Health Insurance and Risk

Introduction and Context

Health care is characterized by uncertainty. Individuals do not know whether they will become ill, when illness will occur, or how expensive treatment will be. A single hospitalization, cancer diagnosis, premature birth, or chronic disease episode can generate costs far beyond a household’s liquid resources. Health insurance is an institutional response to this uncertainty: many individuals make relatively predictable contributions so that the costs of the few who experience high losses can be financed collectively.

Insurance is not merely a financing mechanism. It changes prices faced by patients and providers, affects utilization, influences provider competition, redistributes income and health risks, and creates governance challenges. Its design therefore requires simultaneous attention to economics, actuarial feasibility, equity, clinical quality, and administrative capacity.

The central economic problem is to provide protection against uncertain financial loss while limiting inefficient changes in behaviour. Insurance can improve welfare through risk pooling, but it can also produce adverse selection, moral hazard, excessive administrative costs, and strategic market segmentation. Public and private insurance arrangements address these problems in different ways.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Explain uncertainty, risk aversion, expected utility, and willingness to pay for insurance.
2. Describe how risk pooling converts individual uncertainty into more predictable aggregate expenditure.
3. Distinguish adverse selection, moral hazard, cream skimming, and risk adjustment.
4. Analyse deductibles, copayments, coinsurance, out-of-pocket maximums, exclusions, and benefit limits.
5. Compare public insurance, social insurance, employer-sponsored insurance, and individual private insurance.
6. Evaluate managed care and provider payment as instruments for controlling expenditure and improving value.
7. Apply economic reasoning to benefit design, regulation, and health insurance management.

1. Risk, Uncertainty, and Insurance

Risk exists when possible outcomes and their probabilities can be described, at least approximately. Uncertainty is a broader condition in which probabilities may be unknown or difficult to estimate. Insurers convert uncertain individual losses into calculable expected costs by combining information about a sufficiently large population.

Let a person face a monetary loss L with probability p and no loss with probability 1 − p. Without insurance, expected loss is:

Here, E[L] is expected monetary loss, p is the probability of the adverse event, and L is the loss if it occurs. The formula assumes two mutually exclusive states and monetary loss measured in a common currency. In practice, losses may include medical expenditure, lost income, transportation, informal caregiving, and non-monetary effects that are difficult to price.

Insurance pays an indemnity I when the event occurs. If the premium is P, the individual’s final wealth in the loss state is W − P − L + I, while wealth in the no-loss state is W − P. Full insurance sets I = L; partial insurance sets 0 < I < L. A risk-neutral individual evaluates insurance by expected monetary value. A risk-averse individual also values the reduction in uncertainty.

Risk aversion and expected utility

Expected utility theory represents preferences over uncertain wealth using a utility function U(W). Expected utility is:

where pᵢ is the probability of state i, Wᵢ is wealth in that state, and Σᵢ pᵢ = 1. The framework assumes that decision-makers can rank risky prospects consistently and that probabilities are meaningful.

A risk-averse person has a concave utility function, represented by U″(W) < 0. Because of concavity, the utility of certain wealth exceeds the expected utility of a gamble with the same expected wealth:

The difference between certain wealth and the maximum premium a person would pay to avoid a risk is called the risk premium. If initial wealth is W and a gamble produces final wealth Wᵢ, the certainty equivalent CE satisfies:

The risk premium is:

For a risk-neutral person, utility is linear and RP = 0. For a risk-averse person, RP is positive. Risk aversion explains demand for insurance even when the actuarially fair premium equals expected loss.

Actuarially fair and loaded premiums

An actuarially fair premium equals expected claims:

An actual premium usually includes administrative expenses, capital costs, taxes, fraud losses, reinsurance costs, and a margin for uncertainty. A simple loaded-premium expression is:

where λ is a proportional loading factor and F is a fixed administrative amount. The formula is a simplified pricing representation, not a universal regulatory formula. It is appropriate when loadings can be separated into proportional and fixed components.

A risk-averse person may purchase insurance at P > P_f because the certainty gained is valuable. However, if the premium becomes too high relative to expected benefits, coverage may be rejected, especially by low-risk individuals.

2. Risk Pooling and the Logic of Insurance

Risk pooling combines individuals whose health events are imperfectly correlated. When risks are sufficiently independent, total expenditure becomes more predictable relative to the size of the pool. This is the law of large numbers: random variation in average claims tends to decline as the number of insured people increases, although systematic shocks such as pandemics can affect many members simultaneously.

Suppose n individuals each have an independent claim cost Xᵢ with mean μ and variance σ². Aggregate claims are S = ΣᵢXᵢ. Then:

and, under independence,

The standard deviation of aggregate claims is σ√n, while expected claims increase as n. Thus the coefficient of variation is:

This declines as n increases. Independence is a strong assumption; correlated claims reduce the stabilizing effect of pooling. Geographic concentration, common environmental exposures, epidemics, and similar provider-price shocks can create correlation.

Pooling can be organized through taxes, mandatory social insurance contributions, employer groups, community-rated premiums, or voluntary private contracts. The distributional result depends on who contributes, how premiums are calculated, whether government subsidies are used, and how benefits are defined.

Public and private insurance

Public insurance commonly pursues multiple objectives: financial protection, redistribution, universal access, public health, and correction of market failures. Financing may come from general taxation, payroll contributions, premiums, or a combination. Eligibility can be universal, categorical, income-based, or linked to employment or age.

Private insurance generally relies more heavily on contractual terms, underwriting, premiums, networks, and competition among plans. It may supplement public coverage, replace it for selected populations, or cover services outside a public benefits package. The distinction is not absolute: public systems may contract with private insurers, and private insurers may administer publicly financed programmes.

3. Information Problems in Insurance Markets

Adverse selection

Adverse selection occurs before coverage is purchased. Individuals know more about their expected health risks than insurers, or insurers cannot observe all relevant risk factors. People expecting high medical expenditure are more likely to buy generous coverage, while low-risk individuals may remain uninsured if premiums reflect the average risk of the pool.

A simplified pooled premium is:

where q_H is the proportion of high-risk individuals, C_H is their expected claim cost, and C_L is the expected claim cost of low-risk individuals. If low-risk individuals exit, q_H rises and the premium increases further. This can create a “death spiral” in voluntary markets.

Policy responses include mandatory participation, subsidies for low-income or low-risk households, open-enrolment rules, community rating, standardized benefits, waiting-period restrictions, and risk adjustment. Each has trade-offs. Mandatory coverage improves pooling but may require enforcement and subsidies. Community rating promotes solidarity but can intensify selection unless participation is broad. Risk adjustment compensates plans that enroll higher-cost members, reducing incentives to avoid them.

Moral hazard

Moral hazard occurs after insurance changes incentives. A patient may consume more care because the marginal price at the point of use is below the social cost. A provider may also deliver more services when payment rewards volume or when the insurer bears much of the financial risk.

Moral hazard is not necessarily fraud or irrationality. If insurance lowers the price of care, increased utilization may be an efficient response when care has high health benefits. The policy question is whether additional utilization produces benefits greater than its social cost.

The patient’s out-of-pocket price can be represented by:

where C is covered expenditure, D is a deductible, and c is the coinsurance rate between 0 and 1. In a more complete contract, an out-of-pocket maximum M caps patient payments:

These expressions assume that services are covered, prices are measured consistently, and deductible and coinsurance rules apply in the stated order. Actual plans may use separate deductibles, copayments, exclusions, or benefit-specific limits.

Cream skimming and risk selection

Cream skimming is the attempt by an insurer or provider to attract profitable, relatively healthy members and deter costly members. It may involve selective marketing, convenient locations, narrow benefit interpretation, network design, or administrative barriers. Cream skimming differs from adverse selection: adverse selection describes the market outcome caused by unequal information, whereas cream skimming describes strategic behaviour by an organization.

Risk adjustment is a payment method that transfers funds toward plans predicted to have higher costs. A conceptual risk-adjustment transfer is:

where Tⱼ is the transfer to plan j, Rⱼ is predicted risk-adjusted expenditure for its members, and \bar{R} is the market-average predicted expenditure. In practice, models may use age, diagnoses, disability, prior utilization, and socioeconomic variables, subject to legal and ethical constraints. Risk adjustment is imperfect: coding intensity can increase recorded risk without improving health, and unmeasured needs remain.

4. Cost Sharing and Benefit Design

Benefit design determines which services are covered, at what price to the patient, under what clinical conditions, and with which providers. It therefore affects affordability, utilization, risk protection, and equity.

Main forms of cost sharing

• Deductible: the amount paid by the patient before insurance begins paying for specified services.
• Copayment: a fixed amount per service or visit.
• Coinsurance: a percentage of the allowed cost paid by the patient.
• Out-of-pocket maximum: the annual ceiling on eligible patient cost sharing.
• Balance billing: an additional charge when a provider bills above the insurer’s allowed amount, where legally permitted.
• Premium: payment for enrollment, generally owed regardless of whether services are used.

Cost sharing can reduce low-value utilization, but it may also deter high-value preventive care, medication adherence, and early treatment. Patients often cannot distinguish beneficial from unnecessary care, particularly under stress. Uniform cost sharing is therefore usually inferior to clinically informed benefit design.

Value-based insurance design lowers cost sharing for services with strong evidence of benefit and raises it for low-value services, while protecting vulnerable groups. Examples include zero or low copayments for essential chronic-disease medicines, vaccinations, prenatal care, and evidence-based preventive services.

Benefit design must also address exclusions, prior authorization, formularies, referral requirements, network tiers, coverage of mental health and long-term care, rehabilitation, digital health, and catastrophic protection. A low premium with severe exclusions may provide little effective protection. A useful assessment considers total expected financial exposure rather than premium alone.

5. Managed Care and Provider Incentives

Managed care seeks to coordinate services and influence utilization through networks, clinical rules, information systems, and payment contracts. Common instruments include primary-care gatekeeping, preferred-provider networks, utilization review, disease management, case management, prior authorization, clinical pathways, and negotiated prices.

Managed care can correct fragmented care, reduce duplication, and improve chronic disease management. However, it can also create administrative burdens, restrict patient choice, delay necessary treatment, and encourage under-service when providers bear excessive financial risk.

Provider payment strongly shapes behaviour:

• Fee-for-service: pays for each service; supports activity but can encourage volume and fragmentation.
• Capitation: pays a fixed amount per enrolled person for a defined period; encourages prevention and cost control but creates under-service risk.
• Case-based payment: pays a fixed amount for an episode or diagnosis-related group; encourages efficiency but may encourage early discharge or coding manipulation.
• Salary: pays for time or role; can support teamwork but may weaken activity incentives.
• Bundled payment: pays one amount for related services over an episode; promotes coordination but requires clear episode definitions.
• Pay-for-performance: links payment to indicators; may improve measured care but can produce gaming and neglect of unmeasured outcomes.

A sound contract aligns incentives with outcomes, adjusts for patient risk, monitors quality, provides appeals, and avoids placing providers at financial risk for factors they cannot control.

6. Welfare Analysis of Insurance

Insurance has benefits and costs. Benefits include risk protection, access to necessary treatment, improved health, reduced medical debt, and income smoothing. Costs include premiums, taxation, administration, moral hazard, provider-induced demand, and possible reductions in competition.

The welfare effect can be expressed conceptually as:

This is an analytical identity rather than a directly measurable universal equation. Each term must be defined for the policy being evaluated. The risk-protection gain may be estimated using willingness to pay or utility-based methods; health benefit may be measured in outcomes such as survival, quality-adjusted life-years, or functional status; distortion costs include inefficient utilization and taxation effects.

Insurance may also redistribute from healthy to sick, young to old, high-income to low-income, or low users to high users. Whether this redistribution is viewed as desirable is a normative question requiring explicit ethical justification rather than purely economic calculation.

7. Illustrative Worked Example: Choosing Coverage

Consider a household with wealth of 100,000 monetary units. It faces a 2% probability of a medical loss of 20,000. The actuarially fair premium is:

Suppose the household has utility U(W) = √W and is offered full insurance for a premium of 600. Without insurance, expected utility is:

Using √100,000 ≈ 316.23 and √80,000 ≈ 282.84:

With full insurance, wealth is certain at 99,400, so:

Under this specific utility function and these rounded assumptions, the household would not purchase the 600 premium because its expected utility is slightly lower than remaining uninsured. The maximum premium P* that makes the household indifferent satisfies:

Therefore:

The household’s willingness to pay is approximately 400.8, only slightly above the actuarially fair premium of 400. The example illustrates that risk aversion creates willingness to pay above expected loss, but the amount depends on the degree of risk, wealth, and curvature of utility. It also shows why a loading of 200 may make coverage unattractive to this household.

The example is deliberately simplified. It excludes taxes, incomplete coverage, medical benefits, insurer solvency, income loss, borrowing constraints, and the possibility that the household values health itself rather than only wealth.

8. Case Study: Designing Coverage for Diabetes Care

This is an illustrative case, not a report of a specific organization.

Background

A regional health insurer serves 120,000 adults. Hospital admissions for diabetes complications are increasing, medication adherence is poor, and members report that deductibles discourage regular monitoring. The insurer currently uses a uniform deductible and a 30% coinsurance rate for medicines and outpatient services.

Problem

The existing design protects against catastrophic hospitalization but creates substantial point-of-care costs for insulin, glucose monitoring, consultations, and laboratory testing. Low-income members delay care, while the insurer bears avoidable emergency and inpatient expenditure. At the same time, reducing all cost sharing could increase low-value utilization and threaten affordability.

Analysis

The insurer segments services according to expected clinical value. Evidence-based diabetes medicines, monitoring, retinal screening, renal testing, and structured education have high potential value for eligible patients. Non-urgent services with uncertain benefit are reviewed through ordinary cost-sharing and utilization-management rules. The insurer also observes that its claims data underrepresent social needs, so a model based only on diagnosis codes risks misclassifying members.

Approach

The insurer introduces a value-based benefit design:

1. Low or zero copayments for essential diabetes medicines and recommended monitoring.
2. A maximum monthly payment for insulin and essential supplies.
3. Nurse-led case management for members with repeated admissions.
4. Primary-care incentives based on appropriate monitoring, continuity, and avoidable-admission reduction.
5. Risk adjustment and socioeconomic monitoring to prevent providers from avoiding complex patients.
6. An evaluation using a comparison group, medication adherence, HbA1c testing, emergency visits, admissions, total expenditure, and patient-reported burden.

The insurer also establishes an appeals process for clinically justified exceptions and audits provider coding and quality reporting.

Outcome

After implementation, adherence and routine monitoring improve. Emergency admissions decline among enrolled high-risk members, although total pharmacy expenditure rises initially. The insurer does not claim success solely from lower admissions; it examines whether improved outcomes justify additional medicine spending and whether access improved equitably across income groups.

Learning points

The case demonstrates that benefit design is not simply a choice between generous and restrictive coverage. It is a structured allocation of financial incentives. Successful design combines clinical evidence, risk adjustment, patient protection, provider accountability, and evaluation. It also illustrates why short-term expenditure increases may be compatible with long-term value.

9. Regulation, Ethics, and Governance

Health insurance decisions have direct ethical consequences. Excluding high-risk people, denying necessary treatment, or imposing unaffordable cost sharing can undermine solidarity and access. Conversely, unrestricted benefits can threaten the sustainability of a programme and reduce resources available for other patients.

Important governance principles include:

• Transparency: members should understand premiums, exclusions, networks, authorization rules, and appeals.
• Non-discrimination: risk classification must not unfairly penalize protected groups or people with chronic illness.
• Accountability: coverage decisions should be reviewable, with clinically meaningful appeals.
• Confidentiality: claims and health data require strong privacy protections.
• Proportionality: restrictions should be related to legitimate clinical or financial objectives.
• Equity: access should not depend excessively on income, geography, disability, or bargaining power.
• Stewardship: managers should seek value, not merely lower expenditure.

Regulation commonly addresses solvency, licensing, consumer protection, minimum benefits, network adequacy, premium review, portability, disclosure, and data protection. Regulatory design must recognize trade-offs: strict standardization can improve comparability but reduce innovation; extensive mandates can improve protection but raise premiums; weak oversight can permit exclusion and market abuse.

10. Technology, Data, and Contemporary Developments

Digital claims systems, electronic health records, telemedicine, remote monitoring, and predictive analytics can improve underwriting, care management, fraud detection, and risk adjustment. Artificial intelligence may identify members at risk of hospitalization or support prior authorization. These tools require careful validation because historical data may reproduce unequal access and biased clinical decisions.

Managers should ask whether an algorithm is accurate across demographic groups, whether its variables are clinically and legally appropriate, whether patients can challenge decisions, and whether it creates incentives to code disease rather than improve care. Predictive accuracy does not establish causal effectiveness. A model that predicts high expenditure may not identify an intervention capable of reducing it.

Contemporary insurance development also includes pandemic risk, climate-related health events, aging populations, expensive gene and cell therapies, growth of digital therapeutics, mental-health parity, and expanding employer interest in direct contracting. These developments increase the importance of reinsurance, outcome-based contracting, actuarial scenario analysis, and explicit priority setting.

11. Managerial Framework for Assessing an Insurance Product

A health care manager evaluating an insurance product should proceed through six questions:

1. Population: Who is covered, and what is the distribution of health risks and ability to pay?
2. Coverage: Which services, medicines, providers, and settings are included or excluded?
3. Incentives: How do premiums, cost sharing, and provider payment affect behaviour?
4. Selection: Could the design attract healthy members or deter costly members?
5. Protection: What financial exposure remains after premiums and point-of-care payments?
6. Performance: How will access, quality, equity, expenditure, and member experience be measured?

Useful indicators include enrollment and retention by risk group, claims per member, avoidable admissions, preventive-service use, medication adherence, denial and appeal rates, network adequacy, waiting times, catastrophic expenditure, complaints, and outcome measures. Indicators should be interpreted together; a reduction in expenditure accompanied by lower access may represent failure rather than efficiency.

Challenges and Limitations

Insurance cannot eliminate the underlying scarcity of health care resources. It reallocates financial risk and may improve access, but physicians, medicines, hospitals, and public-health capacity remain limited. Expected-cost models can fail during epidemics, technological disruptions, or changes in disease prevalence. Risk adjustment can be manipulated through coding, while utilization controls can shift costs to patients or other payers.

Cost-effectiveness evidence may not capture distributional concerns, caregiver burden, or the value of hope. Consumer choice is limited by complexity, unequal information, and stress. Finally, insurance arrangements that appear actuarially sound may be politically or ethically unacceptable if they undermine solidarity or exclude vulnerable groups. Sound management therefore combines economic analysis with clinical evidence, law, ethics, and stakeholder participation.

Chapter Summary

Health insurance protects individuals against uncertain medical losses by pooling risks and transferring resources across states of health. Risk-averse people may value insurance even when the premium exceeds expected claims because insurance increases certainty. Expected utility provides a formal method for analysing this choice.

Large pools reduce relative claims variability, although correlated risks weaken pooling. Public and private insurance differ in financing, objectives, regulation, and risk-selection rules, but both must address information problems. Adverse selection occurs before enrollment; moral hazard occurs after coverage changes incentives; cream skimming is strategic selection by insurers or providers. Mandatory participation, subsidies, standardized benefits, risk adjustment, and regulation are important responses.

Deductibles, copayments, coinsurance, and out-of-pocket limits influence utilization and financial protection. Well-designed benefits use clinical evidence to protect high-value care while discouraging low-value services. Managed care and provider payment can coordinate care and control expenditure, but they require quality safeguards and risk adjustment. Insurance evaluation must consider welfare, equity, sustainability, access, outcomes, and governance. Data and AI can support management but require validation, transparency, privacy, and bias control.

Key Terms

Risk: Exposure to uncertain outcomes with meaningful consequences.

Uncertainty: A condition in which outcomes or their probabilities are incompletely known.

Risk aversion: Preference for a certain outcome over a risky outcome with the same expected monetary value.

Expected utility: Probability-weighted utility across possible states.

Risk premium: The maximum amount a person would pay above expected loss to avoid risk.

Actuarially fair premium: Premium equal to expected claims under stated assumptions.

Risk pooling: Combining many risks so aggregate costs become more predictable.

Adverse selection: Disproportionate enrollment of higher-risk individuals caused by information asymmetry.

Moral hazard: Behavioural change after insurance that affects utilization, prevention, or claims.

Cream skimming: Strategic attraction of profitable, lower-risk members or avoidance of costly members.

Risk adjustment: Transfers or pricing methods that account for differences in expected expenditure.

Deductible: Amount paid by the insured before specified coverage begins.

Copayment: Fixed patient payment for a covered service.

Coinsurance: Percentage of an allowed cost paid by the patient.

Out-of-pocket maximum: Contractual ceiling on eligible patient cost sharing.

Managed care: Organizational and contractual methods for coordinating care and influencing utilization.

Capitation: Fixed payment per enrolled person for a defined period and scope of services.

Value-based insurance design: Benefit design that aligns cost sharing with expected clinical value.

Risk adjustment: A method of modifying payments to reflect predicted differences in member health expenditure.

Network adequacy: Extent to which covered providers are sufficient, accessible, and appropriate for members’ needs.

Review Questions

Short Answer Questions

1. What is the difference between an actuarially fair premium and a market premium?
2. Why does risk pooling generally become more predictable as the pool grows?
3. Distinguish adverse selection from moral hazard.
4. Define a deductible, copayment, and coinsurance.
5. What is the purpose of risk adjustment?

Long Answer Questions

1. Explain expected utility theory and show how risk aversion creates demand for insurance.
2. Compare public/social insurance with private insurance in terms of objectives, financing, risk selection, and governance.
3. Analyse how cost sharing can reduce inefficient utilization but also harm access to high-value care.
4. Discuss the advantages and risks of managed care and alternative provider-payment methods.
5. Explain how cream skimming can undermine insurance-market equity and describe regulatory responses.

Analytical / Application Questions

1. A plan has a high deductible and low premium but substantial coinsurance after the deductible. Evaluate the product for a low-income household with chronic disease and a high-income household with low expected utilization.
2. An insurer observes that healthy members are leaving after premiums rise. Diagnose the likely market mechanism and recommend a combination of policy responses.
3. Design a benefit package for diabetes management that addresses moral hazard, financial protection, provider incentives, equity, and evaluation.

Multiple Choice Questions with Answers

Question 1

Which condition is represented by a concave utility function?

A. Risk neutrality

B. Risk aversion

C. Risk seeking

D. Perfect information

Correct Answer: B. Risk aversion

Question 2

Under independent risks, which expression represents the variance of aggregate claims for n individuals, each with variance σ²?

A. σ²/n

B. n/σ²

C. nσ²

D. n²σ²

Correct Answer: C. nσ²

Question 3

Adverse selection primarily occurs when:

A. Providers deliver excessive care after payment is received

B. Individuals have different information about their expected risks before enrollment

C. Insurers improve their claims-processing systems

D. Patients reach an out-of-pocket maximum

Correct Answer: B. Individuals have different information about their expected risks before enrollment

Question 4

Which policy most directly reduces incentives for insurers to avoid high-cost enrollees?

A. Risk adjustment

B. Higher deductibles for all members

C. Narrower provider networks only

D. Removing open enrollment

Correct Answer: A. Risk adjustment

Question 5

A 20% coinsurance rate on an allowed claim of 1,000 monetary units, ignoring deductibles and limits, requires the patient to pay:

A. 20

B. 200

C. 800

D. 1,000

Correct Answer: B. 200

Question 6

Which provider payment method pays a fixed amount per enrolled person for a defined period?

A. Fee-for-service

B. Salary

C. Capitation

D. Copayment

Correct Answer: C. Capitation

Question 7

Cream skimming refers to:

A. Pooling independent risks

B. Attracting relatively profitable members and avoiding costly ones

C. Reducing the price of preventive services

D. Paying claims after an event

Correct Answer: B. Attracting relatively profitable members and avoiding costly ones

Question 8

What is a principal concern with unrestricted cost sharing for essential chronic-disease medicines?

A. It always eliminates moral hazard

B. It may reduce adherence and increase later complications

C. It guarantees universal coverage

D. It prevents adverse selection completely

Correct Answer: B. It may reduce adherence and increase later complications

Question 9

Which feature provides catastrophic financial protection by limiting eligible patient payments?

A. Premium loading

B. Out-of-pocket maximum

C. Risk adjustment transfer

D. Provider salary

Correct Answer: B. Out-of-pocket maximum

Question 10

In the worked example, the actuarially fair premium for a 2% probability of a 20,000 monetary-unit loss is:

A. 40

B. 200

C. 400

D. 1,000

Correct Answer: C. 400