Can we design a market for competitive health insurance? CHERE Discussion Paper No 53



CAN WE DESIGN A MARKET FOR COMPETITIVE HEALTH INSURANCE?

THE EVIDENCE ON RISK ADJUSTMENT

In a competitive insurance market, insurers would discriminate on the basis of risk and premiums
would be set accordingly. Risk adjustment would be the concern of the insurer. Low risks would face
low premiums; high risks would face high premiums and may be priced out of the market. However,
under managed competition government will subsidise high risk premiums so that the out-of-
pocket premium consumers face is more related to their choice of insurance coverage rather than
their risk class. Therefore government must provide each individual with a risk adjusted voucher or
entitlement. The appropriate risk adjustment mechanism is absolutely essential to the operation of
managed competition. According to van de Ven and Ellis: (van de Ven, Ellis 2000)

Without adequate risk adjustment it is hard, if not impossible, to achieve both efficiency and
fairness objectives in a competitive health plan market.

If there were a simple relationship between health status and the use of health care, then for
any given health state, there would be a corresponding use of health care services all of which
would be necessary and appropriate. There would be no need for risk adjustment as all observed
health care use would be appropriate, and the funder could reimburse all service use. However
the large literature on medical practice variations and the under or over-utilisation of health care
demonstrates that this is not the case.

There could still be variations in observed expenditure even with no differences in observed use
due to differences in the cost of the factors of production, or in providers’ capacity to secure rents.
Providers can influence costs through influencing use also; as consumers are typically poorly
informed, there are principal-agent problems here. To some extent, there will be differences in
the demand for health care, driven by socio-economic factors such as consumer preferences, and
social circumstances. As an aside, the role of the family and other social networks in the production
of health and health care is rarely acknowledged and less understood. Further, under any form of
insurance, moral hazard comes into play so that observed use is influenced by insurance coverage.
Underlying all of these factors, consumption of health care is valued for not its direct contribution
to consumer wellbeing but for its contribution to improved health so that age, sex and health status
are important explanatory factors in health care utilisation. Finally, the occurrence of illness and
accidents is largely random at the individual level. The result is that health care expenditures are
characterised by large variation, both predictable and random (van de Ven, Ellis 2000).

Random variation at the individual level is addressed by risk pooling. The efficiency gain for
competitive insurance requires insurers to compete on price, quality and service coverage, rather
than seeking an advantage by selecting good risks. The role of risk adjustment is not to explain
perfectly all the observed variation in health care expenditure at the individual level but rather to
eliminate the potential for cream skimming. The issue for the predictable component of expenditure
is whether the insurer can predict risk category better than the government rate setting process.
Insurers may be able to do so, if they hold more detailed information than the government, or if
they can induce the consumer to reveal their risk status indirectly. Therefore, the feasibility of the
data requirements, what is practical to collect from consumers and insurers, of any risk adjustment
technique must be considered. An associated issue is the incentives inherent in reporting; where
some diagnostic category or previous health service use is taken into account, there are incentives
to upgrade the reporting and/or usage to more profitable categories. In addition, in regulating what
types of insurance contract can be offered the opportunity for insurers to gain risk information
should also be taken into account. Good risk adjustment will ensure that the costs of cream
skimming exceed its profitability (van de Ven, Ellis 2000).

Further, the function of risk adjustment is normative, that is, it is required to meet the social
solidarity objectives of health insurance, then there should, conceptually, be some definition of what
the appropriate or acceptable costs are that should be covered. To some extent, defining what is
acceptable can be approached by limiting what is included in the benefits package; for example,
cosmetic procedures, or more contemporary examples in our setting are elective circumcision, or
gym membership. But conceptually, defining the acceptable requires specification not just of the
service but who gets it and under what circumstances.



More intriguing information

1. Structural Conservation Practices in U.S. Corn Production: Evidence on Environmental Stewardship by Program Participants and Non-Participants
2. Improving the Impact of Market Reform on Agricultural Productivity in Africa: How Institutional Design Makes a Difference
3. Searching Threshold Inflation for India
4. Connectionism, Analogicity and Mental Content
5. ESTIMATION OF EFFICIENT REGRESSION MODELS FOR APPLIED AGRICULTURAL ECONOMICS RESEARCH
6. The name is absent
7. EMU: some unanswered questions
8. Strategic Policy Options to Improve Irrigation Water Allocation Efficiency: Analysis on Egypt and Morocco
9. Evaluating the Success of the School Commodity Food Program
10. Federal Tax-Transfer Policy and Intergovernmental Pre-Commitment