Thomas Cox RN, BA, BSN, MS, MSW, MS
Virginia Commonwealth University
School of Nursing
March 26, 2002
This paper
describes the concept of insurance risk assumption by professional caregivers
in the healthcare industry. The concept of Professional Caregiver Insurance
Risk (PCIR) refers to the financial risk that a Professional Caregiver (PC)
(individual provider, professional partnership, group practice, department,
division, or institution) accepts when it enters into specific types of
contracts that have as a primary feature, the transfer of risk as typically
occurs in insurance contracts. A key consideration is the concept of an Average
Cost Based Reimbursement Plan (ACBRP). An ACBRP is any agreement that
substitutes the payment of an estimate of the average cost for providing
services for the unknown actual costs that will be incurred in providing those
services. ACBRP mechanisms include: capitation agreements, Diagnosis Related
Group (DRG) financing mechanisms and Preferred Provider Organization (PPO)
reimbursement rate reductions. All these mechanisms have the effect of shifting
the burden of meeting unanticipated expenses to PCs rather than retaining those
risks in the hands of insurers or aggregators of risk. However, these are not
the only relationships that involve PCIR transfers. Budgeting inadequacies
within institutions also represent risk transfers. This latter form of risk
transfer has traditionally affected nurses and nursing departments more than
any other group in the health care environment.
The
Oxford online dictionary defines “professional” as a person:
“That is trained
and skilled in the theoretic or scientific parts of a trade or occupation, as
distinct from its merely mechanical parts; that raises his trade to the dignity
of a learned profession.”
The
Oxford online dictionary defines a “risk” as:
“The chance or hazard of
commercial loss, spec. in the case of insured property or goods. Also (freq.
without article), the chance that is accepted in economic enterprise and
considered the source of (an entrepreneur's) profit.”
And
uses ‘uncertainty’ as a synonym, defined as:
“The quality of being
uncertain in respect of duration, continuance, occurrence, etc.; liability to
chance or accident. Also, the quality of being indeterminate as to magnitude or
value; the amount of variation in a numerical result that is consistent with
observation.”
Risk has also been described as 'uncertainty' about
outcomes (Ray, 1986, Angell, & Pfaffle, 1980).
As well, the Oxford online
dictionary defines “insurance” as:
“The act or system of
insuring property, life, etc.; a contract by which the one party (usually a
company or corporation) undertakes, in consideration of a payment (called a premium)
proportioned to the nature of the risk contemplated, to secure the other
against pecuniary loss, by payment of a sum of money in the event of
destruction of or damage to property (as by disaster at sea, fire, or other
accident), or of the death or disablement of a person…”
There
is no extant definition for ‘caregiver’, in part a contraction of ‘care’ and
‘giver’ but also imbued with a different meaning as a noun used to describe
both the action and the person doing the act. We see then, that Professional
Caregiver Insurance Risk is a complex concept that involves the aggregation of
different roles. As shall be developed further along in this paper, the roles
of professional caregiver and insurance risk assumer are also likely to be
incompatible.
The following definition is offered:
Professional
Caregiver Insurance Risk is the financial risk that exists when professional
caregivers accept the obligation to provide unknown professional services, at
unknown costs, to a defined cohort of clients, for a specified period of time,
in return for a premium paid by a policy aggregator or insurer. The
professional caregiver has professional skills, training, and credentials,
which attest to their ability to provide such services but they possess neither
the professional skills, nor the financial capacity, nor do they meet statutory
requirements to engage in the business of writing or backing traditional
insurance policies. Since the costs that will be incurred during the course of
the contract life are unknown at the time the contract is negotiated, the
uncertain financial consequence of the contract reduces the uncertainty of the
outcome, and increases the predictability of the financial experience for the
ceding insurer and increases the uncertainty of the outcome, and decreases the
predictability of the financial experience for the accepting professional
caregiver.
Model, Contrary, Related, and Borderline Cases
Model Case: The clearest example of
an ACBRP that meets the criteria for being a case of a PCIR transfer is the
capitation contract. A
capitation contract is an agreement, within contract law, between two parties:
First, a party that provides health related services, on behalf of third parties, for a pre-determined monthly payment,
Second, a party that provides pre-determined monthly payments in-lieu of purchasing billed services (Anderson, 1999).
Under the conditions of a capitation contract a professional caregiver agrees to provide an unknown quantity of services to a cohort of clients for a fixed cost. Since the actual costs are unknown and unknowable at the time the contract is negotiated, the uncertainty about the costs meets the classic definition of an insurance transfer.
Contrary cases: PCIR as a
concept does not describe the fact that every contractual relationship,
including those, which theoretically guarantee a profit, such as defense
contracts between the Department of Defense (DOD) and defense industry
contractors, entail risks. All of these contracts have risks in them. The
government may decide that they want to eliminate a weapons system, the
government may reject cost overruns on a subsequent review, and may make
allegations of fraud or incompetence after the contract period ends.
As well, were a PC to enter
into a contract that was specifically an insurance contract, as might occur if
the PC became an underwriter in a Lloyds syndicate, this would not represent an
instance of PCIR assumptions because the central concept of being conducted in
the course of their PC activities would not hold. Likewise, if a PC were to
perform as an agent or broker of insurance, this too would not represent an
instance of PCIR since it would involve their non-PC skills, talents and
licenses rather than those that underlie their roles as a PC.
With every
economic relationship that a professional caregiver (PC) enters there is an
implicit, and often explicit, assumption of risk. The purchase of a new
computer system may prove to be a financial loss if it is inadequate to meet
the PCs needs, fails to operate as expected, or becomes obsolete sooner than
anticipated. Each new employee exposes the PC to the possibilities of employee
theft, malingering, sub-standard performance, claims of harassment and
discrimination hence, the possibility of financial loss and even financial
ruin. The distinction rests, in part, on the fact that it is not the intention
of the contractual relationship to transfer risk. Ostensibly, each of the
contracting parties has a reasonable expectation that they will both benefit by
the nature of the contractual relationship and the risks that occur are
incidental rather than.
Related
case: Diagnosis Related Group financing is related to PCIR in the sense that
the costs are dictated by the government, the services are aggregated together
and often indivisible, and PCs either agree to the terms of the DRG provisions
or they opt out of the agreement with an attendant loss of potential clients.
As well, some organizations are required to accept DRG provisions because of
other relationships that they have with the federal government. However, a
fundamental difference is that the PC may not be solely responsible for care to
a defined cohort of patients. This is most clearly the case when viewing a
hospital that is subject to DRG provisions. If the hospital is not the sole
provider in the area, it can influence the costs it subjects itself to, by
simply not providing services it cannot provide economically. So open-heart
surgery may be offered where neurosurgery is not. In this way, since at least
theoretically, consumers have other options, the hospital is able to mitigate
the detrimental effects of DRG provisions in ways that a capitated provider
cannot because the capitated provider has to provide all contracted services to
a defined cohort of clients or suffer financial consequences that may be more
severe than the costs to provide the services.
As suggested earlier other forms of risk transfer exist. If a nursing department is normally inadequately budgeted and yet required to continue to meet performance targets that require additional funding, there is a form of risk transfer involved. However, the manifestations of this form of risk may vary greatly. It may manifest in overtime work requirements that sap the strength of caregivers. It may result in individual harms as might occur when a nurse manager is held accountable and punished for failing to meeet performance standards that are impossible to meet given the budgetary limits imposed. As well, there may be risks that accrue to professional staff such as censure by disciplinary boards for failure to meet professional licensing standards in the performance of their professional duties. This latter risk might occur if inadequate staffing leads to inadequate care that would not have occurred had the organization properly staffed the unit. In particular, this might occur in a situation where the professional realizes that the unit is inadequately staffed but has two untenable choices: leave the facility in which case they are guilty of patient abandonment, or stay in the understaffed environment and assume the risk that they will not be able to adequately meet the needs of all the clients in their care. However, none of these cases assume an actual duty to incur financial liability for unfunded costs of care so they do not meet the requirements for PCIR.
Borderline cases: Preferred Provider Organizations (PPOs) are on the borderline, containing some features that are similar to PCIR and some features that are not. In a capitation contract, the fees for particular services are subsumed in an aggregate premium. The capitated provider does not know exactly how many units of a particular service will be required and the costs for those services are not directly addressed in the contract. In a PPO the fees for particular services are set in advance. The provider has, in effect, had an opportunity to determine that they are willing to provide the required service at the contracted rate. In this sense the PPO contract does not represent a risk transfer. However, it is unlikely that a PC would contract for a narrow range of services. The most likely basis for a contract is that the PC agrees to provide a range of services, each of which has been contracted for at a particular reimbursement rate. Hence, it is unlikely that each of the services a caregiver has agreed to under the contract is equally profitable. Hence, it is likely that the PC has agreed to provide some services that are unprofitable at the agreed rate in order to be able to provide other services that are profitable for them to provide at the agreed rate. As an aggregate then, their future experience is uncertain. If they get far more unprofitable requests than anticipated they incur unexpected costs. As well, if they get fewer profitable requests than anticipated they will lose anticipated revenues. While they have agreed to provide services at specific reimbursement rates it was not their intention to provide a particular mix of services that would be unprofitable. Hence, it meets the same basic criteria of insurance risk transfer despite the fact that there is not a defined cohort of clients. This is true because the insurer that wrote the original polices has been protected from uncertainty as a consequence of the PC accepting that uncertainty.
Another borderline case occurs with Health
Maintenance Organizations (HMOs). A HMO assumes financial risks related to
providing its members with specific medical services for a fixed price and for
a given period of time. Each HMO client must choose a physician as their PCP
(primary care physician). The PCP functions as a "gatekeeper" to
specialists. HMOs typically do not
allow subscribers to see physicians outside their network, or at least will
neither pay for such services nor accept the opinions of such providers. The
HMO controls its caregivers by playing a central role in decision making. By
exerting greater control over access to service, the HMO can better control
costs. As well, by paying providers on a capitated basis and monitoring costs
and reviewing utilization rates, referral patterns, and treatment patterns, the
HMO keeps its costs down and rationalizes its operations. While the capitated
providers in an HMO network fit the definition of PCIR, the HMO itself does
not. The HMO may retain responsibility for meeting health care costs even if
the PCP is derelict.
In
order to understand PCIR it is necessary to have at least a very basic
understanding of how legitimate insurance works. The theory of utility (TU) is
the basis for two parties entering into a risk transfer contractual
relationship that involves mutual gain. Philosophers and mathematicians
elaborated the general principles of utility theory long ago. The theory of
probability and statistics was developed, in part by philosophers and
mathematicians to address the situations, in games of chance and in issues of
ethical choice, where determinate outcomes were unlikely or where the outcomes
presented different levels of desirability or undesirability. However, as often
happens, as these analyses progressed, philosophers and mathematicians began to
realize that games of chance and the analysis of them were defining
characteristics of the human condition and the nature of the universe with very
deep philosophical and mathematical implications.
What we call utility theory was
Daniel Bernoulli's (1738) solution to the St. Petersburg Paradox (1713). The
solution involved two ideas that have since revolutionized economics:
1. People's utility from wealth,
u(w), is not linearly related to wealth (w) but rather increases at a
decreasing rate - diminishing marginal
utility;
2. A person's valuation of a
risky venture is not the expected return of that venture, but rather the expected utility to be derived from that
venture.
Utility Theory has two predominant philosophical precursors. One of these
had to do, as described above, with the analysis of games of chance and under
what conditions a rational actor would be willing to pay a pre-determined
amount to play. On the other hand, it also has roots in the area of religion,
government, law, and ethics as a consideration of the values assigned to
different outcomes and the decision-making process (Bentham, 1970; Bentham,
1988; Mill, 1992). Mathematical utility theory is used to explain why people
take risks or behave in certain ways. The argument is that people will take
risks if they believe that what they are risking is not dear to them or if the
reward they expect is very great, and they will avoid taking risks if what they
place at risk, in that process, is very dear to them or the reward they expect
is very small.
In a typical employment relationship one party, the employer, has a
surplus of cash and a surplus of work to be done. They believe that they will
be better off by paying someone else to do the work. Being freed from the need
to do the work may carry a great utility, or value, for them. The employee on
the other hand, may have a surplus capacity of work ability and a limited
amount of wealth. By entering into the employment contract the employee uses
their ability to do work to secure the additional wealth they want. Since
unremunerated work capacity is generally not considered valuable, whereas cash
is, they benefit from giving away something they do not value as dearly as they
value the money. Unlike zero-sum games, in which one player gains only at the
disadvantage of the other, the theory of utility makes transfers between
actors/actresses in the market place valuable for both.
This is also generally true in the case of insurance contracts. There are
two reasons that insurance, as a social and economic risk transfer mechanism,
works so well. First, insurers benefit from the Central Limit Theorem (CLT) and
the Law of large Numbers (LLN): two statistical and probabilistic laws that
govern insurance theory as well as statistical sampling theory. The more
similar policies an insurer writes, the more accurately the insurer can
calculate its likely losses for the entire collection of policies. This is a
simple application of statistical sampling theory, based on the idea that the
standard error of the estimate decreases in magnitude as the size of the sample
increases. Typically, insurers write many policies and can very accurately
estimate their aggregate or average losses on those policies. As well, the
insurer, because it can accurately estimate its losses, can charge a relatively
small risk premium for its risk assumption services. This risk premium is
generally so low as to be attractive to the purchaser of insurance. Coupled
with the portion of the insurance premium that covers average costs and the
amount of the insurance premium applied to expenses of the insurer exclusive of
losses, the total premium (loss costs + expense cost + risk surcharge) is far
lower than the insured would need to have to cover their potential losses in
the event of even a relatively small, partial loss. This frees the insured from
having to maintain a large amount of capital in reserve to cover possible
losses, allowing the investment of the difference between the insurance premium
and the reserve they would need to maintain, to other desired ends. In the
business world, this means that more money is available to invest in the
ongoing, profit-making, activities of the firm. In the case of individual
consumers there may simply be no way of planning for risk management that is
economically feasible.
Average
Cost Based Reimbursement Plans as Insurance Transfers
In the case of ACBRPs, such as capitation contracts, the usual premise is
that the provider is willing to take a small risk of a loss for the potential
of a relatively large gain, if the care actually rendered to the cohort of patients
costs less than the ACBRP payments that are received. However, several features
that protect insurers in the above description act against the assumption of
insurance risks by PCs. Since ACBRPs generally involve a small number of
patients in the portfolios, relative to the number of patients in the entire
collection of insureds, the LLN and the CLT are positioned in the wrong
direction. The risk of loss that is transferred by health care financing
organizations, referred to in this paper as Policy Aggregators (PAs), actually
increases rather than reduces the collective risk when it goes to the provider.
This occurs because the standard error of the mean, i.e. the uncertainty in the
accuracy of the estimate of average or aggregate costs is based on a smaller
sample size and hence is larger when considered from the PC’s position than
when calculated from the insurer’s position.
Policy
Aggregators and Insurance Companies
Conceptually,
there is a dramatic difference in the health care insurance marketplace.
Distinctions customarily made between different categories of relationships
between PCs and entities that finance health care services are meaningful but
insufficiently descriptive of the essential differences that exist in the
marketplace. The most meaningful distinctions that should be made are between
insurers, entities that accept, retain, and manage risk on the basis of the
traditional mechanisms of the insurance industry, the LLN and the CLT and what
may be called Policy Aggregators (PAs).
PAs have little or no intention of performing the
insurance function. In the extreme, they are little more than insurance agents,
writing policies and passing them along to a willing insurer. The difficulty is
that the distinction has not been adequately drawn and insurance regulators,
legislators, accountants, and actuaries, by benign neglect, or intent, maintain
and promote the mistaken impression that PAs are insurers. PCs, ill-prepared to
understand the mechanisms of insurance, are also inclined to view PAs as though
they were assuming the role of an insurer. However, PAs transfer much of the
risk that they should be managing to PCs through ACBRPs. This has the effect of
placing these same PCs in the insurance business. In effect, the PA functions
as an insurance agent might, writing policies and passing the responsibility
for meeting the terms of the policy to the actual insurer, the PC. In the
process the PA takes a substantial portion of the insurance premium as a
commission, covers its expenses and any retained losses, and passes a
healthcare loss premium to the PC in return for protection from the risks in
the policies it has written. In the process, the PA rationalizes its
operations, streamlines its activities, and effectively exits the insurance
business.
Two fundamentally different types of PAs are of
particular note. First, are those organizations, which entered the health
insurance marketplace in the last three decades and looked very much like an
insurance company. Health Maintenance Organizations (HMOs), Managed Care
Organizations (MCOs) and Preferred Provider Organizations (PPOs) are exemplars
of this first group. The longer standing exemplars are health maintenance
organizations that function as PAs in the sense that they accept responsibilities
for caregiving and then redistribute that responsibility internally.
Kaiser-Permanente would be an example of such an organization.
However,
different examples would include hospitals, nursing homes, long term care
facilities, and clinics, who have entered into contracts that require them to
provide services to a cohort of clients that are unknown at the time the
contract is signed. Depending on the mix of clients served and the variability
in the caregiving requirements and financial experiences of the units to which
costs and responsibilities are transferred, significant risk may be transferred
along with clients. This is critically different than agreeing to provide a set
amount of service, where the costs may be estimable in advance. In risk
transfer contracts the number of units of service and the types of service
required would not be known at the outset of the contract.
Insurance companies (ICs), unlike PAs,
actually retain and manage risk. Characteristic features of ICs are that they
write large numbers of similar or identical insurance policies, distribute
their activity over wide geographic areas, employ insurance professionals such
as actuaries and underwriters to monitor, analyze, and insure solvency and rate
adequacy, conform to Financial Accounting Standards Board (FASB) standards
regarding financial accounting and reporting, comply with state and federal
regulations of the insurance industry, and, usually only enter into
re-insurance agreements with other, regulated, financially capable, and solvent
insurers, where the rates are calculated as a sum of loss cost estimates,
expense estimates and a risk premium.
Unfortunately, PAs have an unfair advantage
in the marketplace because they have reduced their least predictable operating
costs, the risk bearing, by virtue of ACBRP contracts. Since they are able to
limit their exposure to risk, by establishing a maximum cost for the contracts
they write and pass on to PCs, they are virtually guaranteed profitability regardless
of the actual experiences of the PC. Under a fully capitated agreement between
a PA and several diverse PCs, the PA’s costs are fixed at a level that does not
exceed the amount in the premium they charged that is earmarked for losses.
This leaves the expense and profit portions of the premiums they charged their
clients, intact. This means that the PAs have not compensated the PCs for the
assumption of the risks that have been transferred. In fact, they cannot
compensate the PCs for the risk assumption because the cost of doing that would
exceed the loss component of the premium they charged the insured due to the
inverse effects of the CLT and LLN when risks are transferred from larger
entities to smaller entities. To fairly compensate the PCs, the PAs would have
to pay an aggregate amount to them that exceeds the loss and risk premium
charged to the insured. The reason that this is the case is that there is
relatively little risk when large numbers of policies are aggregated, but when
the same portfolio is carved into smaller units the risk increases.
Insureds
There are two categories of insured entities
that are noteworthy. The first insured entity is often an employer, a labor
union, or a social collective that contracts for health insurance on behalf of
a well-defined group. These insureds may properly be referred to as Purchasing
Insureds (PIs). The second entity is the consumer of health services, that is
the individual/family beneficiary of the aggregate purchase of the insurance
contract. . These insureds may properly be referred to as Consuming Insureds
(CIs). These two different categories of insureds often have mutually
inconsistent aims. The PIs are usually interested in minimizing costs, while
the CIs are interested in maximizing benefits, particularly when they actually
need the benefits of the contract. However, most CIs do not focus on the fact
that they may need the benefits some day and often are unmindful, even outright
careless about insuring that the benefits will be available when they are
needed. As well, PIs are often more focused on immediate costs of the insurance
contract rather than the issue of whether benefits will actually be available
at some undetermined time in the future. The difficulty is that the situation
changes dramatically when the CIs actually need the benefits. Frequently,
neither the PIs nor the CIs pay much attention to issues such as managed care,
second opinions, policy exclusions, HMOs versus third-party payers,
experimental versus clinical procedures, and co-payments until such as time as
a health problem exists. Faced with a life threatening condition, an expensive
treatment, or an acute condition, precisely the events that the insurance was
purchased to cover, the exclusions and procedures that were accepted in the
insurance contract might come back to haunt unwary insureds. The CI, may
discover that they do not have any coverage for a wide variety of conditions,
or that they will have to fight to obtain the rights they assumed were theirs.
The Interplay Between Insurers, Aggregators
and Insureds
The
insurance marketplace is a dynamic and competitive economic market. Thousands
of insurers exist, though the market is clearly dominated by a handful of the
largest entities. As well, there are a variety of economic factors that create
opportunities and liabilities in the marketplace. State and Federal laws
require the provision of health insurance to certain classes of employees and
by certain classes of employers. Such laws serve as an inducement for insurers
to enter the marketplace since purchase of their products is obligatory. As
well, such regulations have exerted a positive economic effect in that as PIs
are forced to buy insurance, rates for insurance products may be reduced. As
well, state and federal governments are active participants in providing
insurance products and competitive services in some markets such as homeowners
insurance in flood plains, or state sponsored Joint Underwriting Associations
when automobile insurance premiums reach politically untenable levels or when
insurers are refusing to insure high-risk drivers. In the health care arena,
federal programs such as Medicare and state sponsored programs such as Medicaid
provide government sponsored insurance.
Another source of competitive pressure is
care provided by state and local governments, such as: public health clinics,
TB hospitals, and state run geriatric and mental hospitals. Additional forms of
competitive insurance products include: health insurance programs for elected
government officials in the House of Representatives and the Senate, Veteran’s
Administration facilities and services, and CHAMPUS for military personnel, all
of which exert negative economic pressures on the insurance market place –
tending to reduce business opportunities and decrease potential premium income
for insurance products. In addition, a variety of state, local and federal
health care services such as public psychiatric hospitals, public health
clinics and prison facility health programs assume some roles in the provision
of ‘insurance’ to people who are either unable or unwilling to secure private
insurance. This latter situation is actually an advantage to the private,
profit making insurance industry since these clients are likely to generate
costs that exceed their ability to pay in the form of premiums.
Moral Hazards
One aspect of the contract that is very
troubling is that it is theoretically and practically possible for the PC to
affect their costs merely by failing to correctly diagnose, or once correctly
diagnosed, to refuse appropriate treatment to the client. In normal insurance
contracts the insurer is concerned that the insured may cause a loss to occur
merely to be able to collect the coverage the insurance provides. This might
happen is the insured has a car that has mechanical problems that would be very
expensive to repair, but if the car is stolen the net value to the insured will
be higher than the value of the automobile minus the repair costs. In fact
there is a specific term, moral hazard, which describes this situation. When
PCs become insurers the moral hazard issue operates in the opposite direction.
The PC can affect the actual costs they incur merely by not providing needed
services or by delaying the diagnosis and treatment of the clients they serve.
The only impediments to doing this are the PC’s ethical commitment since there
is no party other than the provider and consumer involved in the caregiving
relationship. An informed consumer may become aware that a denial of service is
occurring but will not, in general, have an ability to influence the situation
in a short period of time.
Conclusion
A
new concept Professional Caregiver Insurance Risk, an exemplar, contrary and
borderline examples have been presented. Consequences for professional
caregivers, consumers and insurers have been elaborated. As well, the possible
ethical consequences implicit in this concept have been briefly described and
discussed. An understanding of Professional Caregiver Insurance Risk is
critical for nursing and other health care professions. Absent an understanding
of the true nature of these contracts, professional caregivers may be tempted
to enter into such contracts with policy aggregators. Understanding the fact
that these agreements are a form of insurance for large corporations and that
the contract provisions may entail the possibility for financial as well as
professional harm, ethical conflict and potentially unlimited liability for
loss are critical. Professional caregivers may want to enter into such
agreements if they can accurately predict and manage their potential losses
under such arrangements. However, the ability to predict and manage the risks
these contracts entail are difficult for insurance professionals such as
actuaries and underwriters. Professional caregivers should critically evaluate
their own skills and abilities incident to these contracts before agreeing to
any average cost based reimbursement agreement.
References
Anderson, G.F.P. (1999). Methods of reducing the
financial risk of physicians under capitation. Archives of Family Medicine,
8(2), 149-155.
Angell, F.J. & Pfaffle, A.E. (1980). Insurance
and risk management. New York: Insurance Advocate.
Bentham, J. (1988). A Fragment of Government.
Cambridge: Cambridge University Press.
Bentham, J. (1970). Of Laws In General.
London: Athlone Press.
Cox, T. (2001). "Risk theory, reinsurance and
capitation." Issues in Interdisciplinary Care, 3(3): 213-218.
Health Symphony (2002). Retrieved March
25, 2002, from
http://www.healthsymphony.com/insurancedefinitions.htm
Mill, J.S. (1992). On liberty and utilitarianism
New York: Knopf.
Oxford English Dictionary (2002).
Retrieved March 25, 2002, from http://etext.lib.virginia.edu/oed.html.
Ray, P.H. (1986). Risk,
values, and lifestyles. Menlo Park: Values and Lifestyles Program.