Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Health Insurance Deductibles Copayments and Coinsurance Design

Abstract. Deductibles, fixed copayments, percentage coinsurance, and out-of-pocket ceilings determine the distribution and timing of medical costs between insured people and financing institutions. Their effects depend on benefit eligibility, negotiated prices, accounting order, service classification, household liquidity, information, and provider behavior. This monograph establishes a source-critical account of the benefit mechanics, actuarial interpretation, causal evidence, financial protection, and international institutional variation of health insurance cost sharing. It derives the cash-flow structure of common nonlinear contracts, distinguishes point-of-care payment from expected annual liability and actuarial value, evaluates evidence from randomized experiments and quasi-experimental designs, and explains the distributive and clinical consequences of moving medical prices onto patients. The RAND Health Insurance Experiment demonstrates utilization responses to cost sharing; subsequent employer, pharmacy, and public-insurance evidence shows that reductions in use may include valuable care and that medical offsets, comprehension, and access vary across populations. Current institutional examples from the United States, the Netherlands, England, Australia, Singapore, and global comparative health financing establish how similar payment labels operate within different rights, subsidies, ceilings, and service systems. The synthesis identifies contract-level, individual-level, and system-level measurements required for a defensible assessment of a cost-sharing schedule. All numerical examples are illustrative unless expressly attributed to a specified source, jurisdiction, and year. This work advances an original analytical interpretation of established evidence without proposing a new proprietary theory.

Keywords: Health insurance; deductible; copayment; coinsurance; out-of-pocket maximum; nonlinear pricing; actuarial value; utilization; moral hazard; financial protection; pharmaceutical benefits; value-based insurance design; comparative health financing; health equity; causal inference.

Introduction

Health insurance converts an uncertain individual expenditure into a combination of predictable financing and contingent claims. Cost-sharing provisions specify how much of an eligible medical expense remains with the insured at the moment of service. They are legal payment rules, actuarial determinants of expected claims, behavioral prices facing patients, and distributive devices within a risk pool. A contract cannot be evaluated by its deductible alone: identical thresholds generate different expenditures when accompanied by different premium levels, negotiated prices, covered service categories, exceptions, accumulators, and out-of-pocket limits.

The operational distinctions in the federal health-coverage glossary locate cost sharing within covered services and separate it from premiums and uninsured expenses. A deductible requires payment toward eligible costs before specified coverage stages begin; a copayment is ordinarily a stated amount associated with a service; and coinsurance is a percentage of an eligible amount. The Marketplace definition of a deductible explicitly allows preventive benefits before satisfaction of the threshold and separate prescription-drug deductibles. Policy language governs whether a deductible also applies to a particular copayment or service, while insurer network terms establish the amount on which percentage participation is calculated.

Three research questions guide the analysis. How do multiple cost-sharing provisions interact within one claims schedule and across an insurance year? How does a nonlinear schedule change utilization, treatment decisions, spending, and provider incentives? How do these responses affect welfare, access, and the allocation of financial burdens within and between insured populations? The inquiry combines formal contract arithmetic with verified institutional documents and empirical research; it distinguishes mathematical identities, credible causal estimates, descriptive statistics, and policy judgments.

The unit of analysis changes with the question. A claim is the appropriate unit for adjudication; a person-year or family-year for accumulated liability; a care episode for clinical substitution; a drug-course for adherence; a plan-year for actuarial value; and a jurisdiction-period for comparative law. Pool-wide savings do not automatically imply patient welfare gains, and a lower point-of-care bill does not automatically imply lower total financing costs.

Foundations and Source-Critical History

Medical expense sharing as institutional allocation

Payment at the point of care existed long before modern insurance contracts. In contemporary systems it combines at least four functions: limiting the share of routine spending financed by pooled premiums or taxes, discouraging some use at the margin, differentiating covered services or provider channels, and allocating residual financial risk. These functions do not have the same measurement criterion. A pooled insurer may record fewer reimbursed claims even if a patient continues to receive care and pays the entire negotiated amount below the deductible; conversely, lower reimbursed spending may reflect care forgone entirely.

The distinction between contractual liability and the real price of care is central. Consider a visit with an undiscounted bill of 300 monetary units and a network allowed amount of 180. If the plan's percentage participation is 20%, the coinsurance base ordinarily concerns the eligible 180 rather than the provider's original list price. If an excluded service is outside the benefit package, its payment cannot be inferred from the plan's ordinary coinsurance rate. The Summary of Benefits and Coverage framework formalizes comparable disclosure of deductibles, copayments, coinsurance, exclusions, and coverage examples in the U.S. private-plan setting.

Experimental origins of modern evidence

The most influential experimental source is the U.S. RAND Health Insurance Experiment of the 1970s and early 1980s. Families were randomly allocated to health coverage with different financial participation and expenditure limits. A RAND retrospective of the experiment reports that higher cost sharing reduced utilization while reducing both more effective and less effective services. Manning and colleagues' randomized utilization study reports per-capita expenditures in the free-care plan 45% above those of the 95% coinsurance plan with an annual spending limit. This is a comparison of specifically randomized benefit designs, populations, and historical prices, rather than a permanent elasticity estimate for every health system.

RAND's comparative health findings also require distributional reading. Average health indicators changed less than utilization, while certain low-income and medically vulnerable participants benefited from free access, as summarized by Newhouse's early experimental review. The result challenges an interpretation that every avoided visit has the same social value. It also makes attention to subgroups essential for judging contemporary deductibles.

The shift toward nonlinear insurance and choice

Later insurance design increasingly linked initial high exposure, subsequent partial insurance, and a terminal ceiling. This creates a sequence of marginal prices throughout the year. A service at a January appointment may cost the individual almost its entire allowed price, while the same service in December may cost little if annual accumulators are exhausted. Expected annual price, the currently quoted copayment, and the person's perceived price can therefore diverge. Such divergence motivates the distinction between economically rational full-year calculations and observed decisions under uncertainty.

Evidence from an employer's conversion to high-deductible coverage shows why this distinction matters. Brot-Goldberg and colleagues (2017) found in their high-deductible natural experiment a spending reduction of approximately 11.8%–13.8%, attributable to quantity reductions rather than observed price shopping; reductions extended across potentially valuable and less valuable care. Because the sample was one firm's employees under one transition, the estimate is a causal finding within that setting, not a universal effect of imposing a deductible.

What the historical record supports

The strongest historical inference concerns a stable causal channel: when eligible care becomes more expensive to the patient, the probability or quantity of use often falls. The size, clinical composition, and eventual health consequences depend on population, service type, and institutional design. Contemporary analysis should retain this channel while rejecting a mechanical inference that lower observed spending necessarily means better allocation.

Deductible and Copayment Mechanics

Contractual vocabulary and eligibility order

A deductible is an accumulator threshold on specified eligible charges. A copayment is a predetermined amount for a specified covered event or unit, subject to the governing contract. Coinsurance is proportional participation in an allowed or otherwise contract-defined amount. An out-of-pocket maximum limits qualifying cost-sharing liability over a specified coverage period. In the U.S. Marketplace setting, the out-of-pocket maximum explanation excludes premiums, noncovered services, generally out-of-network expenses, and charges above the allowed amount. These exclusions make an out-of-pocket cap a precise coverage promise, not an unlimited ceiling on every possible medical bill.

A transparent adjudication sequence asks: Is the member eligible at the service date? Is the service covered? Does a network agreement or statutory reimbursement schedule apply? What is the allowed amount? Is any preventive, clinical, or income-based exception triggered? Which deductible accumulator receives the charge? Does the benefit use a fixed copayment, percentage coinsurance, or a combination? Which charges count toward the out-of-pocket limit? What insurer or public payer absorbs the residual? These questions must be answered in this order or through the exact order mandated by the applicable benefit contract.

The federal copayment definition illustrates a fixed payment of 20 monetary units on an allowed charge of 100 after the deductible has been satisfied, while noting service-specific variation. The federal coinsurance definition identifies percentage participation following the deductible. Neither description licenses adding both a full copayment and coinsurance to the same claim without an explicit coordination clause.

A standard continuous deductible-coinsurance calculation

Define aj≥0a_j\geq0 as the allowed eligible cost of service jj; D≥0D\geq0 as the annual deductible; dj∈[0,D]d_j\in[0,D] as the amount already credited to it before the service; and c∈[0,1]c\in[0,1] as the patient's post-deductible coinsurance rate. For a contract applying coinsurance after the remaining deductible, set rj=max⁡(0,D−dj)r_j=\max(0,D-d_j) and bj=min⁡(aj,rj)b_j=\min(a_j,r_j). The member's pre-cap liability on that service is

Lj=bj+c(aj−bj).L_j=b_j+c(a_j-b_j).

If MM is the annual qualifying out-of-pocket limit and mjm_j is the amount already credited toward it, a simplified cap-compliant member charge is

Yj=min⁡{Lj,max⁡(0,M−mj)}.Y_j=\min\{L_j,\max(0,M-m_j)\}.

The payer's liability for the eligible allowed charge is Ij=aj−YjI_j=a_j-Y_j, provided the benefit contains no other reimbursement phase, exclusions, or third-party payment adjustments. The deductible updates by the portion bjb_j credited under the contract and the out-of-pocket accumulator by qualifying YjY_j. This calculation is an illustrative special case, not a claim that all national policies follow this order. It assumes eligible costs, no independent copay, no coordination of benefits, and a single person-year accumulator.

For example, with a deductible of 1,000, earlier deductible credit of 800, allowed cost 500, and 20% coinsurance, the remaining deductible is 200; L=200+0.2(300)=260L=200+0.2(300)=260. A plan in which the member has 100 of qualifying out-of-pocket room remaining collects only 100 from the member for that covered service; the payer bears the other 400. If the plan has separate network and drug accumulators, the same arithmetic must be performed for the appropriate account rather than on one assumed global balance.

Copayment and combination schedules

A fixed post-deductible copayment ksk_s for service category ss is properly defined by the plan's benefit table. In a common simple arrangement, once the applicable deductible has been fulfilled, the member owes min⁡(ks,aj)\min(k_s,a_j) on the eligible service, subject to the remaining cost-sharing maximum; the insurer pays the balance. Other plans waive the deductible for visits with specified copays, treat a copay as additional to coinsurance, assess a per-day hospital charge, or apply a copayment only after a pharmacy tier threshold. The 2026 employer benefit survey documents such heterogeneity: 10% of covered workers had both copayment and coinsurance requirements for a hospital admission, in addition to any general deductible.

Drug formularies can use a fixed amount for generic medicines, a higher fixed amount for preferred brands, percentage participation for specialty products, or distinct pharmacy deductibles. A percentage without a dollar ceiling magnifies exposure to very expensive therapies. A fixed payment may simplify an individual claim's predictability, but repeated fixed payments can still create cumulative financial burdens. Some benefits pair coinsurance with minimum or maximum per-fill amounts; these terms must be explicitly parameterized.

Individual, family, and service-level accumulation

Family contracts may combine an embedded individual deductible and an aggregate family deductible, with rules governing whose expense satisfies each threshold. They may also maintain distinct person-level and family out-of-pocket ceilings. A correct benefit engine requires member identifiers, family membership and effective dates, plan-year boundaries, network status, and a reproducible ledger of accumulator events. A single printed family deductible is insufficient to infer one dependent's obligation.

Timing matters beyond the calendar-year arithmetic. A deductible can cause concentrated early-year expense even where annual liability is eventually bounded. A plan change, job change, interruption of entitlement, or different renewal date may reset or disallow carryover under applicable terms. The resulting burden depends on liquid wealth, credit, income volatility, and the probability of needing care before the threshold has been met.

Premium, expenditure, and actuarial value

Member total financial exposure for a year is the sum of premiums actually paid, eligible cost sharing, and any noncovered or other lawful charges. An actuarial value measure instead concerns the share of a defined reference population's covered allowed costs paid by the plan. Those are different estimands. A 70% actuarial value does not imply that a particular insured person receives 70% reimbursement on every service, spends 30% of her income on care, or has a 30% coinsurance rate.

U.S. metal-tier descriptions in Marketplace plan categories present approximate population-average plan-cost shares, and actual patient liabilities vary across services and utilization states. Likewise, different premiums are not evidence of identical expected insurer costs: contract subsidies, tax treatment, risk composition, and provider prices may differ. Designing deductibles entails pricing the entire joint distribution of care and risk, not multiplying a single average service bill by a coinsurance percentage.

Legal constraints and patient-facing information

Some benefits are exempt from otherwise applicable cost sharing. The Marketplace preventive-services rules explain that most U.S. plans cover specified in-network preventive services with no copay or coinsurance even before the deductible; the precise qualifying service and billing context matter. Income-based reductions can change the effective plan schedule: Marketplace cost-sharing reductions reduce deductibles, applicable copays or coinsurance, and the maximum out-of-pocket burden for qualified eligible enrollees who select the applicable Silver plan.

Regulatory amounts are plan-year and program specific. For 2026, the Marketplace out-of-pocket guidance states maximums of $10,600 for self-only coverage and $21,200 for family coverage; these are upper statutory limits for the relevant plans, not the deductible in every contract, and some plans set lower limits. The applicable federal cost-sharing regulation establishes the broader regulatory structure, including coordination with preventive requirements. An international comparison must never transplant these U.S. amounts into an unrelated insurance or health system.

Utilization Response and Elasticity

The nonlinear effective price

In economic models, a patient chooses use given expected health benefit, disposable resources, nonfinancial barriers, and the policy's marginal financial contribution. With a deductible the contemporaneous marginal contribution may be 100% of an eligible allowed cost until the threshold is met, then a fraction cc, and eventually zero for qualifying cost after reaching an applicable out-of-pocket maximum. This stylized piecewise structure is a contractual identity. Whether patients understand, anticipate, and respond to its different segments is an empirical question.

The demand elasticity of utilization relative to an effective patient price is the proportional response of quantity to a proportional change in that price, under a specified identification method. Elasticity is not identical to the percentage fall in total medical spending, because spending also changes with provider prices, mix of services, coding, visit severity, and selection. Research must specify whether it estimates demand at the point of contact, episode initiation, drug refill, or total annual consumption. For zero copays and large deductibles, a logarithmic elasticity at zero price cannot be used without an appropriate alternative specification.

RAND and the experimental utilization margin

The RAND final utilization findings show that utilization responds to financial incentives, with a particularly important change between free care and 25% coinsurance. Average effects conceal clinical and income heterogeneity. Because RAND included expenditure caps, the treatment was a nonlinear price schedule even when a study reports a single coinsurance rate. Reusing one coefficient outside its historical context without adjusting for income, care technology, benefit design, and covered population is an identification error.

Interpretation should distinguish a reduction in initial contacts from reduced service intensity after a contact. RAND's reported variation was especially apparent in medical contacts, and total spending changes need not follow a simple constant-elasticity curve. Modern emergency and specialty referral pathways may create different elasticities because patients have different information and control over quantities.

High deductibles, spot prices, and shopping

Brot-Goldberg et al. (2017) exploit a mandated plan switch to identify a spending response that largely occurred through using less care, with limited evidence of price shopping. Their spending-dynamics analysis also finds substantial response to prices encountered during the year rather than the economically relevant expected year-end marginal price. This exposes a practical design problem: an instrument conceived as a shopping incentive can operate principally as a deterrent to contact. The evidence does not imply that comparison shopping never occurs; it shows what was observed in that institutional setting and time period.

Price transparency and clinical guidance can change the mechanism. A patient who can compare elective imaging prices may substitute a lower-priced provider. An acutely ill patient in an emergency department has little opportunity to shop. Provider-directed referrals, concentration, and surprise billing may further separate the patient's decision from the price component that insurers seek to discipline. A credible evaluation must therefore distinguish price choice from the decision to receive care.

Adherence, prevention, and medical offsets

An increase in drug copays may reduce fills and adherence, creating downstream effects that are absent from the pharmacy's immediate claims total. Chandra, Gruber, and McKnight (2010), in their elderly cost-sharing study, document reduced physician and drug use associated with a California retiree benefit change and increased hospitalization, with savings and offsets falling on different financing parties. The result illustrates a fiscal externality: shifting the initial price to one population or payer can reallocate future costs to another payer.

The MI FREEE trial offers a complementary test of lowering charges for secondary prevention. Choudhry and colleagues (2011), in a randomized medication-coverage trial, assigned plan sponsors to full coverage or usual cost sharing for selected post-myocardial-infarction medications. The primary composite outcome was not significantly improved, while medication adherence and certain secondary vascular outcomes improved, patient costs fell, and the study found no statistically significant increase in overall spending. The distinction between primary and secondary endpoints matters: an analytically appealing intervention should not be described as having succeeded on an endpoint that failed the prespecified statistical test.

Coverage expansion is a different treatment

The Oregon Medicaid lottery tested eligibility for expanded insurance access, rather than isolated variation in a deductible within unchanged coverage. Baicker and colleagues' two-year clinical outcomes study found more preventive use and diabetes detection, less financial strain and depression, but no statistically significant changes in several measured physical health indicators over the observation horizon. It offers powerful evidence for insurance and financial protection, yet should not be pooled numerically with studies of an individual copayment change as though they estimated the same treatment.

Causal interpretation and welfare relevance

A meaningful evaluation requires counterfactual outcomes, not a before-and-after expenditure comparison alone. Employers adopting high deductibles may simultaneously change networks and prices; populations selecting generous plans may differ in anticipated risk; an increase in hospital costs can coincide with a decrease in pharmacy expenditures because disease severity changes. Randomization, natural experiments, difference-in-differences with defensible parallel trends, regression discontinuities at eligibility thresholds, or transparent instrumental variables may address selected confounders. Each method still requires a sharply defined exposure and credible exclusion of simultaneous policy changes.

Lower spending has multiple possible explanations: price reductions, substitution toward equivalent lower-cost services, elimination of low-value care, deferral of beneficial treatment, or reduced insurance reimbursement while the patient pays the same allowed price. Only some mechanisms unambiguously improve total welfare. Clinical outcomes, avoidable admissions, treatment discontinuation, and person-level spending are required to interpret the utilization change.

Equity and Welfare Tradeoffs

Risk sharing, timing, and household liquidity

A deductible is a transfer of first-dollar financial risk to the member. The incidence of that transfer depends on medical need, savings, income stability, and insurance subsidy rules. Annual caps bound some liabilities but do not remove a January liquidity shock. Patients may rationally decline or postpone care they cannot finance immediately even when the expected long-term benefit exceeds the price. Payment timing is therefore an independent dimension of financial protection.

Consider two members with identical expected annual claims but very different accessible reserves. The same 1,000-unit deductible may be manageable for one and prevent the other from accessing a covered diagnostic test. Measuring the average deductible as a share of mean income obscures that difference. A benefit analysis should use the joint distribution of income, liquid assets, underlying disease, claim timing, family size, and actual premiums.

The World Health Organization's 2025 universal coverage monitoring reports that 2.1 billion people experienced financial hardship from out-of-pocket health spending in 2022, using its specified updated global methodology. This is a worldwide financing statistic, not an estimate of the number harmed specifically by private-insurance deductibles. It establishes the scale at which direct patient payment and financial protection must be studied.

Clinical value is not uniform across services

If price sensitivity is high for both highly effective and marginal care, uniform cost sharing will reduce both. The RAND findings and the high-deductible employer study substantiate the need to distinguish care by clinical value and timing. Removing payment barriers for proven high-value services, designing exemptions for chronic-condition maintenance, and avoiding percentage charges on essential high-cost medicines may improve access, but the benefits depend on correct classification, uptake, and sustainable pooled financing.

The MI FREEE evidence supports examining targeted medication affordability rather than indiscriminately treating every service as equivalent. A plan can make a low-value elective intervention more expensive while keeping essential preventive care affordable, subject to clinical evidence and legal requirements. The resulting design should be tested against possible overtreatment, administrative burden, and changes in prescriber behavior.

Distribution across risk groups and payers

Cost-sharing schedules redistribute claims between relatively healthy and high-use members in any given year. Their ex-ante fairness depends on how premiums, taxes, risk adjustment, subsidies, and health shocks interact. Higher deductibles may lower the average pooled claim expenditure, but high-need members can face larger direct bills. Health-based differences are partly predetermined by chronic conditions and disability; the equitable consequence of a common price therefore depends on more than formal equal treatment.

The WHO European financial protection review, covering 40 countries with principally pre-pandemic household information, reports that medical out-of-pocket hardship disproportionately burdens the poorest groups and that outpatient medicines, medical products, and dental care are recurrent drivers. Portugal's 2025 WHO country assessment identifies heavy percentage copayments on outpatient medicines and limited automatic exemptions as continuing problems in that system. These findings are descriptive and institution-specific; their strongest implication is to analyze who remains charged for which essential goods.

Information, comprehension, and selection

Choice complexity creates its own welfare channel. Bhargava, Loewenstein, and Sydnor (2017) examined 23,894 employees choosing among options that differed in premiums and cost sharing and found widespread selection of financially dominated plans, with especially pronounced mistakes among lower-income employees. Their dominated-plan choice study is evidence that understanding the premium–deductible tradeoff cannot be presumed. An insurance choice menu may expose households to unnecessary total cost without any change in the medical services delivered.

Consumers should be able to identify predictable annual premiums, worst-case covered out-of-pocket exposure, provider network exclusions, prescription tiers, typical utilization scenarios, and which amounts accumulate. The U.S. official coverage-summary framework is one institutional response. Better disclosure, examples, and decision tools have to be evaluated for actual comprehension rather than for document availability alone.

Actuarial and social tradeoffs

A complete welfare assessment has at least four dimensions: ex-ante insurance value against catastrophic shocks; ex-post clinical and financial access; efficiency in the level and composition of care; and distribution of costs across persons and fiscal institutions. A schedule that lowers premiums by shifting small but frequent payments can be attractive to members with readily available savings. The same schedule can create care delays among people with high chronic need. Pooled savings may also diminish when costly hospitalization follows reduced outpatient care, or when reimbursement shifts to public programs.

The policy question is a constrained allocation problem: how to finance high-value care, preserve risk protection, moderate lower-value utilization, and sustain a viable pool with understandable terms. Empirical evidence informs the tradeoffs; a universal optimum cannot be extracted from a single elasticity or spending average.

Data, Measurement, and Validation

Reconstructing the eligible benefit transaction

A scientifically auditable dataset connects each service or drug line to the member, date, covered benefit, network and provider, clinical category, billed price, allowed price, deductible charge, copayment, coinsurance, plan liability, third-party contribution, and out-of-pocket accumulator. It also records plan versions, enrollment spans, family relationship, insurance year, and benefit phase. Without these data, a change in insurer-paid claims is ambiguous between utilization reduction and financial reassignment.

Paid claims alone are incomplete. They may omit all cash-paid services, denied claims, noncovered care, pharmacy coupon interactions, informal out-of-pocket spending, and services outside a reporting institution. The correct denominator for a plan-level analysis differs from a population-level analysis. Provider invoices, claims encounters, EHR records, patient surveys, and financial hardship surveys observe different portions of the process and must be reconciled carefully.

The OECD extent-of-coverage measure separates population entitlement, services included in benefit packages, and share of expenditures covered. Its 2025 report describes differences across hospital, outpatient, and pharmaceutical spending and identifies the coverage comparison as contingent on financing scheme and service definition. A national coverage percentage must not be interpreted as an individual plan's actuarial value or as an elasticity estimate.

Cost-sharing variables

An empirical protocol should record nominal and inflation-adjusted deductibles; deductible-to-income ratios; remaining accumulator balances before every service; fixed and percentage copayments; benefit-specific caps; preventive exemptions; person/family out-of-pocket ceilings; network tiers; and annual premiums. The relevant treatment price may be the contemporaneous allowed-charge liability, the expected end-of-year marginal price, or a patient-reported perceived cost. Mixing them without distinction biases conclusions about responses to nonlinear contracts.

Provider prices should be decomposed into allowed prices and quantities. Service mix should be stratified by clinical indication, urgency, chronicity, and expected net benefit. For pharmaceuticals, track days supplied, medication possession, adherence, substitutions, discontinuation, and the timing of refills. For hospitalization, report admissions, preventable admissions, length of stay, and downstream utilization, not only total inpatient spending.

Estimands and comparison groups

A baseline policy evaluation might estimate the average causal effect of a copayment increase on the probability of a completed service, the amount paid directly by the member, and total allowed annual spending. These outcomes require distinct estimands and follow-up horizons. A treatment effect among people who already use care cannot be generalized to the entire insured population without modeling initiation. An instrumental-variable estimate may apply to compliers at a particular threshold rather than to all enrollees.

Pre-registering denominators is particularly valuable. Specify enrollment-month adjustment, censoring, family clustering, multiple-testing policy, and imputation of missing service records. Where policies change by age, drug tier, income or plan anniversary, regression discontinuity or difference-in-differences designs can be informative, but manipulation, concurrent changes, and regression to the mean must be assessed. Report statistical intervals and real monetary units alongside point estimates, with subgroup analyses identified as prespecified or exploratory.

Separate mechanisms: fewer visits, lower prices, payer shifts

For expenditure S=∑jpjqjS=\sum_j p_jq_j, a change in total spending can reflect a change in qq, a change in pp, and shifts in the service mix. A benefit redesign can also change only the division of SS between payer and patient. These are four analytically different outcomes. A claim that deductibles encourage price comparison needs evidence that provider choice and prices actually changed; a claim of aggregate savings needs total allowed expenditure measured across relevant payers; a claim of welfare improvement needs health and access information.

An insurer may gain from reducing outpatient reimbursements while another insurer, employer, family, or public payer bears offsetting costs. The hospitalization-offset evidence supplies a concrete warning. Research spanning provider sectors and institutional budgets is required to distinguish social cost savings from accounting transfer.

Robustness and bias controls

The dossier recommends the following as established research practice, not a new proprietary standard: compare results under alternative allowed-price definitions; inspect changes before the policy implementation date; test negative-control services when appropriate; evaluate the distribution rather than only the mean; reconstruct amounts near deductible and annual cap thresholds; check coding changes; and separate incomplete enrollment from genuine zero utilization. Sensitivity analyses should vary the assumed behavioral price, income measurement, and missing claims correction.

Claims with impossible negative member liability, payer plus member payments exceeding the allowed amount without a documented adjustment, accumulator reductions without reversal codes, or out-of-pocket charges after a qualifying cap should trigger review. Exceptions for coordination of benefits, refunds, corrections, and multiple payers must be explicitly modeled rather than discarded as anomalous data.

Source hierarchy and limits

Legal obligations require the law and current regulatory text for a named jurisdiction and period. Benefit arithmetic requires an actual schedule of benefits. Causal effects require identification strategies and supporting empirical data. International trends require harmonized statistics with known denominators and survey coverage. Reviews and institutional summaries help frame the literature but cannot replace original research where available. This dossier uses publisher abstracts, official explanatory pages, and original research records; it does not claim independent reproduction of individual-level datasets or clinical adjudication.

International Synthesis and Scientific Conclusions

United States: contracts, statutory ceilings, and pharmacy phases

U.S. private plan design combines premiums, network-negotiated allowed charges, varied deductible structures, service copays and coinsurance, and regulated in-network out-of-pocket limits. The 2026 KFF employer survey reports a mean coinsurance rate of 20% for covered worker hospital admissions among plans using coinsurance and emphasizes plan variation. This is a statistic for covered workers in a specified survey and benefit class, not an American universal coinsurance rule.

The Marketplace's cost-sharing reductions illustrate how an income-based subsidy changes the effective point-of-care terms while leaving the insurance category recognizable. Separate federal schemes use entirely different benefit structures. CMS's 2026 Medicare Part D instructions establish a defined-standard $615 deductible, 25% coinsurance in the initial phase, a $2,100 annual out-of-pocket threshold, and zero beneficiary cost sharing for covered Part D drugs in the catastrophic phase, subject to the statutory structure and plan variations. These parameters must be cited as Medicare Part D 2026, never as a generic Marketplace plan.

Netherlands: deductible and contribution as distinct rules

In the Dutch statutory basic-insurance setting, the national government's explanation specifies a mandatory annual excess of €385 for adults for relevant standard-package care and explains that general practitioner visits do not fall under that excess. It also distinguishes the excess from separately payable patient contributions for certain goods and services. The case illustrates how an annual deductible can coexist with service-specific exceptions and charges that follow different legal logic.

England: item charges and exemptions

England's NHS prescription arrangements use a fixed item charge rather than a general insurance deductible on all care. The NHS prescription guidance lists the current charge at £9.90 per applicable item and explains that some products and eligible categories are exempt or free. A prescription-prepayment certificate may limit repeated charges for qualifying holders. As the NHS and territorial charging arrangements differ, this is a specific England prescription rule and does not describe all care throughout the United Kingdom.

Australia: pharmaceutical charges and safety nets

Australia's Pharmaceutical Benefits Scheme features per-prescription patient contributions and an additional safety-net mechanism. The Services Australia 2026 thresholds report a general-patient safety-net threshold of AUD 1,748.20 and a concession threshold of AUD 277.20, with maximum ordinary contributions before the threshold of AUD 25.00 and AUD 7.70 respectively. The arrangement demonstrates a benefit-specific combination of fixed charges, differentiated income/social eligibility, and cumulative protection; the PBS safety net is not an all-service annual medical out-of-pocket cap.

Singapore: deductible and coinsurance against large bills

Singapore's MediShield Life description describes national lifetime insurance protection for large hospital bills and selected costly outpatient treatment, designed around subsidized public-hospital settings. The relationship between coverage limits, patient financing, and the setting of care matters when an individual selects higher-class or private treatment. Its risk-pooling model should be compared using actual eligibility, subsidy, and benefit rules rather than mapped onto U.S. individual-market terminology.

Comparative interpretation

These cases demonstrate several different uses of direct charges: a broad first-loss deductible, a fixed medicine charge, a percentage share of covered claims, a program-specific annual safety net, and a tiered coverage phase. The label “copayment” may refer to quite different underlying patient obligations. A cross-national table that reports only deductible size, without including exemptions, covered service basket, annual limits, subsidies, financing contributions, and expected medical prices, can reverse the substantive comparison.

For policy research, the meaningful comparative question is the effective distribution of timely access and realized out-of-pocket spending for comparable patients with comparable needs, conditional on actual entitlement and total financing. The World Health Organization's European financial protection evidence and the OECD's coverage comparison jointly demonstrate why legal entitlement, benefit comprehensiveness, and household financial protection must be observed separately.

Integrated analytical findings

First, benefit design is a piecewise transfer rule whose marginal charge depends on eligibility, allowed price, accumulation history, clinical exceptions, and a ceiling. A quoted deductible is one parameter of this rule.

Second, cost sharing generates measurable utilization effects, but expenditure reductions may originate in foregone contact, delayed essential medication, lower prices, or mere payer reassignment. Evidence can distinguish these mechanisms when all-payer allowed spending and clinical utilization are observed.

Third, the insurance value of a schedule depends on both annual tail protection and the liquidity needed to obtain care before the terminal cap. Exemptions and income-sensitive contributions alter that distribution without changing the underlying concept of insurance pooling.

Fourth, patient comprehension and provider behavior are causal intermediaries, not background noise. Real-world outcomes depend on information, referral networks, diagnosis, and the way the benefit is communicated.

Fifth, evaluations must match scale to evidence. A randomized trial of one pharmacy benefit establishes effects of that intervention; a national health-financing statistic describes a system; a legal schedule defines entitlements. Combining these into a single undifferentiated “impact of coinsurance” would erase the very institutional mechanisms under investigation.

Technical Appendix: Worked Benefit Calculations

Episode with an unmet deductible

For illustration, take an in-network covered episode with an allowed amount of 2,400 units; annual deductible 1,000; no previous deductible credit; 20% post-deductible coinsurance; and a remaining out-of-pocket allowance above the claim's initial liability. The patient bears 1,000 plus 20% of 1,400, totaling 1,280. The insurer bears 1,120. This computation assumes one eligible episode, complete deductible applicability, no fixed copay, and no additional riders.

Episode after a partly satisfied deductible

If 800 units had already been credited and a later service had an allowed price of 500, the remaining deductible would be 200. Liability would be 200 plus 20% of 300, or 260; the insurer would pay 240. If only 100 remained beneath the annual qualifying out-of-pocket maximum, member liability would be 100 and plan liability 400, provided all relevant amounts qualified toward the same maximum.

Fixed copay after the deductible

For a covered visit whose allowed amount is 180 and whose post-deductible copayment is 30, the member pays 30 and the plan pays 150, assuming the copay is the sole applicable member charge and the annual maximum has not altered the result. If the same contract has an unmet deductible applying fully to the visit, the result depends on its specific visit exception and coordination text; one cannot simply add 30 to the deductible obligation by assumption.

Annual comparison of two invented plans

Suppose Plan A has annual premium 3,600, deductible 500, coinsurance 20%, and annual covered out-of-pocket maximum 4,000. Plan B has annual premium 2,400, deductible 2,000, coinsurance 20%, and covered out-of-pocket maximum 6,000. For a hypothetical person with 1,000 of covered allowed annual spending, no copays or exemptions, and all claims in the same accumulator, Plan A's patient spending is 500 plus 20% of 500 = 600; total premium plus covered cost sharing is 4,200. Plan B's covered liability is 1,000; total is 3,400. For covered allowed spending of 20,000, Plan A would reach its 4,000 cap, so its total is 7,600. Plan B's uncapped deductible-plus-coinsurance liability is 2,000 plus 20% of 18,000 = 5,600, so its total is 8,000. The crossing of annual cost curves shows why plan choice depends on spending distribution and risk preference; these figures are invented numerical demonstrations, not market quotes or evidence of a universally superior plan.

Validation arithmetic

For every illustrative example, the sum of patient and payer covered liability equals the eligible allowed price and the member's accumulated qualifying cost sharing remains within the assumed cap. Tests should reproduce exact boundary cases at zero allowed cost, at the deductible threshold, at the out-of-pocket limit, and when the deductible and limit are exhausted in the same episode. For actual claims, contract-specific features and multiple payers require additional reconciliation.

Conclusion

The scientific object of health insurance deductible, copayment, and coinsurance design is the complete contractual distribution of eligible medical spending over time and across insured persons. A coherent description contains the covered benefit, allowed price, service-specific payment rule, deductible and out-of-pocket accumulators, premium, network and clinical exceptions, and the jurisdiction's statutory constraints. These terms jointly determine the patient's effective price at the point of care and the insurer's contingent obligation.

The causal literature establishes that changing the patient price can substantially change health care use. The RAND Health Insurance Experiment supplies randomized evidence of this response; high-deductible employer research shows that reduced utilization can dominate price shopping; the elderly retiree study identifies potential hospitalization offsets; and the post-infarction medication experiment shows why targeted reductions in cost sharing can matter for adherence, patient expenditure, and selected outcomes. These studies have different populations and treatments. Their synthesis supports the conclusion that the clinical value and distribution of forgone care are indispensable to economic assessment.

International systems employ different deductibles, fixed charges, percentage payments, subsidies, exemptions, and safety nets. Their comparability depends on the patient's enforceable entitlement and total financed benefit, not on surface similarities in terminology. Research and policy evaluation must report financial protection, timely access, price and quantity changes, risk-pool effects, and heterogeneous consequences across income and health status. The appropriate scientific conclusion is therefore a methodologically bounded one: cost sharing is a consequential instrument of insurance design whose effects emerge through a nonlinear contract, institutional rules, actual patient decisions, and the distribution of medical need. The original contribution of this monograph lies in systematically integrating these evidentiary levels and their implications, without claiming a newly discovered empirical law or instituting a new doctrine.

References