|        Polky Roewoh
WORKING PAPERS
Country OperatIons
Europe and Central Asia
Country Department I
The World Bank
September 1993
WPS 1181
Reforming Health Care
A Case for Stay-Well
Health Insurance
Zeljko Bogetic
and
Dennis Heffley
Many health care reform proposals expand insurance coverage
without fundamentally changing the structure of health insur-
ance. The stay-well plan used in Mendocino County, California,
since 1979, offers an alternative insurance structure that pro-
vides direct incentives for consumers to control utilization and
adopt healthier lifestyles.
PdbuyR sch WiingPape dinam newt  pfwik in p   S aWwwwngetlwuzdmnpofides aon danksaff and
aua _immud indevelpmatuumnoepapa, dibutod by teRecb Adviy Staff. cay nmn of beam, idlI
*niyw,wdldbsud abnd diwd a.oaodngly.Th.Bndingutqmatiaes, and c usiantbeautma own. hey iuld
nM1w w t W<ld Bmi. lu lame dD_imcn. lwanaunns eran di inmwm_i e



Pollay Rhwo
Country Operations
WPS 1181
This paper- a product of the Country Operations Division, Europe and Central Asia, Country Department
I - is part of an effort in the region to strengthen the emphasis on the analysis of social sector issues. Copies
of tle paper are availablM free from the World Bank, 1818 H Street NW, Washin.ton, DC 20433. Please
contact Faith Smith, room H5-245, extension 36072 (September 1993, 17 pages).
All countries - whether industrial, developing,   a general, flexible form of insurance that sub-
or in transition to a market economy - are        sumes most conventional plans as special cases.
interested in health care reform. A central focus  The rewards for low use might take many forms.
of reform everywhere is to make patients more     As in the Mendocino plarn, die rewards might be
responsive to health care costs without diluting  a credit to a retirement account, but they could
the protection offered by public or private       just as easily be annual cash rebates or credits
insurance.                                        against out-of-pocket expenses that exceed an
individual's or family's spending goal in a future
Conventional insurance offers customers       period.
little incentive to monitor their own use of health
care services or to adopt and maintain better         Administration of the stay-well plan appears
health habits.                                    not to be unduly complex. If anything, incorpo-
rating stay-well incentives in a single-payer or
Bogetic and Heffley describe an alternative   national health care system would be simpler
health insurance structure first adopted in       than incorporating them in a self-insured fund.
Mendocino County, California, in 1979, and        The success of the plan hinges on whether
compare it with conventional forms of insurance.    incentives shift the frequency distribution of
The Mendocino or "stay-well" plan offers          health care spending by reducing unnecessary
consumers direct incentives to control their use  utilization in the short-run and through better
of health care services and to adopt healthier    health care habits, reducing long-run costs.
lifestyles. How well this insurance can contain
health care costs depends on the size of the          Despite additional payments to low usrers,
incentives and consumer responsiveness to them.    the stay-well plan could be less expensive than
conventional plans with similar coverage. As in
Conditions in some developing countries and    any insurance plan, solvency is enhanced by
in many countries moving to market-based          larger groups, better risk-pooling, economies of
economies - overuse of services, poor health      scale in administration and clains processing,
habits, and declining real incomes - improves     and grater bargaining power with health care
the likelihood of a favorable rtsponse to such    providers.
incentives.
How to structure the stay-well system
depends on the country, but the stay-well plan is
The Policy Research Working Paper Series disseminatew the fndings of work under way in the Bank. Anobjectiveof the series
is to get these findings out quickly, even if presentations mre less than fully polished. The findings, interpretations, sno
conclusions in these papers do not necessarily represen. official Bank policy.
Produced by the Policy Research Dissemination Center



Reforming Health Care:
A Case for Stay-Well Health Insurance
v
by Zeljko Bogetic and Dennis Heffley*
Country Operations Division
Department I
Europe and Central Asia
Zeljko Bogetic is a country economist, Europe and Central Asia,
Country Department I, World Bank. Dennis Heffley is a professor of economics
at the University of Connecticut. The authors gratefully acknowledge thorough
and constructive comments from Alexander Preker, World Bank.



TABLE OF CONTENTS
Section 
1.  Background.                                           1 -3
II.  A Recurrent Question: How Can Patients Be Made More
Responsive to the Cost of Health Care Without Reducing
Insurance Protection?                                3 -5
III.  Properties of Conventional Health Insurance.         5 - 6
IV.  An Innovative Form of Health Insumnce: The Mendocino
or Stay-Well Plan.                                   6- 9
V.  Extending the Stay-Well Concept to a National Health
Care System.                                        9 - 12
VI.  Concluding Remarks.                                 12 - 13
References.                                        14 - 15
Figure 1: Conventional Insumance Plans.
Figure 2: The Stay-Well Plan.



REFORMING HEALTH CARE
7,eljko Bogetic and Dennis Heffleyl
I.  BACKGROUND.
'There is a pervasive feeling of imminent crisis with respect to the delivery of medical services
in the United States. Expenditures on medical care ae rising without any offsetting decline in our
mortality rates -- which compare poorly wvith tho " of many other countries. The costs of medical
caro am increasing at an accelerating rate. Many big-city hospitals, which service the poor, are in a
state of financial distress. Medicare and Medicaid, the programs developed to enable the
disad-antaged to obtain medical care, set off the cwrent price spiral by creating a iapid increase in
the demand for medical services while the supply of those ser-ices could be incmased only slowly.
The middle class now finds it difficult to obtain the kind of medical care it has conwe to expe:t,
namely that of an interested, warm physician who is familiar with the family history. There is a
perceived over-all sho.tage of nurses and physicians. Moreever, many concerned individuals believe
that the existing physicians are not distributed well. Th;y argue that there are relatively too few
primary physicians and too many specialists, too few physicians in the inner cities and in rural
areas, and too miany in the affluent suburbs.' (Lave and LAve, 1970; pp.253-254).
Much attention is being given to possible health care reforms in the UJnited States2 The
preceding description of problems, written more than two decades ago, sug-. sts that the problerms
of the health care system are not new. Nor are the problems unique to the United States. Efforts
to rethink and reform health care systems are widespread in other developed countries, Eastern
Europe, and less developed countries in general.3
1.   Zeljko Bogetic is a Country Economist, Europe and Central Asia Region, Country Department 1, The World
Bank. Dennis Heffley is a Professor of Economics at The University of Connecticut. The views expressed in this
paper are the authors' and should not be construed as official views of their respective institutions. We gratefully
acknowledge thorough and constructive comments from Alexander S. Preker, The World Bank.
2.   Health care reform has emerged as a critical element of the Clinton agenda, but even the previous
administration's economic advisors had come to view reform as an important issue. See Chapter 4 ('Tlhe Economics
of Health Care') of the January 1993 Economic Report of the President for a thorough overview of the issues and a
discussion of various reform proposals.
3.   Health has been selected as the primary focus of the 1993 World Dwvelopment Report, reflecting the concen
jf The World Bank ard other international agencies for improving health care in developing countries and ber
understanding the role of health in rtimulating economic development. In a survey of the link between health and
development and the contribution of health research, Birdsall (1990) identifies 'Analysis of altemnative financing



2
Escalating expenditures and lack of universal insurance are the focal points of American
health care reform, while in Britain, where expenditures are substantially lower and services are
available to all, the length of queues, quality of services, and internal misallocation of medical
resources have driven reform efforts. Many Eastern European nations face the dilemma of trying
to maintain universal access to health care services in the face of sharp de lines in public sector
revenues associated with general economic reforms. The health care systems in these transitional
economies often suffer from overstaffing, unnecessanly large physical facilities with inadequate
medical equipment, and overutilization of services.4 Health care reform in the LDCs is focused
more cn the fundamental problerns of inadequate medical resources (providers, facilities,
equipment and supplies), the geographic distribution of these resources, and the failure of
financing mechanisms to instill providers and patients with adequate concern for the cost of care.
Between and wit-iin these blocks of countries there clcarly are differences in the urgency, nature,
and scope of health care reform efforts. Yet common elements do exist, and lessons learned Di one
context may help to solve related problems in other settings. This, of course, is why current U.S.
health care reform efforts are being monitored closely by other countries and international agencies
such as The World Bank.
In addition to common concerns about how to finance and control rising health ca;e
expenditures,5 many countries face similar problems in encouraging healthier lifestyles that
would reduce the need for medical intervention. While some developed countries report significant
reductions in per capita consumption of tobacco and distilled spirits,6 personal health habits
remain poor in many areas, including Central and Eastern Euwope. For example, a recent review
of Bulgaria's publir sector notes that: "Indicators of the health of the population in Bulgaria show
an overall pattern that compares unfavorably with W- 2em Europe but is typical of the countries in
anXtral and Eastern Europe." The report further notes that "...the main source of differences in life
expectancy between Bulgaria and other countries is in diseases of the circulatory system, with
nechanisms for health care, and alternative mechanisms for controlling costs' as a key area of potentially fruitful
research.
4.   For a detailed discussion of the problems and prospects for health care reform in these countries, see Preker
and Feachem (1993).
5.   Korea, for example, ' experienced very rapid increases in health expenditures. In the 1980-1985 period, real
health expenditure. grew at an annual rate of almost 17 percent, of which 15.4 percent was attributable to rising
utilization and intensit , only 1.4 percent of the increase was due to demographic factors. Overutilization appears to
be a significant problem in Korea, despite considerable 'out-of-pocket' fmincing (51 percent in 198S). For further
details of the Korean experience, see De Geyndt (1991).
6.   See, for example, Tables No.201 (p.123) and No.209 (p.127) of the Statistical Abstract of the United Star
(1991).



3
particularly hi"h levels of stroke." (The World Bank, 1993a; eh. 6). In Hungary, where mortality
rates are among the highest in the world, "...expenditures have not secured large gains in health
standards, primarily because of unhealthy lifestyles and a bias toward curative hospital care."
(Preker, 1993; p.4). Similarly, in a recent World Bank study of Romania, it was noted that
morbidity patterns reflect "...(a) poor individual preventive health measures; (b) unhealthy
behavior patterns (i.e., smoking)..." and that "circulatory diseases are the most impottant cause of
death and disability, and account for 10 percent of hospitalizations.(The World Bank, 1992,
p.107). Similar patterns exist in other Eastern European countries, but the incidence of
behaviorally-induced diseases is also high in some Latin American countries. These problems
have contributed to rising utilization rates and the escalation of health care costs, causing a number
of countries (e.g., Chile, Costa Rica, and Brazil) to embark on major hea)th reforms.7
The purpose of this paper is not to discuss U.S. health care reform, about which much has
been written, but rather to describe an innovative, more flexible, and more general form of health
insurance that has been owerlooked in current discussions of reform. Some ideas about how this
type of insurance might be adapted to a broad range of health care systems, regardless of whether
insurance is privately or publicly funded, also will be given. The remainder of the paper is
organized as follows. Section II describes a fundamental problem in most health insurance
arrangements and poses a question that is central to the success or failure of proposed reforms.
Properties of conventional health insurance are described in Section m, while Section IV outlines
the essential features of an innovative plan that has operated successfully in one California county
school system for more than a decadc. Adaptation of this plan to national health care systems is
discussed in Section V, and some final remarks are offered in Section VI.
II. A RECURRENT QUESTION: HOW CAN PATIENTS BE MADE MORE
RESPONSIVE TO THE COST OF HEALTH CARE WITHOUT REDUCING
INSURANCE PROTECTION?
To ensure greater access to services and reduce the risk of financial loss from illness, many
health care systems have severed the link between the utilization and financing of health clre
services. This severance occurs in systems of socially provided care, as in Britain, as well as in
systems where government serves only as the principal insurer, such as Canada. In both cases,
health care services ultimately are financed through general tax revenues rather than direct user
charges. Under both arrangt .nents, effective prices at the time of utilization are negligible. The
consumer is insulated from risk but also becomes less aware of the true social cost of providing
health care. The fiully insure, consume;    may have less incentive to undertake personal
7.  See World Development Report (1993b), chapters 3 and 7.



4
activities that reduce the need for medIical care. Both responses -- more utilization of covered
services and less risk avoidance -- have been desc-4hed as 'moral hazard," the tendency for
consumers to behave differently when insured.8 Because of these moral hazard effects,
insurance increases the demand for medical services.9 If the supply of services cannot satisfy
the demand forthcoming at a near-zero price, some form of non-price rationing must be adopted or
inevitably will emerge.
Even in the U.S. health care system, where the private sector has played a more important
role in insuring consumers and providing health care, the introduction of major public insurance
programs (Medicare and Medicaid) and the spread of private insurance, especially through the
workplace,10 have expanded the demand for care, raised concerns about moral hazard and
socially inefficient levels of utilization, and stimulated the search for offective ways to control
utilization and costs. These cost containment approaches include greatsr 'cost-sharing" (highez
coinsurance rates), various forms of utilization review under the rubric of "managed care,' ana the
introduction of ""lternative delivery systems' (e.g., health maintenance rganizations or HMOs)
that are thought to have stronger incentives or more effective ways to contxol utilization. However,
any such effort to raise coinsurance rates or to limit access to services often is perceived by
consumers as a benefit reduction and loss of insurance protection.'1 Indeed, the natural
8.   Pauly (1968) generally is credited with explaining the economic incentives underlying the moral hazard
problem. While moral hazard has been extensively discussed within the context of private health insurance plas,
the same problem plagues public health insuance systems. For example, moral hazard appears to be a particularly
acute problem in Easter. Europe. For an overview of the problem in the Bulgarian health care system, which is
quite similar to those of other Eastem European co..utries, see The World Bank (1993a), Ch. 6.
9.   If the agency relationship between consumers and providers of health care extends beyond medical decisions to
include financial interests, providers also may exhibit a form of moral hazard by recommending services to insured
patients that would not be recommended to uainsured patients. We focus on the problem of structuring insurance to
give consumers appropriate incentives, but the problem of structuring provider reimbursement- to induce cost-
effective forms of health care also merits attention.
10.  In 1960, consumers directly paid 55.9 percent of U.S. personal health care expenditures. By 1990, this figure
had declined to 23.3 percent [see Levit et aL (1991), p.50]. This average coinsurance rate for the entire population
may understate the degree of insurance coverage among insured individuals, since approximately 35 million
Americans have no insurance [Eonomic Report of the President (1993), p. 134].
11. Beyond the problem of consumer resistance to cost-sharing, there is considerable debate about the effectivwess
of higher coinsurance rates or copayments on utilization and expenditures. The Rand Health lsurance Study is
viewed as one of the most comprehensive attempts to evaluate the effects sf cost-sharing on utiliz..ion, expenditures
and L.alth status. Rand researchers conclude that "...the response to cost-sharing is nontrivial.' [Manning et aL
(1987)1. Implied price elasticities of demand in the Rand study are roughly -.2, but such estimates vary across



5
resistance by consurners to such changes has been an obstacle to meaningful health care reform in
the U.S. and elsewhere. The success of reform efforts in many settings .nay well hinge on one
central quesdon: Can wefind more positive ways to encourage the rational use of health care -
ways that reward healthy lifestyles and pndent consumption of health care services -- without
substantially reducing thefinancial -rotection offered by he -'th insurance? Finding aii answer to
this question may require fundamental change3 in the stin'iture of public and private health
insurance.
III. PROPERTIFS OF CONVENTIONAL HEALTH INSURANCE.
Consumers derive proit-ztion from health insurance by knowing that when expenses are
incurred they will not have to pay the full amount out-of-pocket. Indeed, under full insurance
(zero coinsurance, c = 0), common in mary public insurance programs, out-of-pocket expense
(E1) is zero regardless of the full expense (E). In some public programs and most private
insurance plans, some fraction of the total expense (c < 1) must be paid by the patient. The larger
the coinsurance rate, the less the insurance protection. For each of these conventional types of
insurance, full coverage and partial coverage, the relationship between out-of-pocket expense and
total expense is shown in Figure 1.
[FIGURE 1]
In the absence of insurance xc = 1), out-of-pocket expense equals the full expense, as
depicted by the 45 degree line in Figure 1. Under partial insurance plans, E. rises with increases
in E, but the lower the coinsurance rate (c), the less the increase in E will be felt by the consumer.
Under full coverage, Eo is always zero and the consumer is entirely insulated from the full cost c'
care. Thus, conventional insurance, particularly full coverage, reduces the financial risk associated
with illness, but also diminishes the consumer's sensitivity to the true cost of care; this, in turn,
may lead to excessive utilization and reduced personal efforts to remain healthy (i.e., moral
hazard). One way to discourage unnecessary utilization in convendonal insurance plans is to
impose a deductible, where coverage only begins after the consumer has expended a specified
different types of services and may be sensitive to the unusual insurance structure used in the experiment. In general,
one would expect cost-sharing to be most effective in containing utilization of elective or non-emergency medical
services, but the effectiveness of cost-sharing is asso likely to vary with existing levels of price, coinsurance, income
and other factors that influence the demand for medical services. For a thorough review of the demand-side literatue,
see Chapter 3 of Rosko and Broyles (1988).



6
initdil amount (Ed) on health care.12 In Figure 1, this case is depicted by the kinked relationship
between Eo and E Note that any increase in the coinsurance rate or the deductible level will be
viewed by the consumer as a reduction in insurance benefits.
A general problem withl most conventional forms of insurance is that low utilization by
consumers goes unrewarded. In partial or full covtrage plans, low utilizers may be reassured by
the fact that, if the need arises, a portion or even the full cost of care will be paid by the insurer.
But invariably there may be a strong feeling that benefits of the insurance are "lost' if no medical
services are consumed. Thwre typically is no way to convert this foregone insurance benefit to
some oither form of reward. Apart from the psychic security of knowing he or she is insured, there
is no way for the insured person to benefit from conventional forms of coverage unless services
are utilized. This lack of reward for low utilization encourages unnecessary use of the systein.
A second, related problem in conventional insurance plans is the lack of direct incentives for
consu-ners to adopt healthier lifestyles. Although some private pl'ans offer lower premiums to
non-smokers, non-obese, or other groups who exhibit lower health risks, once an individual is
enrolled in such plaus there is little, if any, reward for maintaining or further improving personal
healtt. habits. The significance of this second flaw in the structure of conventional i urance
grows as we acquire additional empirical evidence that lifestyle strongly influences the incidence of
major diseases and the cost of health care. 13
IV. AN INNOVATIVE FORM OF HEALTH INSURANCE: THE MENDOCINO
OR STAY-WELL PLAN.
Health insurance markets in the U.S. have changed significantly in the past decade or two.
Of particular note are insurers' efforts to monitor and control costs in more direct ways. Ilese
efforts, among others, include: the introduction of requirements for second opinions by physicians
or pre-approval by the insurer for certain types of treatment; consolidation of health insurance and
health care provision in the form of prepaid HMOs; and direct bargaining by the insurer for service
fee discounts from a specific group of providers (preferred provider organizations or PPOs).
Since the 1980s also has been a period of unprecedented increase in U.S. health care expenditures,
12.  One problem with imposing a deductible is that it is properly viewed by the consumer as a reduction in
insurance protection; the insured is now fully exposed for the first Ed dollars of expenswe. Hospital expenses are often
subject to an initial deductible with additional expenses fully covered. In Figure 1, this insurance structure would
appear as a kinked relationship following the 45 degree line to Ed and then becoming honizontal for E > Ed.
13. For a detailed analysis of the social cozts of poor health habits, see Manning ei al. (1991).



7
there is little primafacie evidence that these structural chan-cs have helped to contain outlays on
health care.14
This lack of apparernt success in cortrolling U.S. health care costs may reflect the fact that,
even in prepaid plans, the fundamnental problemns of health ins.rance, cited earlier, have not been
addressed. For example, once enrolled, the HMO or PPO member faces zero o. very small
copayments and has little incentive to control utilizatio:n; thus low utilization remains unrewarded.
S;milarly, the near complete coverage in preEaid plans provides little incentive or personal changes
in lifestyle that might reduce the need ior servicoes. Unfortunately, popular reform proposals do
little more to alter the basic structure of heath insurance. Health insurance may be extended to a
larger percentage of the populaffon to reduce the problem of 'uninsureds," but the structure of the
insurance is apt to look mnuch like one of the couventional forms of insurarce depicted abcvc in
Figure 1. Alternatives, however, do exist.
The plan of interest is one that originated ir California in 1979. In the late 1970s, California
local and county governments were seeking ways to bring their costs into line with state-wide
limitations on property-tax revenues. As a result, education officials in Mendocino County
devised an innovative health insurance plan for their employees. The plan, which remains solvent
and in effect today, provided incentives to limit unnecessary utilization and to adopt healthier
lifestyles, without sacrificing relatively complete coverage.
Before the stay-well plan was introduced, Mendocino County school employeer had
enjoyed 'first-dollar coverage" which, in 1979, cost about $1260 per employee per year. The
County terminated this plan and signed a new contract with Blue Shield of California, a non-profit
insurer. Tlhe new contract called for a lower-premium ($780 per employee per year) policy with a
$500 deductible. The savings ($480 per employee) were set aside by the County in an interest-
bearing, self-insurance fund that would be used to pay the first $500 of any employee's annual
medical bills. Total annual chorges in excess of $500 were to be paid by the new Blue S"..eld
policy. Thus, from the employee's perspective, he or she still enjoyed complete coverage. 'Tiis
was an important aspect in eliciting employee acceptance of the new plan.
The most interesting and innovative component of the Mendocino Plan was a further
stipulation that any portion of the $500 in the self-insurance fund not used by an employee during
the year would accrue in his or her name and be paid to the employee upun quitting or retiring from
14. Barth and Heffley (1992) give an overview and analys.s of health care policy, receot expenditure paterns, and
reform proposals. A more technical analysis of the potentially adverse effects of HMOs on health care prices is
given by Beaaglou and Heffley (forthcoming).



8
County service. For example, an employee with only $200 in claims during the year would not
only have those claims paid by the self-insurance fund, but would receive a future bonus of $300
(plus any other accumulated amounts) at the time of separation. This inceitive component of the
plan, similar to the "no claims bonus" found in some auto insurance policies, directly rewarded the
employee for not overutilizing health care, but it also provided some incentive to adopt healthier
lifestyles that might reduce the need for medical care. In addition to significantly reducing
claims,15 the rAay-well plan reportedly prompted a number of employees to increase their
exercise, lose weight, quit smoking, reduce alcohol intake, etc. These, of course, are precisly the
types of responses we would like a health insurance plan to evoke, yet conventional full-coverage
plans are likely to reduce the insured party's incentive to adopt hea!th-preseving habits.
Thc employer (M.-ndocino County) appears to have been very satisfied with this plan. The
new, lower premiums were stabilized due tw the favorable claims performance of the group and,
although sorie of the premium savings placed in the self-insurance fund were paid out to r,meet the
first $500 of each employee's claims, not all of the funds were used, and the interest generated by
the fund was retained by the employer.
Employees, too, benefitted. Despite some initial resistance of employee groups who feared
a reduction in benefits, it soon became obvious that workers were truy better uff. Effective first-
dollar coverage was preservec for high-utiliL_rs, since the first $500 in claims were covered by the
self-insurance fund and excess claims were covered by Blue Shield. Moreover, low-utilizers now
had something that they had not enjoyed under the previous conventional plan. If a worker's total
Oaims on the self-insurance fune were less than $500, the 'unused" portion of this self-insurance
was converted to a future cash benefit for the employee. This reward for lcw utilization did not
exist under the previous conventional insurance plan.
The Mendocino or stay-well plan received a good deal of national at.ention in the early
1980s,16 and a few other employers adopted similar plans. Yet, despite Mendocino County's
apparent success and sat.faction with the plan, it has not been implemented on a much larger scale
in the U.S.17 This may well be due to resistance from insurers to any plan that significantly
reduces premium cash flow. For many insurers, investment income from the premium flow, as
15. Conway (1980) reports that only 38 of the 2i$ persons initially enrolled in the plan exceeded the $500 figure
and therefore received no bonus. Each of the remaining 180 persons' had claims of less than $500 and received a
bonus, including 66 who received the naximum bonus of $500 for baving no claims.
16.  See, for example, Conway (1980), Newsweek (1980), and lTm  (1980).
17. Zweifel (1988) describes a premium rebate scheme offered by some private health insurers in Germany.
Empirical results te-nd to confurm that the plan contains costs more effectively than simple increases in coinsurance
rates.



9
opposed to net income from selling and servicing policies, is a major source of final profit. Any
private insurance reform that reduces the premium flow is likely to be opposed by insurers.
There have been concerns about the Mendocino plan. Critics, for example, worried about
the possibility of truly sick employees foregoing needed care, resulting in poorer health and larger
outlays in the future. Mendocino County and Blue Shield tried to minimize the possibility of such
a response by issuing a brochure that encouraged members of the plan to be sensible about seeking
,re and alerted them to various symptoms that should not be ignored. It is also likely, that at least
some of the reported reduction in claims was due to shifting of claims to other policies by
employees with dual coverage under a spouse's health care plan. Yet, in a market setting, even
this response could be beneficial: the adoption of a stay-well incentive by some insurers might
force other insurers to adopt similar provisions to avoid the adverse effects of claims shiftirg.
More recently, concerns also have emerged about the declining real incentives; the $500 maximum
bonus for zero utilization has not been increased since the plan began. Measured in 1980 dollars,
the maximum bonus has decreased from $500 to $293 due to the effects of general inflation. Even
more rapid inflation in medical care prices also has made it increasingly difficult for members of the
plan to qualify for a bonus. These problems, which are readily overcome by appropriate indexing
of the plan, have weakened but apparently not eliminated the initial incentives embodied in the
stay-well plan.
The stay-well approach adopted in Mendocino County seems to have a number of attactive
features including: the preservation of rather complete coverage; direct incentives for individuals to
monitor their utilization without heavy-handed forms of managed care, since the implicit price of an
unnecessary visit or procedure is a smaller future reward; incentives to adopt healthier lifestyles; an
opportunity for low-utilizers and those with healthy lifestyles to convert unused portions of their
health insurance into cash benefits; and potentially a lower overall outlay on health care for the
insured group.
V. EXTENDING THE STAY-WELL CONCEPT TO A NATIONAL HEALTH
CARE SYSTEM.
In order to see how the the stay-well incentive might be applied to a national health care
system, as opposed to a particular group of workers, it is important to understand how this plan
differs in structure from the conventional forms of insurance depicted in Figure 1. The stay-well
plan contains three important parameters: (1) a goal expenditure level (per individual or household)
which may be denoted Eg; (2) a rate of reward, denoted r, for individuals or households with



10
health care expenditures less than Eg; and (3) a rate of coinsurance, denoted c, for total
expenditures in excess of E. Figure 2 illustrates the relationship between out-of-pocket expense
and the full cost of health care for a hypothetical stay-well plan.
[FIGURE 21
Note that when expenditures are less than the goal (E < Eg), out-of-pocket expense is
negative, reflecting the potential reward to low utilizers. Here, where r = 1, the insuired receives
the entire difference between Eg and E. There is no reason why this value (r = 1) would be
optimal, but the smaller the reward parameter, the weaker the incentive to remain healthy and to
avoid unnecessary utilization. When full cost exceeds the goal (E > Eg), the coinsurance parameter
(c) determines the portion of the excess cost that must be borne by the insured. In Figure 2, c is
depicted as a fraction; full coverage would require a zero coinsurance rate. As in any insurance
plan, the problem faced by the insurer is how to structure the plan or, here, how to select Eg, r,
and c to achieve certain objectives. For commercial insurers, profitability, or at least actuarial
soundness, may be the overriding objective; for public insurers the objective(s) may be more
complex, reflecting concerns for the overal! cost of the plan, its effects on particular groups of
consumers and providers, and its impacts on measures of health status. Given the complexity of
the problem, it is likely that the optimal structure will vary for different systems. However, if
utilization and health habits are at all sensitive to financial incentives, there is a case for this type of
insurance in a varety of institutional settings.'8
Although we cannot say much about precisely how to structure such a plan in a given
country without a good deal more country-specific information, several aspects of this plan are
noteworthy. First, the stay-well plan is a more general and more flexible form of insurance,
subsuming most ccnventional plans as special cases. For example, if Eg is set to zero and c c 1,
we have conventional partial coverage. Similarly, if Eg = Ed, r = 1, c < 1, and the insured is
required to pay a lump-sum "tax' equal to rEg, regardless of the full expense for that individual,
the stay-well structure in Figure 2 is shifted up so that it becomes equivalent to the conventional
deductible plan shown in Figure 1.
Second, the rewards for low utilization might take a variety of forms. As in the Mendocino
plan, th . rewards might constitute a credit to a retirement account, but they could just as easily be
annual cash rebates, or credits against out-of-pocket expenses that might be incurred if the
18. Thbe plan may be particularly relevant to the Eastern European setting, where utilization is high, behaviorally
induced diseases are common, and declining real incomes might cause individuals to respond favorably to new
financial incentives to reduce utilization and modify uihealthy behavios



11
individual or family exceeds the expenditure goal (Eg) in a future period.
Third, administration of the stay-well plan does not appear to be unduly complex. In
Mendocino, where initial expenditures (E < Eg) are paid by the employer from a self-insurance
fund and subsequent outlays (E > Eg) are covered by the insurer, processing and coordination of
claims is done through a local medical foundation. If anything, incorporation of a stay-well
incentive in a single-payer or national health care system would be somewhat simpler. Insurance
records of participants would need to contain information about the depletion (if any) of their
potential reward, based on expenditures to date. At the end of the accounting period, these
records would be used to issue reward checks or to credit the individual in some other fashion.
Fourth, success of the stay-well plan in reducing overall costs to the system hinges on the
capacity of the incentive to shift the frequency distribution of health care spending. The purpose of
the stay-well incentive is to reduce unnecessary utilization in the short-run and, in the long-run, to
further reduce the need for care by stimulating better health habits. If utilization and personal
habits are sufficiently responsive to such incentives, the plan may reduce overall system costs. 19
Fifth, financial solvency of the plan requires that revenues from various sources (premiums
paid by individuals and/or employers, government subsidies, and income from any initial
endowments) be sufficient to cover not only the insured portion of charges for high utilizers
(E > Eg), but also the insured charges and rewards for low utilizers (E < Eg), as well as any
administrative costs and allowed profits. Again, the plan has a better chance of remaining
independently solvent (i.e., not dependent on external subsidies) if parficipants' demand for health
care and their willingness to adopt better health habits are sufficiently responsive to financial
incentives. It should be emphasized that, despite the additional payments to low utilizers, the stay-
19.  One might argue that hospital expenditures, which typically account for a significant proportion of total
health care expenditures, will be less affected by the stay-well incentive. First, the demand for hospital care will be
less responsive to financial incentives than other, less expensive types of services (e.g., routine office visits or
examrs). Second, the power of the stay-well incentive hinges on the size of the reward; the lower the reward, the
lower the power to affect excessive utilization and , particularly, to reduce the use of costly hospital services. We
would argue, however, that this is essentially an empirical question that crucially depends on the expenditure
distribution, demographic factors, and lifestyle characteristics of the population. Relationships between these
features of the population and their responses (both utilization and health habits) to financial incentives would need
to be studied befoe a large-scale stay-well plan is introduced (see the Section VI for some thoughts on this matter).
If, however, poor lifestyle is .n important contributor to illnesses or diseases that ultimately require hospitalization,
or if a portion of inappropriate' hospitalization is initiated by routine physician visits, even relatively modest
incentives provided by the stay-well plan might be effective in reducing hospital costs, particularly in the long run.



12
well plan is potentially less expensive than conventional plans with comparable coverage. If the
incentive causes consumers to reduce utilization, search for less expensive providers, and adopt
better health habits, and if these responses also reduce the frequency of claims and administrative
costs, the stay-well plan may lower overall costs. Also, as in ary insurance plan, solvency is
enhanced by group size, due to better risk-pooling (less exposure to "large loss" effects), potential
economies of scale in administrative costs and claims processing, and greater birgaining power for
the insurer in exacting favorable terms from providers of medical care.
Finally, the stay-well plan has certain intergenerational implications. For example, younger
participants, with better immunological and health characteristics, could be the main beneficiaries
of the plan's financial incentives if a single expenditure goal (Eg) is established. This problem
could be addressed by scaling the expenditure goal to age or other relevant characteristics, thereby
giving a broader range of opportunities for individuals to benefit from reduced utilization and
improved health habits. Such modifications could increase administrative costs, but increasing the
number of parameters of the insurance plan also might reduce medical expenditures by allowing
administrators to better "fit" the plan to characteristics and responses of the population. It also
should be pointed out that, even if they cannot cannot directly benefit from the financial incentive,
'inherently high utilizers' may enjoy lower out-of-pocket costs if savings induced by the stay-weU
incentive can be used to provide more complete insurance coverage (lower coinsurance rates).
VI. CONCLUDING REMARKS.
The fundamental dilemma facing health care reformers in many different settings is the
conflict between expanding or maintaining access to health care and controlling health care costs.
Under conventional insurance, attempts to remove financial barriers to care by lowering
coinsurance rates or extending existing coverage to the uninsurel tend to increase demand for care,
forcing up utilization and prices in a market setting or increasing the need for rationing in a non-
market system. This increased demand stems frorn the lower (perhaps zero) out-of-pocket cost to
consumers, as well as the reduced incentive for consumers to avoid health problems. It follows
that one way to increase access without necessarily increasing health care costs is to alter the
structure of insurance to provide stronger incentives for well-protected consumers to adopt
healthier lifestyles and to avoid unnecessary utilization.
The stay-well plan is a simple, yet more general, form of insurance that rewards lower
util,zation of health care services and encourages positive health habits. Mendocino County's
thirteen years of experience with this type of plan suggest that it may be a useful way of balancing



13
the objectives of cost containment and comprehensive insurance protection. Why this type of
insurance has not spread or come to the forefront in current discussions of health care reform is
unclear. Private insurers may simply lack the incentive to develop or accept innovations in the
structure of insurance that would lead to lower premiums.20 To our knowledge the plan has not
been implemented in a public health care system, but the structure of the plan is general enough to
allow it to be adapted to a wide variety of health care systems.
Further study of the stay-well plan is clearly needed. Much of the information about the
response of consumers to this incentive-based plan is anecdotal in nature. Existing plans ought to
be studied more closely, but such plans appear to be scarce and the study of any existing plan is
subject to questions about the representativeness of the sample. An alternative approach to
acquiring more informnation about the feasibility and effectiveness of stay-well insurance might be
the establishment of demonstration projects, in which individuals or families are randomly
assigned to either stay-well or conventional insurance plans. Establishing sufficiently large
samples and using multivariate techniques to control for personal or demographic characteristics
that might also influence behavior, the effects of the difference in insurance structure could be
estimated. Given the policy importance that has been attached to health care reforn, the cost of
operating such projects mnight be small relative to the potential payoff.
There are no easy solutions to the fundamental question posed in Section II, regarding the
difficulty of insilling consumers with appropriate incentives without significantly reducing the
protection affordad by health insurance. Insured consumers have economic incentives to be less
careful about their health habits and to consume even those services that do relatively little to
maintain or improve health. The more comprehensive the coverage, the stronger these adverse
incentives will be. The stay-well plan provides a countervailing incentive for even fully insured
persons to monitor their own health habits and to avoid unnecessary udlization. We believe that
some additional incentive of this type is needed to obtain the desired balance between the social
objectives of insurance protection and health care cost containmenL There may be other forms of
insurance that would achieve this balance as well or even more effectively than the stay-well plan.
What seems clear is that conventional insurance has failed in this regard and, if we are to resolve
the whealth care crisis," we need to think much more creatively ab(ut the strture of insurance
and the signals it conveys to consumers and health care providers.
20.  The stay-weU plan was accepted by the insurer, Blue Shield of California, but was designod by Ed Nickaeman,
an Assistant Superintendent of Schools in Mendocino County.



14
REFERENCES
Barth, Peter and Dennis Heffley, 1992, "Moving toward national health care policy," inWorkers'
Compensation Health Care Cost Containment, J. Greenwood and A. Taricco (eds.), Horsham
PA: LRP Publications, 285-309.
Beazoglou, Tryfon and Dennis Heffley, "Reevaluating the 'pro-competitive' effects of HMOs: a
spatial equilibrium approach," Journal of Regional Science, forthcoming.
Birdsall, Nancy, 1990, "Health and development: what can research contribute?" World Bank
Policy Research Working Paper Series, No.545, Counry Department 1, Latin American and the
Caribbean Regional Office, Washington DC: The World Bank
Conway, John, 1980, "Payoff on health," Forbes, 126 (September 1): 6.
De Geyndt, Willy, 1991, "Managing health expenditure under national health insurance: the case of
Korea," World Bank Technical Paper, No.156, Asia Technical Department Series, Washington
DC: The World Bank
Economic Report of the President, 1993, Washington DC: U.S. Government Printing Office.
Kutzin, Joseph and Howard Barnum, 1992, "How health insurance affects the delivery of health
care in developing countries," World Bank Policy Research Working Paper Seres, No. 852,
Population and Human Resources Department, Washington DC: The World Bank.
Lave, Judith and Lester Lave, 1970, "Medical care and its delivery: an economic appraisal," Law
and Cotwemporary Problems, 35: 252-266.
Levit, Katherine, Helen Lazenby, Cathy Cowan, and Suzanne Letsch, 1991, "National health
expendtures," Health Care Financing Revew, 13: 29-54.
Manning, Willard, Joseph Newhouse, Naihua Duan, Emmett Keeler, Arleen Leibowitz, and Susan
Marquis, 1987, "Health insurance and the demand for medical care: evidence from a randomized
experiment," American Economic Review, 77: 251-277.
Manning, Willard, Emmett Keeler, Joseph Newhouse, Elizabeth Sloss, and Jeffrey Wasserman
1991, The Costs of Poor Health Habits, Cambridge MA: Harvard University Press.



15
Newsweek, 1980, "Maldng good health pay off -- in cash," 96 (August 25): 73.
Pauly, Mark V., 1968, "The economics of moral hazard," American Economic Review, 58: 231-
237.
Preker, Alexander, 1993, "Hungary: pension administration and health insurance," Health Policy,
World Bank Newsletter, June (No. 2), Population, Health and Nutrition Department, Washington
DC: The World Bank.
Preker, Alexander, and Richard Feachem, 1993 (forthcoming), 'Health care and market
mechanisms," in Human Resources in Transition: Labor Markets and the Welfare State in the
CEE, N. Barr (ed.), Oxford: Oxford University Press.
Rosko, Michael, and Robert Broyles, 1988, The Economics of Health Care, Westport CT:
Greenwood Press.
The World Bank, 1992, Romania: Human Resources and the Transition to a Market Economy,
World Bank Country Study, Washington DC.
The World Bank, 1993a, Bulgaria: Public Finance Review, Europe and Central Asia Region,
Country Department 1, processed, Washington DC.
The World Bank, 1993b, World Development Report, Washington DC.
,Yne, 1980, "Wealth for the healthy,' 116 (August 25): 60.
U.S. Department of Commerce (1991), Statistical Abstract of the United States, Washington DC:
U.S. Government Printing Office.
Zweifel, Peter, 1988, "Premium rebates for no claims: the West German experience," in Health
Care in America: The Political Economy of Hospitals and health Insurance, H.E. Frech HI (ed.),
San Francisco CA: Pacific Research Institute for Public Policy, 323-352.



FIGURE 1: Co:iventional Insurance Plans
Out-of-pocket
expense (Eo)
45
Full expense (E)
Ed



FIGURE 2: The Stay-Well Plan
Out-of-pocket
expense (Eo)
0
45
/                '1    we~ 
0    /                                            .-  Full expense (E)
Eg
r = 1
Reward



Polly Research Working Paper Seriea
Contact
Tlte                             Author                  Date               for paper
WPSI 164 Power, Distortions, Revolt, and  Hans P. Binswanger      July 1993           H. Binswanger
Reform in Agricultural Land Relations Klaus Deininger                       31871
Gershon Feder
WPS1 165 Social Costs of the Transiton to  Branko Milanovic       August 1993         R. Martin
Capitalism: Poland, 1990-91                                                 39026
WPS1166 The Behavior of Russian Firms In    Sirnon Commander      August 1993         0. del Cld
1992: Evidence from a Survey     Leonid Lberman                             35196
Cecilia Ugaz
Ruslan Yemtsov
WPS1 167 Unemployment and Labor Market    Simon Commander         August 1993         0. del Old
Dynamics In Russia               Leonid Lgbrman                             35195
Ruslan Yemtsov
WPS1168 How Macroeconomic Projections     Rashid Faruqee          August 1993         N. Tannan
in Policy Framework Papers for the                                          34681
Africa Region Compare with Outcomes
WPS1 169 Costs and Benefits of Debt and   Eduardo Fernandez-Arias  August 1993        R. Vo
Debt Service Reduction                                                      33722
WPS1 170 Job Search by Employed Workers:   Avner Bar-Ilan         August 1993         D. Ballantyne
The Effects of Restrictions      Anat Levy                                  37947
WPS1 171 Finance and is Reform: Beyond    Gerard Caprio, Jr.      August 1993         P. Sintim-
Laissez-Faire                    Lawrence H. Summers                         Aboagye
38526
WPS1 172 Liberalizing Indian Agriculture:  Garry Pursell          September 1993      D. Ballantyne
An Agenda for Reform             Ashok Gulati                               37947
WPS1 173 Morocco's Free Trade Agreement with Thomas F. Rutherford  September 1993     N. Artis
the European Community:          E. E. Rutstrom                             38010
A Quantitative Assessment        David Tarr
WPSI 174 Asian Trade Barriers Against Primary Raed Safadi         September 1993      J. Jacobson
and Processed Commodities        Alexander Yeats                            33710
WPS1 175 OECD Trade Barriers Faced by the   Bartlomiej Kaminsid    September 1993     J. Jacobson
Successor States of the Soviet   Alexander Yeats                            33710
Union
WPS1 176 Cash Social Transfers, Direct Taxes, Branko Milanovic     September 1993     R. Martin
and Income Distribution In Late                                             39065
Socialism
WPS1 177 Environmental Taxes and Policies    Neil Bruce           September 1993      C. Jones
for Developing Countries         Gregory M. Ellis                           37699



Policy Research Working Paper Series
Contact
Title                           Author                  Date              for paper
WPS1 178 Productivity of Pubikc Spending,    John BafHes         September 1993     C. Jones
Sectoral Allocaion Choices, and  Anwar Shah                               37699
Economic Growth
WPS1 179 How the Market Transition Affeted   Bartlomkjl Kaminski  September 1993    P. Kokils
Export Performance in the Central                                         33716
European Economies
WPS1 180 The Financing and Taxation of U.S.   Harry Huilnga      September 1993     R. Vo
Direct Investment Abroad                                                  31047
WPS1 181 Reforming Health Care: A Case for   Zeliko Bogetic      Septemt ir 1993    F. Smith
Stay-Well Health Insurance      Dennis Hoffley                            36072