(t) denote the probability of transitioning from state i to j over a specified period, and let b_j(t) be the expected policy payment conditional on state j. A stylized present value of benefits is:
EPV(B) = E[ sum over t of v(t) B_t(S_t, utilization_t, prices_t, contract_t) ].
Here v(t) discounts future cash flows. The expression must reflect benefit triggers, elimination periods, caps, taxes where applicable, projected care costs, coverage continuation, and death. It is a valuation framework familiar from multistate insurance mathematics, not a newly proposed theory. Practical calibration requires credible policy-specific incidence, recovery, utilization, lapse, expenses, and mortality data.
Risk of care dependency cannot be inferred from prevalence alone. A high prevalence of limitations among older persons does not measure the rate at which an insured population enters payable claims because the insured cohort has distinctive entry age, underwriting, policy survival, eligibility criteria, and exposure duration. Similarly, historical insurer payments are shaped by benefit designs, claims management, and service supply. Experience studies must separate state incidence, time until claim payment, payment severity, duration, and competing exits.
Private information and selection matter. Finkelstein and McGarry's study shows that long-term care insurance purchasing behavior can reflect multiple dimensions of private information, including risk expectations and preferences for insurance, complicating simple predictions based on a single risk variable. The finding does not imply that every observed portfolio experiences the same direction or magnitude of selection. Underwriting, market institutions, and policyholder attitudes determine the empirical outcome.
Public coverage also interacts with private demand. Brown and Finkelstein's analysis studies how Medicaid's role in long-term care financing may influence private insurance purchase, using an economic model and U.S. program structure. Its numerical conclusions should not be transplanted automatically to different means tests or financing systems. For home-care benefits, the crowd-out question depends on which expenditures a public program covers, the income and asset consequences of eligibility, and whether individuals perceive private care flexibility as valuable.
A defensible premium or reserve model must also recognize the covariation of longevity and dependency duration, wage-driven care inflation, provider supply, and policyholder persistence. Low frequency combined with long episodes can produce heavy-tailed claims costs. Changes in institutional discharge policy can change home-care exposure even when underlying health conditions are unchanged. Data and scenario documentation should therefore accompany any point estimate of an expected benefit cost.
Incentives, administrative controls, and distribution
Benefits can affect decisions made by claimants, relatives, providers, insurers, and governments. Reimbursement may encourage the purchase of additional covered formal hours up to a cap; cash may enable assistance outside a licensed provider network; family-care eligibility may compensate previously unpaid labor. These are plausible behavioral channels. Their direction and size need credible evidence and cannot simply be inferred from the presence of coverage.
Claims verification is part of benefit design because paid home services are dispersed, recurrent, and sometimes difficult to observe independently. Documentation can include care assessments, approved plans, invoices, provider rosters, visit records, electronic verification, and periodic reassessment. Controls should match the risk being addressed. An insurer needs evidence of payable services but must also consider administrative burden, privacy, and the potential to interrupt legitimate care. The optimal scope of verification depends on expected error, fraud risk, service market characteristics, and the costs of delaying assistance.
Self-direction changes who organizes care. The Medicaid self-directed personal assistance option provides an institutional example in which beneficiaries can have greater control over personal assistance within defined program safeguards. Self-direction may improve fit and continuity when supported by a workable budget, recruitment tools, backup arrangements, and administrative assistance. It cannot eliminate local labor scarcity.
Distributional consequences are substantial. A benefit that pays a fixed dollar amount has greater real purchasing power in a low-price service market than in a high-price market. A reimbursement contract that excludes family members may disadvantage households where formal care providers are scarce. A requirement to pay the provider before receiving reimbursement can create a liquidity barrier even for a household with valid insurance. Thus formal generosity, realized generosity, and equitable access must be evaluated separately.
The OECD international affordability analysis shows that long-term care needs can impose major costs relative to older people's resources, with substantial variation across institutional arrangements. For insurance research, the relevant adequacy measure is the distribution of residual needs-adjusted out-of-pocket spending after benefits, not the average nominal face amount alone. Assessment should include households with zero formal-care purchases, because their low spending may reflect informal care, limited need, service shortage, or unmet need.
Continuity of Community Living
The meaning and measurement of continuity
Community living is a multidimensional outcome involving residence, effective support, choice, participation, and stability through changing needs. A person may remain at a private address while receiving inadequate assistance, or may move to a different supported community setting and experience more autonomy. Conversely, a temporary hospital admission does not necessarily mark the end of community residence. Research requires an operational definition that reflects the question being asked.
One indicator is the proportion of observable days spent living in the community rather than in an institutional setting. A second is sustained community residence after discharge from a skilled nursing facility. A third is the probability of a first long-stay nursing facility admission within a defined period. Each answers a different question. Other meaningful outcomes include unmet assistance needs, delayed discharge, emergency admissions, self-reported autonomy, unplanned caregiver work loss, housing stability, and satisfaction with care. The UN Article 19 framework emphasizes the ability to choose residence and access support. In evaluation, counting residence days alone does not establish that meaningful choice is present.
Continuity has distinct transitions: stable home care, escalation of needs, temporary hospital or rehabilitation admission, return home with augmented services, transfer to assisted living, entry into long-term residential care, and death. A claimant's insurer may cover some transitions and exclude others. A policy that pays for home care but suspends benefits during hospitalization can produce fragmented financing at the precise moment when discharge coordination is necessary. A flexible care plan and clear rules for resuming coverage may matter as much as the monthly payment maximum.
The relationship between benefits and residence is also bidirectional. The service package may influence discharge feasibility, while a person's residence and clinical severity determine which services become authorized. A population with high community-care spending may consequently have greater underlying need rather than better protection. Analysts should define when an individual becomes at risk of an event and how later claims are attributed to the exposure that preceded it.
Comparative institutional designs
U.S. Medicaid finances a diverse set of HCBS arrangements, with federal statutory authorities and state implementation determining services and enrollment. The 1915(c) waiver design connects community support to an institutional level-of-care criterion and approved service plans, but allows state-specific targeting and enrollment limits. The framework can support community discharge and avoidance of institutional admission, yet statewide headline program availability is an inadequate measure of an individual's actual access.
Waiting lists reveal a distinct access margin. KFF's 2026 HCBS waiting-list survey reported a 34-month average period before accessing services among responding states and programs with the relevant data. The data combine waiting and interest lists, differ in eligibility screening, and include retained information for nonresponding states. The figure is a survey statistic about recorded lists, not a national actuarial estimate that each listed individual had a clinically verified need, an enforceable entitlement, or the same wait. In a causal study, list definitions, state policies, and censoring require individual-level investigation.
The Medicaid HCBS settings standards address community integration and participant choice. These requirements affect what counts as an appropriate funded setting, beyond the insurer's expenditure alone. A service may be formally offered but functionally unusable if its provider cannot be selected or if everyday routines are excessively restricted. Rights-sensitive measurement therefore asks whether eligible people can exercise practical control over their support arrangements.
German social long-term care insurance uses care grades and differentiated cash and service benefits, creating options for professional home care and household organization. The 2026 statutory benefit summary shows grade-specific amounts, respite-related support, day services, and housing-adaptation assistance. Germany's combination benefit adjusts cash in relation to used professional services, rather than simply paying full cash and full service maxima together. This structure makes the distribution between informal and formal provision an explicit choice within statutory limits.
Singapore's CareShield Life severe-disability benefit insures a severe dependency contingency with cash payouts. The cash mechanism is compatible with several care settings and arrangements. It grants flexibility in use of funds while leaving the question of sufficient formal care unresolved. For 2026 the published base monthly payout is SGD 689 for a claim reaching the qualifying severe-disability threshold in that year, with individual payout growth subject to program rules. This should not be read as the monthly cost of full-time personal assistance or as the amount every eligible cohort necessarily receives.
Australian Support at Home contribution rules show another design. Clinical supports can carry no participant contribution, whereas other categories may involve means- and service-related contributions. The official description records full government funding of approved personal care services from October 1, 2026, while other independence and everyday-living contributions vary. This changing classification makes edition date part of any comparison: an account based on program commencement in 2025 can misstate the cost exposure applicable in late 2026.
The comparison illustrates four dimensions relevant across systems: who qualifies, what service or cash is financed, how payment is constrained, and how public or private institutions convert eligibility into delivery. Institutional designs occupy different positions on those dimensions. They should be compared as conditional arrangements rather than ranked by a single benefit amount.
Workforce constraints and informal care
Home care is labor intensive, and assistance often must be supplied at a precise time. A schedule with six required daily visits cannot be replaced with the same aggregate number of hours delivered in a single weekly session. Workers may be unavailable because of travel distance, turnover, unsocial hours, or a shortage of particular skills. Insurance value therefore depends on scheduling reliability and substitutability of providers, not merely the reimbursement rate.
A basic care-purchasing calculation makes this explicit. Let H denote required qualifying hours per month and w the effective price per hour, including travel or agency charges when they are eligible. Required formal-care spending is H × w. Let L be the monthly payment limit. The financed proportion is min(L, H × w)/(H × w), conditional on accessible providers and qualifying expenditure. If the provider supply ceiling is Q hours below H, the observed service hours are at most Q even if the benefit could finance more. Unspent entitlement can thus coincide with significant unmet need.
Informal support introduces a second production system. Family members may provide daily assistance without monetary transactions, reduce paid formal-care purchases, or become exhausted by unpredictable schedules. A design paying family caregivers can change the composition of remuneration and the distribution of economic burden, even when underlying hours do not change. A reimbursement-only design might encourage formal substitution but may also leave households exposed when licensed providers are unavailable. Cross-country comparisons require time-use measurements rather than equating zero formal spending with independent functioning.
Respite and backup services are economically relevant because their value may be realized through preventing the breakdown of another caregiver's supply. The WHO long-term care package views support systems and integrated service provision as part of long-term care capacity. In an insurance context, the effective design of respite includes eligibility, covered duration, provider substitution rules, and emergency arrangements. Evaluations should measure both the insured participant's outcomes and the consequences for unpaid caregivers.
Housing is another limiting factor. Transportation, accessible bathrooms, safe transfers, heat and power reliability, and proximity to a qualified provider can determine whether home residence is feasible. Insurance benefits that finance practical adaptations may reduce required assistance hours for some people, yet those effects vary by dwelling and impairment. A pre-existing inaccessible home can absorb a large benefit without creating an adequate care environment.
Empirical evidence and competing explanations
Empirical findings provide conditional support for the relevance of HCBS generosity to community outcomes. Wang and colleagues' 2024 study examined dual-eligible individuals discharged from U.S. skilled nursing facilities between 2010 and 2013. Higher HCBS breadth or intensity was associated with a greater probability of remaining in the community and lower risks of nursing-home admission; certain measures were also associated with lower short-run hospitalization. The analysis used linked administrative data and facility fixed effects. These associations are important evidence, but residual geographic, programmatic, and patient differences complicate a claim that expanding one particular insurance benefit causes the full observed change.
The causal problem arises because states choosing richer HCBS programs may simultaneously differ in nursing-home bed supply, hospital practice, housing, underlying health, fiscal capacity, and caregiver networks. Service availability can influence who is discharged from a facility in the first place. Without comparable populations and appropriate identification, apparent gains after discharge may partly reflect selection into community discharge.
For a carefully designed causal study, the exposure should be defined before the outcome window. Candidate designs include a difference-in-differences comparison around a plausibly exogenous state policy change, a narrow eligibility threshold analysis where assumptions are supported, or an instrumental variable with a defensible exclusion restriction. A claim of causation would require testing pre-intervention trends, possible spillovers, composition shifts, policy co-interventions, and the timing of implementation. Those conditions are not established simply by an official benefit expansion date.
The distinction between institutional placement and quality of life also requires attention. Avoiding a facility admission can be valuable when it preserves chosen residence with adequate support. It can also conceal unmet care if a participant lacks access to alternative services. Outcome measurement should include adverse events, hospitalization, caregiver burden, participant choice, and access to necessary assistance. The OECD affordability report shows why costs and accessibility must be considered together: households can face severe financial pressure even when services exist.
The current evidence thus supports a conditional conclusion. Home and community-based benefits can expand the set of feasible community care arrangements. Their realized contribution to continuity depends on the interaction of financing, actual service supply, household support, functional need, and residential environment. A stronger conclusion about any particular contract feature requires a design that isolates the relevant mechanism.
Evidence Validation and Sensitivity
Denominators and multistate experience
The first requirement of an actuarial or epidemiological analysis is a denominator that matches the proposed claim. A proportion of older adults reporting limitations is not a claim incidence rate. The number of monthly recipients is not a measure of distinct newly payable episodes. A count of Medicaid participants is not directly comparable with private-policy insured person-years. Differences in underwriting, waiting periods, contract survival, care thresholds, and country coverage change what those figures mean.
For claim incidence, define exposure at risk and the number of new qualifying episodes during a specified calendar period. For claim duration, identify episode start and end dates, reopening rules, and competing exits. For expenditure, distinguish charges submitted, approved charges, insurer-paid amounts, public payments, family payments, and unpaid assistance. These components are related but cannot substitute for one another.
The multistate formulation uses transitions between independent residence, community-based care, institutional care, recovery, and death. In practical insurance data, additional states may include waiting-period qualification, suspended claims, exhaustion, and policy lapse. A multistate model with absorbing death can separate mortality from recovery and institutional transfer; otherwise a general 'claim termination' rate can create systematic duration error. Censoring arises when the observation period ends before the claim does. Left truncation occurs when individuals enter the observation data after earlier care episodes have begun. Both require explicit treatment.
An estimate of community residence duration should include death as a competing event instead of treating all deaths as uninformative censoring. If death is frequent among individuals with severe dependency, simple Kaplan-Meier estimates of long-term institutional admission that censor deaths can overstate absolute admission probabilities. Cause-specific hazards, cumulative incidence functions, and restricted mean community residence time answer related but distinct questions. An analyst should justify the selected estimand before choosing a procedure.
Service utilization should be counted at the level appropriate to the economic question: visits, care hours, care days, claim dollars, or people. Aggregating visits into dollars mixes variation in physical assistance with price changes. An apparent increase in claim severity may reflect wage inflation, more hours, a shift toward certified providers, or an older and more impaired claimant mix. Decomposition is therefore essential for pricing, reserving, and policy comparison.
Model specification and financial stress
A transparent scenario analysis can illuminate relationships without purporting to estimate real experience. Consider a hypothetical insured individual requiring 100 formal-care hours per month, with a contractual reimbursement cap of $3,500 and a qualifying hourly price of $40. The eligible expense is $4,000; the insurer pays $3,500 and the individual bears $500, assuming no other offsets or deductibles. At 20% higher hourly cost, eligible expense becomes $4,800 and the residual rises to $1,300 while the capped insurer payment remains $3,500. Under an uncapped 80% coinsurance design with otherwise identical eligibility, insurer spending would rise from $3,200 to $3,840 and the residual from $800 to $960. The two designs transmit price shocks differently. Each illustration holds need and utilization fixed, which is itself an explicit modeling assumption.
A second hypothetical scenario introduces a provider shortage: the same individual requires 100 hours, but only 60 qualifying hours can be delivered. At $40 each, observed spending is $2,400, and a reimbursement insurer may pay $2,400. The gap of 40 hours is a service deficit despite a theoretical $3,500 benefit maximum. The lower claims bill does not demonstrate better health, lower dependency, or adequate insurance. A purely expenditure-based assessment would miss this welfare loss.
A third scenario examines inflation over five years. If eligible hourly care cost grows at 5% annually while a capped benefit grows at 3%, purchasing power measured in care hours changes by a factor (1.03/1.05)^5, approximately 0.908. The ceiling buys about 9.2% fewer hours after five annual adjustments, before any change in the person's dependency level. The result depends on those hypothetical rates and an unchanged mix of eligible services. A policy evaluation must separately model wage escalation, general price change, utilization intensity, and benefit-indexation rules.
A fourth scenario concerns waiting-period design. Suppose an eligible person needs $120 per day of payable assistance. If a 30-calendar-day elimination period begins when eligibility is certified, an otherwise identical claimant may finance $3,600 of early support before payment eligibility, assuming continuous care. If the policy counts 30 service days and services are scheduled only on weekdays, the waiting interval can last substantially longer than 30 calendar days. Contractually permitted counting rules and care patterns therefore drive liquidity requirements. The $120 amount is invented.
An illustrative severity decomposition is E[annual paid cost] = expected qualifying claim count × expected discounted payment per qualifying claim, subject to cohort definitions and state dependence. This arithmetic can guide sensitivity categories but does not allow one to multiply unrelated aggregate population prevalence and arbitrary provider prices. A credible forecast requires exposure-based incidence, episode duration, utilization, provider payment rates, offsets, lapse, mortality, and uncertainty in the underlying assumptions.
The OECD cross-country projections describe rising demand for care as populations age. They provide context for stress testing, but their public-system projections should not be imported directly as claim cost assumptions for a single private policy. Differences in contract definitions, service eligibility, national wages, and population mix can dominate aggregate trends.
Source criticism and identification
Source verification has three levels. Identity verification confirms the author or institution, document, date, and URL. Substantive verification confirms the exact proposition, relevant passage or abstract, and limitations. Transport verification asks whether the proposition remains valid when applied to another jurisdiction, calendar period, insured cohort, benefit type, or outcome.
For U.S. Medicaid, the operative unit may be a specific state waiver or state-plan amendment. A federal 1915(c) authority description establishes permissible structures but does not by itself prove a given state's current service caps. For private insurance, a regulator's long-term care guidance establishes recognized contract distinctions, but proof of a particular insurer's liability requires the applicable policy form, endorsements, claim facts, and law.
For comparative public programs, amounts and service categories must be dated. The German 2026 benefits schedule is a dated official schedule. The Singapore official insurance description identifies current payout and qualification principles. Australia's October 2026 personal-care update illustrates a within-year policy change that must be checked against publication timing. Comparability requires documenting the exact service category and participation assumptions.
Scientific literature has different limitations. The 2006 selection study concerns private information in a particular long-term care insurance setting; it does not supply a universal loading factor. The Medicaid interaction analysis develops a specific model of U.S. private insurance demand; it is not a measurement of provider supply or future reimbursement risk. The 2024 discharge-outcomes study provides associations within its observed U.S. dual-eligible discharge population, not a randomized estimate for all people needing community care.
Selection into insurance, into benefit eligibility, into a provider network, and into a discharge cohort can occur at different times. Conditioning on post-treatment service receipt can introduce collider bias: both severity and provider supply may influence whether a service was received. Similarly, excluding people who died before a follow-up survey can alter the survivor population and apparent care quality. Research designs must draw these pathways explicitly, state the target estimand, and retain competing outcomes.
Comparative results are strongest when the inquiry uses matched definitions, overlapping periods, jurisdiction-specific law, and verified source metadata. Cross-country tables without these controls risk a false precision. The choice to show an intentionally limited set of credible comparisons rather than many unqualified national entries is a substantive editorial decision favoring accurate scientific inference.
Substantive and editorial verification
The research evidence is organized at the level of named external works, legal or supervisory sources, and precisely scoped checked claims. The companion ledger records source identity, publication details, claim relevance, geographical or legal scope, and the depth of access. Every scientific inline hyperlink is associated with that ledger and with a complete reference entry. Short anchors identify the relevant document or factual phrase; source URLs are not presented as unexplained brackets in the prose.
The internal scientific review examines five issues: whether care need has been separated from eligibility, authorization, and delivered service; whether legal instruments are confined to their jurisdiction and temporal scope; whether observed statistics are distinguished from hypothetical arithmetic; whether the causal language matches each study design; and whether the multistate specification treats recovery, mortality, institutional transfer, and censoring coherently.
Editorial review separately examines exact title identity, unnumbered chapter and subsection hierarchy, American-English usage, terminology consistency, link readability, bibliographic completeness, and preservation of publication provenance. The author and publisher are identified consistently. No date of public publication is assigned to this research manuscript. The final technical gate is an independently executed local source-link check plus byte-level SHA-256 verification; its actual result must be recorded in the machine-readable passport and the scientific Notion record rather than inferred from this description.
Scientific Synthesis and Research Agenda
Integrated findings
Home and community-based care insurance finances a contingent combination of functional states, eligible service arrangements, and future payment obligations. The insured object has at least four stages: assessed need, claim eligibility, authorized care, and delivered care. The central analytical implication is that additional benefit generosity at one stage may fail to produce additional realized care at another stage when the limiting factor is provider supply, liquidity, administrative friction, or housing accessibility.
The comparison of reimbursement, cash, and service-unit forms supports a second conclusion. Reimbursement is naturally aligned with verifiable eligible costs but exposes the policyholder to residual needs, provider restrictions, and prepayment burdens. Cash offers flexibility across household arrangements but creates a potential gap between money and adequate formal services. Service-based funding defines deliverables more directly but relies on local provider capacity. The relevant design assessment involves both expected liability for the payer and actual feasible care for the beneficiary.
A third conclusion concerns time. Long-term care episodes are dynamic and uncertain in duration. Waiting periods, inflation protection, benefit exhaustion, service transitions, and death all influence the value of a promise. Comparing two plans at their initial monthly maximum obscures their behavior when care lasts many years or becomes more intensive. Valuation and consumer comparison need projected care-hour purchasing power and benefit continuity under multiple plausible state transitions.
Fourth, international public programs show multiple valid institutional arrangements. U.S. Medicaid, German social care insurance, Singaporean severe-disability cash insurance, and Australian assessed home-support programs differ in enrollment logic, service definition, fiscal architecture, and benefit delivery. A single ranking by nominal dollars would combine incompatible objects. Jurisdiction-specific comparison is informative when it discloses eligibility, financing, service supply, and household residual cost.
Fifth, the empirical evidence supports studying community residence as an outcome, but it does not allow a direct causal conclusion for every policy feature. The observed discharge associations are compatible with beneficial effects of richer HCBS arrangements and with some remaining differences among states, facilities, and populations. The research therefore distinguishes the evidence of an association from an intervention-specific causal estimate.
Insurance, public-policy, and rights implications
For insurance contract design, the most informative disclosure is a worked example that follows a claimant through eligibility, a care plan, delivered services, invoices, and actual payment. The example should expose limits, waiting-period counting, provider qualifications, out-of-pocket costs, and the consequences of care interruptions. Insurers and regulators can improve comparability by defining common terms while preserving contractual distinctions. The NAIC insurance information provides a relevant supervisory reference for understanding consumer-facing contract features.
For actuarial practice, a model should connect the insured cohort to the correct incidence denominator, separate severity and duration, and include changes of care setting, provider prices, benefit utilization, and competing exits. A stress test should include simultaneous wage inflation, prolonged care episodes, and formal-worker shortages rather than varying only one average cost factor. Such scenarios can reveal when a contract continues to fund care adequately even if insurer reserves remain mathematically sufficient for the promised dollar benefit.
For public policy, financing arrangements should be assessed jointly with service capacity and administrative implementation. The Medicaid waiver requirements include individual plans and health-welfare safeguards, demonstrating that effective delivery involves operational institutions as well as budget appropriations. Program evaluation must also consider whether waiting lists, service documentation, and provider availability restrict practical access.
For rights analysis, community residence should be evaluated in terms of choice and inclusion alongside expenditures and admission rates. The Article 19 community-living standard provides the normative context for that multidimensional evaluation. The empirical question is whether financed assistance expands feasible choices under real environmental conditions. An individual who stays home with severe unmet needs has a different outcome from one who remains there with reliable support and meaningful control.
For households, the fiscal significance of informal care deserves explicit measurement. Cash transfers and professional services can alter who bears the cost of care, including opportunity costs from reduced employment or caregiving hours. Measured insurer spending is a partial economic account. Evaluations of value should identify the person whose burden changed, the other expenditures displaced, and the benefits that depend on sustained unpaid work.
Testable research questions
A mature research program should address the following specific questions through original policy-level and person-level data:
- For people with similar functional profiles, how do reimbursement versus fixed cash benefits affect delivered formal-care hours, unmet needs, caregiver time, and claimant financial exposure? Identification must account for voluntary policy selection.
- Does adding a payable family-care or respite provision change the probability of sustained community residence, and is the effect mediated by reduced caregiver turnover or improved service continuity?
- How much of regional variation in paid claims is explained by wage rates, provider supply, claim incidence, and utilization rather than underlying functional need?
- Do different elimination-period counting conventions generate distinct delays in actual receipt of care for people with intermittent needs?
- How does benefit indexation compare with local personal-care wage growth over five-, ten-, and twenty-year horizons?
- Which changes to provider qualifications and self-directed care improve realized access without materially increasing unsupported claims, after accounting for verification burden?
- How do institutional discharges, care-setting transitions, and mortality alter the expected present value of benefit pools and the apparent success of community-care programs?
- What proportion of authorized home-care services is actually delivered within the specified window, and which operational barriers explain the gap?
- Do observed community-residence associations withstand longitudinal designs with well-defined counterfactuals, competing risks, and state-specific program implementation dates?
- How do legal choice and person-centered planning standards translate into measurable recipient experience, especially for people requiring cognitive supervision?
Answering these questions requires longitudinal records that connect assessment, authorization, service delivery, payment, residence, and outcomes; jurisdictional laws and contract terms; and credible identification strategies. This agenda derives directly from the reviewed evidence and clarifies future empirical work. It is not presented as a new theory or independent doctrine.
Conclusion
Home and community-based care insurance transforms functional dependency into contingent access to services and financial transfers across time. Its meaningful economic unit is the supported care arrangement: a person with specified needs, a qualifying entitlement, a set of authorized supports, actual available providers, and a payment mechanism subject to temporal and monetary limits. A benefit's nominal ceiling does not by itself describe its capacity to sustain community residence.
The comparative evidence establishes the importance of legal specificity. U.S. Medicaid home- and community-based authorities govern eligible populations and funded services under defined federal and state requirements. German social insurance provides grade-specific cash and service benefits. Singapore's CareShield Life provides severe-disability cash payments. Australia's Support at Home program funds categorized services with evolving participant-contribution rules. These programs illuminate different mechanisms rather than interchangeable measures of one identical insurance product.
Actuarial analysis must distinguish claim incidence, duration, utilization, price, transition between settings, recovery, and death. Economic analysis must account for service supply, informal labor, liquidity, selection, and residual spending. Evaluation of social outcomes must include meaningful community choice and the adequacy of delivered care as well as institutional admission rates.
The central finding is that protection is realized through the interaction of entitlement and delivery. A scientifically adequate benefit design and a scientifically adequate evaluation both trace that interaction explicitly. The present monograph contributes a documented comparative synthesis and a reproducible analytical structure for subsequent empirical inquiry, while leaving causal estimation, portfolio calibration, and jurisdiction-specific policy adjudication to studies with the relevant original data.
References
Australian Government Department of Health, Disability and Ageing. (2026). Support at Home participant contributions. Australian Government. https://www.health.gov.au/our-work/support-at-home/charging-for-support-at-home-services/support-at-home-participant-contributions
Brown, Jeffrey R., and Amy Finkelstein. (2004). The interaction of public and private insurance: Medicaid and the long-term care insurance market. NBER Working Paper No. 10989. https://doi.org/10.3386/w10989. Later published in American Economic Review, 98(3), 1083-1102 (2008). https://www.nber.org/papers/w10989
Centers for Medicare & Medicaid Services (CMS). (2014; accessed 2026). Home and Community Based Services Final Regulation. Medicaid.gov. https://www.medicaid.gov/medicaid/home-community-based-services/home-community-based-services-guidance-additional-resources/home-community-based-services-final-regulation
Centers for Medicare & Medicaid Services (CMS). (n.d.; accessed 2026). Community First Choice (CFC) 1915(k). Medicaid.gov. https://www.medicaid.gov/medicaid/home-community-based-services/home-community-based-services-authorities/community-first-choice-cfc-1915-k
Centers for Medicare & Medicaid Services (CMS). (n.d.; accessed 2026). Home & Community Based Services 1915(i). Medicaid.gov. https://www.medicaid.gov/medicaid/home-community-based-services/home-community-based-services-authorities/home-community-based-services-1915i
Centers for Medicare & Medicaid Services (CMS). (n.d.; accessed 2026). Home & Community-Based Services 1915(c). Medicaid.gov. https://www.medicaid.gov/medicaid/home-community-based-services/home-community-based-services-authorities/home-community-based-services-1915c
Centers for Medicare & Medicaid Services (CMS). (n.d.; accessed 2026). Long-term care. Medicare.gov. https://www.medicare.gov/coverage/long-term-care
Centers for Medicare & Medicaid Services (CMS). (n.d.; accessed 2026). Self-Directed Personal Assistant Services 1915(j). Medicaid.gov. https://www.medicaid.gov/medicaid/home-community-based-services/home-community-based-services-authorities/self-directed-personal-assistant-services-1915-j
Central Provident Fund Board (Singapore). (2026 accessed). CareShield Life: Healthcare financing and benefit details. CPF Board. https://www.cpf.gov.sg/member/healthcare-financing/careshield-life
Central Provident Fund Board (Singapore). (2026 accessed). What is CareShield Life? CPF Board. https://www.cpf.gov.sg/service/article/what-is-careshield-life
Finkelstein, Amy, and Kathleen McGarry. (2006). Multiple dimensions of private information: Evidence from the long-term care insurance market. American Economic Review, 96(4), 938-958. https://doi.org/10.1257/aer.96.4.938. https://doi.org/10.1257/aer.96.4.938
German Federal Ministry of Health (Bundesministerium fuer Gesundheit). (2025, February 25; accessed 2026). Long-term care insurance. Federal Ministry of Health. https://www.bundesgesundheitsministerium.de/en/themen/pflege/online-ratgeber-pflege/long-term-care-insurance
German Federal Ministry of Health. (2025, December 11; applicable 2026). Leistungsansprueche der Versicherten im Jahr 2026 an die Pflegeversicherung im Kurzueberblick [Insured benefits in long-term care insurance in 2026]. Official benefit table. https://www.bundesgesundheitsministerium.de/fileadmin/Dateien/3_Downloads/P/Pflegeversicherung_Leistungsbeitraege/Uebersicht_Leistungsbetraege_2026_VA.pdf
German Federal Ministry of Health. (2026 accessed). Ambulante Pflegesachleistungen [Outpatient in-kind care benefits]. Bundesministerium fuer Gesundheit. https://www.bundesgesundheitsministerium.de/pflege-zu-hause/pflegesachleistungen
KFF. (2026, October 5). A look at waiting lists for Medicaid home- and community-based services from 2016 to 2026. KFF Medicaid HCBS Survey. https://www.kff.org/medicaid/a-look-at-waiting-lists-for-medicaid-home-and-community-based-services-from-2016-to-2026/
National Association of Insurance Commissioners (NAIC). (2026, July 23). Long-Term Care Insurance. NAIC Insurance Topics. https://content.naic.org/insurance-topics/long-term-care-insurance
Organisation for Economic Co-operation and Development (OECD). (2024, October 29). Is Care Affordable for Older People? OECD Health Policy Studies. OECD Publishing. DOI: 10.1787/450ea778-en. https://www.oecd.org/en/publications/is-care-affordable-for-older-people_450ea778-en/full-report.html
United Nations. (2006). Convention on the Rights of Persons with Disabilities, Article 19: Living independently and being included in the community. United Nations. https://www.un.org/development/desa/disabilities/convention-on-the-rights-of-persons-with-disabilities/article-19-living-independently-and-being-included-in-the-community.html
United States Congress. (2026 accessed). Internal Revenue Code, 26 U.S.C. section 7702B: Treatment of qualified long-term care insurance. Office of the Law Revision Counsel. https://uscode.house.gov/us/usc/119-120/t26/s7702B/
Wang, Sijiu, Helena Temkin-Greener, Adam Simning, R. Tamara Konetzka, and Shubing Cai. (2024). Outcomes after community discharge from skilled nursing facilities: The role of Medicaid home and community-based services. Journal of Applied Gerontology, 43(10), 1536-1543. https://doi.org/10.1177/07334648241242942. https://pubmed.ncbi.nlm.nih.gov/38581163/
World Health Organization (WHO). (2001; accessed 2026). International Classification of Functioning, Disability and Health (ICF). WHO. https://www.who.int/classifications/international-classification-of-functioning-disability-and-health
World Health Organization (WHO). (2024, May 8). Long-term care for older people: package for universal health coverage. WHO. https://www.who.int/publications/i/item/9789240086555