Sa4E9 - Health Insurance

Front 1

Outline – Short

Back 1

1. Motive for Health Insurance – Expected Utility Theory
2. Health insurance and moral hazard
3. Health insurance and income effects
4. Moral hazard revisited
5. Critcisms
6. Policy implications

Front 2

Outline – Long

Back 2

1. The Motive for Health Insurance – Expected Utility Theory (5)
2. Health Insurance and Moral Hazard
a. Arrow (1963)
b. Pauly (1968)
c. Cost-sharing – Feldstein (1973), Manning and Marquis (1996)
3. Health Insurance and Income Effects
a. EU theory & assumptions (2)
b. Income effects (4)
c. EU model – risk aversion & access motive
d. conclusion
4. Moral Hazard Revisited
a. Estimated income effects
b. Welfare gains from insurance
c. Welfare loss from insurance
5. Criticism
a. Blomqvist
b. heterogeneity of health services
c. SID
5. Policy implications
a. EU – risk aversion model
b. EU – risk aversion + income effects model
6. Conclusion

Front 3

The Motive for Health Insurance – Expected Utility Theory

Back 3

1. Minimize variation in income 2/2 medical expenditures
2. Individual level – large variation – unpredictable
3. Law of large numbers – small variation – predictable
4. Risk averse individual increases utility by exploiting the gains of trade by risk-pooling and financial-pooling to minimize risk of catastrophic health expenditure
5. P = pL+l – estimated empirically

Front 4

Health Insurance and Moral Hazard

Back 4

1. Arrow (1963) – as p → 1 (eg ger, poverty), extensive market failure for insurance; therefore the government should provide health insurance (Medicare, Medicaid)
2. Pauly (1968) – criticized addressing market failure with a failed institution – noted that subsequent to prepayment for health care created ex post asymmetry of information in the form of moral hazard – over consumption of health care
3. Cost-sharing – reintroducing a + MC of consumption returns the price signal limiting welfare loss; Feldstein (1973) & Manning and Marquis (1996) estimated, respectively, 66% and 50% coinsurance rates to minimize welfare loss

Front 5

Health Insurance and Income Effects – Short - 1

Back 5

1. EU theory assumptions (2)
2. Income effects (5)
3. Expected Utility Model of HI

Front 6

Health Insurance and Income Effects – Long - 1

Back 6

1. EU theory assumptions
a. assumption: loss (health expenditure) exogenously determined
- implies that HE s HI = HE c HI – price elasticity of care
b. assumption: income elasticity = 0
- RAND health insurance experiment (Manning & Marquis, 1987) estimate income elasticity of demand = 0.22
- Nyman therefore argues, the income effect of financial pooling motivates the demand for health insurance and changes levels of health expenditures after insurance

Front 7

Health Insurance and Income Effects – Long - 2

Back 7

2. Income effects
a. income transfer from healthy  sick
b. inversely proportional to probability of loss/illness; as p  0, no income transfer
c. access motive – income transfer grants access to otherwise unaffordable care
d. the access value of insurance is different than the risk avoidance value because, by definition, there is no financial risk for unaffordable health care because purchases cannot privately occur
e. As lending institutions will not lend with limited ability to repay and there may not be sufficient time or income to save privately, insurance may be seen as a viable institution for income transfer with welfare loss as transaction cost.

Front 8

Health Insurance and Income Effects – Long - 3

Back 8

3. Expected Utility Model of Health Insurance – Risk aversion and Access motive
a. access motive can be quantified by the gain in utility from access to M (income transfer) in medical expenditures in the event of illness at the cost of P (premium)
b. economists overlooked welfare gain from income transfer, resulting in an overestimation of moral hazard (welfare loss)

Front 9

Moral hazard revisited

Back 9

1. Moral hazard = pure price effect
2. Welfare gain from health insurance
3. Welfare loss from health insurance

Front 10

Moral hazard revisited – Long -1

Back 10

1. Moral hazard = pure price effect
a. Marshallian demand – uncompensated demand curve
b. pure price effect – substitution effect – change in consumption of medical care that would occur if an individual who was already ill (p=1) purchases a contract with a company to reduce the price of care paying an actuarially fair premium
c. income effect – can be eliminated when the probability of illness is one
 should not be included in welfare loss because access value provides welfare gains
d. Hicksian demand = compensated demand curve = estimates the pure subsitution effect by holding the income effect constant
e. employs Slutsky equation to approximate compensated price elasticity
E = n +ae
E pure price elasticity
n Marshalling demand
a share of household income in medical care
e income elasticity
f. estimates -.06  33% of prior moral hazard estimates = pure price effect

Front 11

Moral hazard revisited – Long -2

Back 11

2. Welfare gain from Insurance
a. conventional risk-bearing gain price- & income- inelastic
b. access value – gain from income transfers improves purchasing power providing access to previously unaffordable procedures
c. increased demand for health services from income elasticity of demand + income transfer  more efficient than pure price effect

Front 12

Moral hazard revisited – Long -3

Back 12

3. Welfare loss from insurance
- moral hazard from pure price effect
- could be considered a transaction cost for the welfare gains realized from risk-aversion and access motive

Front 13

Criticisms

Back 13

1. Blomqvist – overestimates income effect – poor estimates of
a. total health care spending: does not account for stop-loss clauses
b. elasticity of income: employs nursing home data > RAND estimate
c. point elasticity less efficient parameter than arc elasticity
d. Blomqvist estimation – 13% of previous welfare loss calculations from income effect
2. heterogeneity of health services – low cost services eg primary care and preventive care may not possess the same access value
3. supplier induced demand

Front 14

Policy conclusions

Back 14

1. MH  CS
2. SID  utilization reviews
3. oppose expansion of HI by gov in cases of market failure due to welfare loss
1. cost-sharing may decrease welfare gain and welfare loss; heterogeneity of services – low cost services may have a higher component of pure price effect – justifies cost-sharing
2. opposes utilization reviews insofar as may bar access to cost procedures negating access motive; address SID with assessing cost-effective treatment and not necessarily just high cost treatment
3. tax subsidy for health insurance premiums – welfare loss from pure price effect is a transaction costs for income transfers to ill