ECON 364: Health Insurance and the Role of the Government

Demand for Health Insurance

Demand for Health Insurance is principally derived from the uncertainty or randomness
with which illnesses befall individuals. Consequently, the derived demand for health
insurance is to protect the individual from the financial risk created due to the illness and
the consequent cost of care needed to return to health if possible. With a health insurance
market, individual’s best bet would be to self-insure through saving, or through a network
of family and friends that would assist in times of need. Further, the latter includes a
network that is capable of monitoring deviant behavior which the group may choose not
to insure against. Further, the highly informal nature ensures that nothing is verifiable to
the courts.

1. Why do we dislike Risky Events?

In general, economic agents are typically risk averse, in the sense that if given a
chance, they would rather do without a risky venture. This explains why agents
often pay insurance companies more than their average loss they may confront to
secure certainty.

We know that the marginal utility from consumption of private goods, and
medical goods, which gives rise to health, is positive. This in turn means that the
marginal utility from income is also increasing, but increasing at a decreasing rate
due to the fact of diminishing marginal utility. This means that if we were to plot
the relationship with respect to income, we would get a increasing but concave
curve. Another way of thinking about this is that given two income levels, the
lower with certainty, and the higher without, the individual would pick the lower
income with certainty.

2. Risk Aversion

Utility
C’
B
C’’
E(C)
C
C is the average outcome
derived from having B
A occur with some
probability (1-a), and A
with probability a.

Risk Premium

Income

The above diagram describes why the idea of diminishing marginal utility is
synonymous with risk aversion. Consider a gamble where on the low side, you’d
ECON 364: Health Insurance and the Role of the Government

get an income level of A, and a high winning the same as from B. The expected
winning is what is typically termed a convex combination or weighted average of
A and B, which I have arbitrarily denoted at point C. Note that although this level
of expected income yields a higher income with uncertainty, it has the same level
of utility as a situation where a lower level of income occurs with uncertainty.
Comparing to another situation where the individual could get the same level of
income with certainty at the same average income level, the individual would
obviously prefer the certainty situation. The horizontal distance between C and C”
is the risk premium, or the premium that an individual is willing to pay to avoid
risky outcomes. (Note that we obtain this outcome without developing any new
microeconomic theory.) The greater the risk aversion is, the greater the risk
premium.

Choice of Insurance Policy

Let C be the coinsurance rate of an insurance policy, than (1 - C)% of a patient’s medical
bills is paid for by the insurance company. Although an insurance policy is far more
complicated, this simplification would still allow us to glean some important insights.

As noted, the insurance contract obliges the insurer to pay (1 − C ) p m m , for the m units of
medical care that the patient obtains. However, as we have realized, with insurance
coverage, the effective price of health care reduces, and consequently you have the
problem of moral hazard. That is as C increases, m increases as well. However, this
behavior is not necessarily a deviant behavior, but a rational response to a reduction in
price. In order to understand the mechanics of deciding on the optimal level of
coinsurance, we must understand what happens with its changes. Some of the key issues
are the following;

1. The effects of insurance coverage on demand for health care feedbacks onto
demand for insurance since as demand for health care increases, the marginal
value falls.
2. The increased consumption leads to the patients consuming more health care (that
gives a lower marginal value, than it costs to provide it. This means then that
there is a fall in welfare for society as a whole. The problem is that health
insurance breaks the link for the patients between the price of health care, and the
costs of provision.
3. At the same time, necessarily, the provision of health insurance raises the welfare
of the patients eliminating the financial risks. Then in deciding the optimal level
of coinsurance, we have to balance the welfare gain versus the welfare loss in 2.
ECON 364: Health Insurance and the Role of the Government

Welfare loss from insurance.

Price without
insurance.

Price with
Excess consumption insurance.
due to insurance
Demand

* Read the specific example on page 326-327 of your text for a more detailed expression
of these welfare ideas.

Further, the expected “moral hazard” loss is in turn dependent on the elasticity of
demand. Intuitively, if the elasticity of demand were small, then the change in pricing
resulting from insurance purchased would mean that the welfare loss (due to moral
hazard) would be small, as compared to the situation when the elasticity of demand is
large. This then mean that the more price elastic is demand for medical services, the less
desirable it is to insure against the risk due to the welfare loss involved.

This leads to two testable predictions

1. The greater the variance or financial risk involved the greater the demand for
insurance.
2. The greater the elasticity of demand, the lower the demand for insurance.

It turns out that all of them hold empirical true, for example, most individuals tended to
purchase insurance against hospitalization which is the type of care that creates the
greatest financial risk. On the other hand, dental care has the lowest incidence of health
insurance on account of the high degree of price elasticity.

It should be noted that the price of insurance as we had noted before is not the premium
paid by the consumers, but it’s the markup above the expected benefit. To use the
notation we had above, the insurance premium is
R = (1 + L )(1 − C ) p m m
where L is the loading fee which insurance firms charges above the expected benefits so
as to cover the cost of risk bearing, and administration fees. Then the larger the loading
fees, the larger the coinsurance rates, C, that the consumer would select. If the insurers
were to include deductibles into the insurance contracts, then the larger the loading fees,
ECON 364: Health Insurance and the Role of the Government

the larger would the consumer pick for the deductible so as to reduce the cost of
insurance. This then reveals that the relationship between the demand for insurance in the
form of coinsurance, and deductibles is just as a typical demand, i.e. is negative with
respect to the loading fee/price of insurance.

The Supply of Insurance

The insurance firm performs essentially two tasks:
1. The processing of claims on insurance contracts. In that sense, the output of a
insurance firm are these payments, and the equipment it needs to perform these
duties such as Actuaries, Insurance Agents, Clerical Staff, etc (And including
equipment such as computers, and office space). The administrative costs that are
to be covered by the loading fees are meant to cover these costs at the least. Just
as there are several types of Hospitals, there are several types of insurance firms,
a. For Profit (Prudential, Manulife etc) and Not-for-Profit (Blue Cross and Blue
Shield). The key difference between the two are that;
a. Not-for-Profit (Blue Plan) firms are exempt from taxation.
b. Exemption from many State regulations such as minimum cash reserves
regulations.
The principal reason why the Not-for-Profit Insurance firms has not flourished
principally is because,
c. Organized on a State or Sub-State level that prevents them from enjoying
or reaping economies of scale (especially where risks within a state are
rather correlated).
d. They also tended to insure only against low risk events.
The above two features then prevent the Blue Plan firms from capturing the
market.
2. Bearing and spreading of risk. This is reliant on the idea that if risks borne
individually are uncorrelated (risk of earthquakes along the inhabitants of the San
Andreas Fault is a correlated risk. One gets it, all gets it.) It should be noted that
although with insurance, average risks falls with pooling done by Insurance firms,
as a society, the transfer of individual risk to an insurance firm does not reduce
the risk in/to our society.

Is the Insurance Market Stable? : The Question of Self-Selection

The basic problem remains in insurance; that causes us to wonder about the stability of
the market. That is the inherent problem that the consumer know more about their own
risk of illness than do the insurance firms; i.e. the problem of asymmetric information.
We have already argued that this may lead to from of adverse selection where individuals
with poor risk end up signing up for the policies.

A Simple Model
A way to “force” individuals to identify their true tendency towards usage is to offer set
plans. Then their selection of the plans would indicate their “type” to the insurance firm.
If such plans could be found, insurance firms would be able to avoid the problem of
ECON 364: Health Insurance and the Role of the Government

adverse selection. However, on the down side, the insurance firm in doing so denies the
healthy individuals from buying an insurance plan they would prefer.

We can think of the budget constraint for an individual choosing between the optimal
level of private consumption good and health insurance as a negatively sloped concave
shaped curve since it is more costly in terms of consumption good lost as an individual
increase his purchase of health insurance.

1-C

Then equilibrium occurs at the point where the individual’s utility is tangent to her
budget constraint. In a perfect world where the insurance company is aware of everyone’s
proclivities, they would know that the budget constraint of a healthy individual has a
larger budget set than that of an individual with ill health since in such a perfect world,
the insurer knows that the probability of ill health is lower for the healthy individual, and
consequently the price charged is lower.
ECON 364: Health Insurance and the Role of the Government

Utility of
X Unhealthy
Individual

Utility of
Healthy
Individual

1-C

This then implies that in equilibrium, the consumption of the healthy individual would
also be higher in all respects.

However, reality does not commiserate with that analogy, but rather is parallel to one of
asymmetric information. If both the above insurance schemes were offered to everyone
under that scenario, if would only be rational that the ill health individual try to pass off
as someone healthy, and consume more of insurance and consumption goods. This then
leads to a higher expected payout for the insurance firm.

To force the high risk individuals to reveal themselves then the insurance firm deny the
portion of the insurance that would make the healthy individuals better off to everyone,
thereby forcing the high risk individuals to reveal themselves.
Utility of
X Unhealthy
Individual

Utility of
Healthy
Individual

Budget
Budget constraint of Constraint of
someone unhealthy someone healthy

1-C
ECON 364: Health Insurance and the Role of the Government

This is reflected in the above diagram as the bold line. In that scenario, the high risk
individual is ambivalent between the policy offered to herself, and that offer to someone
healthier. This however reduces the set of insurance that a healthy individual could
purchase, and make her worse off. This is known as a Separating Equilibrium.

Other Possible Pareto Improvement Schemes

An alternative solution to the problem of asymmetric information is to force insurance
companies to offer the same rates to everyone in the community, called “community
rating”. Such a scheme causes everyone’s budget constraint to be the same, and creates a
budget constraint that is between the two above. The exact location of which is dependent
on whether it is individuals with ill health, or healthy individuals that dominate. The
greater the proportion of healthy individuals, the greater the proximity of the “community
rated” budget constraint is to the healthy individual’s. This is a pareto improving scheme
since everyone is made better off compared to the separating equilibrium, but it also
means that the healthy are paying for the purchase of insurance for the unhealthy.

Yet another possibility is through insurance firms selling insurance policies directly to
employers since the reason for an individual joining a particular firm are for purposes
(such as high pay, matching of skills desired, etc.) other than to purchase insurance,
thereby averting the consideration of self-selection/adverse selection. Virtually, all
employers have a “community rating” scheme within a firm. However, if within a
society, the problem rests in coverage of long term care, such insurance plans would fail
to solve the problem for individuals dominant in that sub-market, the elderly. In
summary, group insurance 1. provides for economies of scale (since it eliminates the
“sale of insurance” individually, negates the collection of individual health histories,
etc.), and 2. avoids adverse selection.

Empirical Estimates of Demand for Insurance

The estimation of demand for insurance has two approaches:

1. By examining choices made by individuals and groups, since the differences in
income allow estimation of how demand varies across income groups.
2. By examining aggregated data over time, and estimating how total insurance
premiums respond to changes in income, loading fees, etc.

The summary of these finding are as follows:

1. Income elasticities has been found to be positive, and generally small (less than 1)
when using individual level data. Using aggregate data, the findings however
yield measures between 1 and 2 (closer to 2).
2. Price elasticities found using both types of data has yielded values between -1 to -
2. However, studies using differences in the marginal tax across households to
determine the effect of price on insurance has yielded much smaller estimates of -
0.2.
ECON 364: Health Insurance and the Role of the Government

General Considerations for a National Health Policy

1. Should such a system have universal coverage, and if so, how should it be
achieved?
2. How would such a coverage policy be financed?
3. What should be the scope and cost sharing scheme in such a policy?
4. How would expenditure be controlled?
5. How would new technologies be introduced?
6. How would quality of care be maintained?

Issues regarding Universal Coverage

Why should we have universal coverage?

1. Medical Care is a “merit good”, whose consumption by fellow citizens yield
increases in ones utility. If true, this would make the societal demand far larger
than individual private demand for medical good and service.
2. Individuals without insurance free ride of individuals who do have through their
inability to acquire it. It is claimed that eliminating private insurance and
instituting universal insurance eliminates this problem. Do you think universal
insurance really eliminates free riding? Consider universal unemployment
insurance; does it eliminate free riding?
3. Market failures due to asymmetry in information. Universal insurance such as
“community rating” eliminates this problem by providing a solution that is pareto
welfare enhancing. How can we really think about or measure the loss of
wealth for individuals of good health really?

How can Universal Insurance be achieved?

1. Government provision of Health Insurance (Canada), or Health Care provision
(United Kingdom).
2. Legislate mandatory purchase of health insurance, and subsidizing needy
individuals and households in this endeavor (Germany and Japan)
3. Legislate mandatory purchase of group health insurance by employers (as
proposed in the United States).

How to Finance Universal Health Insurance?

1. Taxation
2. Subsidies

However these financing schemes necessarily have a redistributive effect

dependent on who bears the greatest burden in the tax increase (read 553-559 of
your text). Further, insofar as the taxes are on either individuals and/or firms, this
means that there will be distortions created in individual choices. Consider an
increase in taxes to finance universal insurance. We know that labor supply would
fall if the income effect dominates the substitution effect, thereby reducing an
economies production, and hampering the society’s ability to sustain the
provision.
ECON 364: Health Insurance and the Role of the Government

Features of the Canadian Health Care System and its Consequences

Some Stylized Facts
1. Prior to 1971, Canada spent 7.4% if its GNP on healthcare compared with 7.6 in
the United States.
2. Prior to 1971, the most common found of insurance in Canada are group
insurance provided for by employers, with some government coverage. Doctors
and Hospitals operated as private entities then, and where Doctor’s fees are
determined by the market, while that of the hospital was through controls and
bargaining with the government.
3. In 1971 Canada instituted a universal Medicare System, which then caused a
deviation between the two countries. Due to a strong and comprehensive control
on spending growth, and its share of GNP dedicated to healthcare, Canada today
only dedicates 10% of its GNP on healthcare, compared to 15% in the United
States.
4. Financed by Value-Added-Tax (VAT) (neutral in redistributive effect) and
Income taxes (progressive in redistributive effect).
5. Scope of benefits includes almost all standard services and charges for care are
small.
6. Hospitals are paid negotiated budgets, and doctors are paid on the basis of fees
negotiated between the government and medical societies.
7. Capacity controls exists that constrains the provision of care, limits cost increases
through time, and also provides for filters on the introduction of new
technologies.

How does it work?

1. Insurance is Universal in Canada today, and the plan is organized around its
provinces (i.e. provided for by the Provincial governments), where each province
has its own Medicare system. Although there are idiosyncrasies across provinces,
common ground is imposed by the federal government on the provinces. With this
private insurance has disappeared.
2. Hospitals came under the authority of the provincial government, with a budget
cap established by the government for all hospitals. Within that cap, every
provincial hospital receives a direct budget.
3. Physicians receive a fee according to a negotiated schedule within the province,
but continue to function as independent firms.

What is it like today?

1. Health care cost growth in Canada is this an entirely political decision, with no
role for the markets to set prices. Since 1971 when US saw an increase of
proportion of GNP received by physicians by 40%, Canada saw only 10%.
2. Hospital spending has also increased at a lower rate than in the US.
3. Admission rates in Canada and US has been the same, which means that the cost
cuts does not seem to have dampened quantity.
4. US length of stay in hospital is lower than that of Canada’s.
5. Doctor office visits have remained the same in both countries.
ECON 364: Health Insurance and the Role of the Government

6. Because funding comes from the Government, the rate at which the Canadian
system has adopted technology has been notably slower than in the US. This
hence explains a part of the cost differential between the two countries.
7. A substantial proportion of consumers in Canada also seek health treatments
across the border to bypass the waiting period associated with the Canadian
system. This is evident from the thriving health markets catering to Canadian
demand in cities that are close to the Canada-US border.