Faculty of Economics, Rijeka, Croatia
Faculty of Economics, Rijeka, Croatia
Faculty of Economics, Rijeka, Croatia
The agency problem in healthcare and the
importance of incentives
The agency problem in healthcare is caused by information asymmetry between the principal
and the agent and is different than the agency problem in other economic fields due to the
specificities of healthcare systems. Although other principal-agent relationships are present
in healthcare, the most common one is the one between the patient - principal, and his
physician - agent. Bearing in mind that there is a certain conflict of interests between the
patient and the physician, the ultimate goal is to create an incentive compatible contract
which will maximize utility of both the principal and the agent. A well designed payment
system in healthcare has a great importance in influencing physicians’ behavior and when
determining the payment system which will contribute the most to the motivation of the
physician to maximize the patient’s utility, the guiding principles should be optimal use of
resources and effective resource allocation to ensure an efficient healthcare system.
Key words: information asymmetry, incentive compatible contract, health care efficiency,
resource allocation, healthcare payment systems
The agency relationships in healthcare and the agency problem associated with the ability to
create health services by physicians which does not provide maximum benefits to the patient,
result from the characteristics of healthcare itself. Although there is a disagreement among
economists regarding the extent to which healthcare differs from the conventional economic
goods and the relative importance of various characteristics of healthcare, the prevailing
notion is that there are four major characteristics of healthcare. Namely, the demand for
healthcare is derived from the demand for health, and then there are externalities, the
asymmetry of information between healthcare providers and recipients of services and the
uncertainty of the need for healthcare together with the uncertainty of the effectiveness of
This article is the result of the scientific project "Approaches and Methods of Cost Accounting in
Croatian Public Sector" no. 081-0811272-1276, financed by Croatian Ministry of Science, Education
The paper will discuss the agency relationships in healthcare, resulting from information
asymmetry between the providers and the recipients of health services, mostly the one
between the patient and the physician. Due to the complexity of the diagnosis and treatment
procedures, and partly because the patient can not collect and process all the necessary
information about his medical condition and treatment options, he has to rely on the
information provided by health professionals. That puts physicians in a monopolistic position
in terms of information and gives rise to the possibility of physicians’ induced demand.
However, there is no widely accepted physician agency model in theory because the
applicability of the pure profit maximization paradigm to the physician agency model is rather
questionable, mostly because the behavior of physicians is largely motivated by other factors
such as status, medical education, professional ethics, altruism, etc.
2 The agency relationship and the divergence in
goals of principal and agent
An agency relationship arises in a situation where one or more persons (the principal)
engages another person (agent) to perform certain work on their behalf, which among other
things, includes delegating authority to make decisions. In the business economy agency
relationships are part of everyday business practice, and the most common one is called the
pure agency relationship between the managers (the Board) and the shareholders. In the
focus of economic studies is the behavior of managers in large corporations where the
management is separated from the ownership and the agency problem arises if the agent
fails to act in the principal’s best interest, respectively if the both sides tend to maximize their
own utility with conflicting cost functions.
The basic setting of agency theory is that the agents are selfish opportunists who, if they are
not monitored successfully, use their position for their own interest. Business ethics theorists
object to that postulate and therefore criticize the agency theory and are against the adoption
of the economic model of action (rational choice theory) that assumes that individuals are
selfish and that they act solely out of selfish reasons. However, the economic model of
rational choice does not imply such conclusions because the utility is inherent to the
individual, and consists of personal preferences that can be both selfish and altruistic.
Accordingly, the agency problem does not occur because of egoistic, but because of different
preferences, i.e. different motivation of the agent and the principal causes the risk of agency
problem and the conflict of interest.
The basic question of agency theory is how to motivate the agent to act in a manner that will
maximize the principal's benefits. The ultimate goal is to create an incentive compatibile
contract between the principal and the agent. Such an agreement should meet the incentive
compatibility constraint, i.e. maximize the utility of both the principal and the agent, while the
principal's utility maximization must be compatible with achieving a minimum acceptable level
of expected utility for the agent. Theoretically it is possible to form an incentive compatibile
contract which will motivate the agent to act in the best interest of the principal and
simultaneously achieve his expected utility, within the following assumptions of agency
theory (Preker, 2007):
1. both the principal and the agent want to maximize their utility, and their utility
functions are independent of each other, i.e. their utility functions have no common
2. income and effort are the two main arguments of the agent's utility function. While
income increases utility, more effort by the agent reduces his utility and, therefore, the
agent always strives to maximize income and minimize the effort invested;
3. the principal knows exactly which action the agent should take, but the information
about the action actually taken is available only to the agent, while the principal could,
directly or indirectly, attain the necessary information;
4. the information about the outcome is directly available to the principal, and the
outcome is perfectly correlated with the agent's effort;
5. the principal and the agent enter into the relationship voluntarily.
The agent is better informed than the principal and therefore has an advantage over him.
The principal can limit the agent's adverse actions by establishing an adequate system of
incentives for the agent, together with the monitoring of the agent's actions, which cause
monitoring costs for the principal. Accordingly, the existence of agency relationship, the
divergence of the physician's and the patient's objectives and the information asymmetry,
result in certain agency costs that consist of the principal's monitoring costs, bonding costs of
the agent and the residual loss (Jensen and Mackling, 1976).
3 The allocation of resources and the agency
problem in healthcare
The basic question of health policy is how to improve the health of the population by using
the available resources so that such an allocation of resources ensures maximum outcomes.
Because of the rising health spending
, there is more and more discussion on the amount of
funds allocated to healthcare, on the purposes which they are allocated to, inside which
services and healthcare programs, all of which aim at ensuring a rational distribution of the
resources between different sectors, services, or specific programs.
As resources are managed by people, the optimal allocation of resources won't be attained
unless the decision makers are motivated to act in this direction. The existence of agency
relationships in the health market, as a consequence of the asymmetry of information and
knowledge, can result in inefficient allocation of resources, which implies an important role of
the physician in providing cost-effective healthcare. Different research indicate that
physicians and hospitals have an important role in the allocation of health resources and it is
estimated that 70% to 80% of the consumption of healthcare resources is associated with
physicians' decision (Reinhardt, 1985), while the hospital sector absorbs 30% to 50% of
health spending in developing countries, and even more so in the developed countries (Mills,
3.1 Incentive schemes and the efficiency of healthcare system
The behavior of the physician is one of the central issues of health economics, and creating
an incentive compatibile contract between the principal and the agent – physician is one of
the fundamental goals of the efficient healthcare system. The motivation of physicians will
influence their decisions about hospitalization, treatment and the time spent on setting a
proper diagnosis. Over a longer time period, the physician's motivation will also influence the
choice of the geographic location, the choice of specialization, etc. For this reason, the
healthcare payment system should encourage the agents to use the available resources
reasonably to ensure the efficient allocation of healthcare resources. In general, the
According to the World Bank data (WB, 2011), total health spending in 2009 in the United States
reached a level of 16.5% of gross domestic product, while the developed countries of Europe spent
10-12% of GDP on healthcare. 7.8% of GDP is spent on healthcare in the Republic of Croatia, which
is very close to the spending of the developed countries, but little more than spending of the south-
eastern countries in the region.
healthcare payment system can be classified into three different categories (Hankins & Baker,
2004): fee for service, prospective payment and the capitation system.
Fee for services represents the coordination of both the patient's and the physician's
interests, because in this way the physician can provide the patient with everything
necessary without any external constraints. The end result for the physician is the difference
between the total cost (investment effort) and the total price charged, therefore, there is an
implicit incentive for the activities that require less effort invested (Rigo, 1999), such as the
prescribing of drugs. Also, there is an incentive to provide specialized services because the
cost of such services is higher.
According to the prospective payment system, the price of the health output is previously
determined, meaning that the price of services is determined before the services are
performed and is independent of the level of the consumed resources. The most widely used
prospective payment system is called the payment system according to Diagnosis Related
Groups - DRG and is commonly referred to as "pay-per-case" and used in all the EU
countries. The basic idea of the payment according to DRG is the payment per episode, with
episodes being the period from the admission to the discharge from the hospital. The
duration of the episode may vary, and all the costs incurred in this period are included in the
price. This means that all the healthcare utilization is included in the episode and that there is
only one payment per episode. The payment system according to DRG meets the criteria of
a good payment system because it (Strizrep):
- encourages and rewards the efficient actions of healthcare providers,
- ensures equality for all the hospitals and the patients,
- ensures the transparency of contracting health services,
- the payment is based on the best available information.
Although the "pay-per-case" represents a good system of the payment for hospital services,
it is not appropriate for the physicians' payment because it is impossible to determine the
contribution of the individual physician in most cases.
In the Republic of Croatia, the National Strategy for the Development of Health 2006 – 2011
announced the introduction of DRG mechanism for all hospital operations. Since 2009 the
services for treating acute patients are paid according to DRG, while chronical diseases
services are paid according to days of hospitalization. The main goals of introducing DRG
are the reduced costs, price uniformity in hospitals, better control and quality of healthcare
and the greater transparency of the system. However, in Croatian hospitals DRG system is
primarily a justification of the remitted funds, which usually exceed the maximum amount of
funds agreed, but still isn't used for the division and the distribution of funds. The latter
function is the responsibility of hospitals, but adequate monitoring, based on which the
viability of services and hospital operations would be assessed, is not provided.
The third method of payment is the capitation system, according to which a certain amount of
funds is paid for the total number of the reported patients in a given time, regardless of the
amount or the type of health services. This system of payment for health services creates at
least three types of incentives for the physicians (Rigo, 1999):
- an incentive to accept and retain as many patients,
- an incentive to limit the use of services by limiting as much as possible the use of
auxiliary materials, technologies and personnel,
- an incentive to select patients of low health risk, that way avoiding patients who would
use medical services in larger amount.
Generally, the assumption of incentive schemes highlights the importance of financial
rewards in motivation that will lead to better results. However, the quality of the results
depends on various internal and external motivational factors, and financial reward is only
one of the goals of human actions. Therefore, the incentive scheme that emphasizes
financial incentives and excludes all other forms of incentives is incomplete and theoretically
and practically obsolete.
3.2 The physician's utility function: single and dual agency
When his utility function is taken into consideration, the physician's behavior is characterized
by the three basic approaches respectively:
1. the utility maximization model,
2. the model of income maximization and
3. the model of the target income.
The proponents of the utility maximization model (Feldstein, 1970; Eustaugh 1992) include
the following elements in the physician's utility function: profit (net income), leisure,
professional status, internal ethics, the complexity of the case, the time spent on improving
knowledge, the number of medical staff that is supervised by the physician, etc. Thus,
income is an important, but not the only element because the physician's behavior is too
complex to be explained only by income. The utility maximization hypothesis implies that the
physician will make effort to the point where the marginal utility of the effort invested is equal
for different elements of the utility function. According to the model set this way, the health
policy focused on elements of the utility different from income can be just as successful, and
the physician's behavior can be directed by coordinating multiple measures of health policy
because the exclusive use of financial incentives can result in an insufficient motivation.
The model of income maximization assumes that income is the dominant element that
affects the physician's behavior, i.e. that the physician's behavior is driven by income. The
proponents of this model (Sloan 1976; Baumol 1988) argue that income has a dominant
influence on the physician's behavior, therefore the simplified income model is sufficient in
explaining his behavior, meaning that the utility maximization model makes the empirical
implementation harder and is unnecessarily confusing and complex. Thus, an appropriate
healthcare payment system and financial incentives may act to direct the physician's
behavior in the desired direction.
A third and widely accepted model of the target income represents a combination of the
aforementioned models and points to the existence of an expected level of income, one
associated with the level of market income of physician of the same rank. Implicitly, the
model assumes that, in a situation where the achieved level of income is below the expected
level, the physician’s behavior will largely be driven by income, while he will consider other
elements of utility more significantly once this level is reached. According to this hypothesis,
if a physician is paid less than his expectations, he will act according to the principles of
income maximization and attempt to increase his income. On the other hand, if he is paid in
accordance with his target level of income or more, he will act in a manner that will enable
him to meet other needs and desires. In economic terms, prior to reaching his target income,
a physician will value the marginal utility of income per invested unit of input (e.g., invested
effort) more than other gains, while after reaching his target income, the marginal utility of
other gains (such as status) per invested unit of input will be valued more than additional
gain in income.
When observing the behavior of the physician and his utility function, it is important to also
emphasize the characteristic of double agency. Namely, when making medical decisions, the
physician must take into account both the patients' interests and the interests of his employer.
On one hand, his utility function is faced with the demands of cost reduction, as well as
achieving the highest possible efficiency set forth by the employer, while on the other hand,
he is faced with realizing the best possible health outcomes for the patient.
4 Deviations from the basic assumptions of the
Although the agency theory, which deals with the situations where exists information
asymmetry, uncertainty of desired outcomes and interdependence of the various outputs of
health services, represents a theoretical base and a general framework for the formation of a
motivation compatible payment system for healthcare, special features of healthcare bring
into question the applicability of the assumptions of standard agency theory in regards to the
patient - physician relationship.
The first assumption of agency theory is the independence of the agent’s and principal’s
utility function. However, their utility functions are to some extent dependent, and there exist
overlapping arguments in their utility function. Therefore, the physician’s invested effort need
not be in a negative relationship with his utility, but his marginal utility of additional effort will
decrease, and ultimately be negative.
Among the most prominent features of healthcare is the asymmetry of information and
knowledge between the recipient and provider of health services, something which thereby
undermines the premise on which the principal - patient knows which action the agent -
physician should take. Even if there is a possibility for the physician to be supervised by the
patient, he can not pass judgment on whether the physician’s action is appropriate because
of the high level of uncertainty of health outcomes, while the outcome itself is not a perfect
function of the received healthcare due to many factors that affect health. For this reason, the
assumption of perfectly correlated outcomes and effort of the agent are violated.
The agency problem in healthcare, one arising from information asymmetry, is manifested
not only through the physician - patient relationship, but also includes a series of principal -
agent relationships within the health system, which together form the so-called relational
network. The unique characteristic associated with healthcare is that changes within one
agency relationship affect all others, an element which makes the formation of a motivation
compatible contract and related incentive compatible remuneration all the more difficult.
Consequently, a perfect remuneration system in healthcare is impossible, while the next best
solution would be a reasonable remuneration system. Such a system should contribute to the
increase of healthcare efficiency because it would (Preker, 2007):
- enable service providers to earn a reasonable income in order to ensure a stable supply
of quality staff,
- motivate service providers to ensure a sufficient volume of needed services of good
- prevent the dissipation of resources and reduce unnecessary healthcare.
Along with a well designed incentive system for agents (health providers), the success of the
agency relationship depends largely on the possibility of the principal (patient) to determine
whether the agent (physician) has met expectations. The agent’s execution must be
measurable and undoubtedly attributable to him. However, healthcare requires cooperation
of the medical staff in order to ensure the required quality of healthcare which is usually the
result of joint effort that is difficult to measure separately. In many activities, that which is
significant can not be measured; for example, a physician’s on-call nights may represent his
work effort, but do not immediately mean better performance and service quality. Therefore,
information on health outcomes, which is available to the patient, is the strongest evidence of
the physician’s performance quality, and represents a good basis for payment of health
services. However, as outcomes in healthcare are uncertain and not perfectly correlated with
effort, and individual efforts of medical staff are dependent on and only part of the total
outcome, a remuneration system based on health outcomes is difficult to form.
The pronounced asymmetry of information and knowledge in the patient - physician
relationship creates the possibility of an agency problem in healthcare, a problem which is
more ‘burdened’ with the specificities tied to healthcare. Because of the pronounced
uncertainty of healthcare that is associated with the emergence of a need for healthcare and
uncertain health outcomes, it is extremely difficult to determine an incentive compatible
payment system that would directly connect the effort of service providers with the level of
benefits in healthcare.
The human factor in healthcare is of paramount importance, and the motives and attitudes of
service providers are important in ensuring the rational consumption and efficiency of
healthcare. Health professionals, in addition to financial motives, are also driven by humane
motives, and the utility functions of patients and physicians are not mutually exclusive.
Although the neoclassical model of profit maximization is not fully applicable in the
healthcare system, the fact remains that healthcare costs are continuously growing and
challenging the sustainability of national health systems. For this reason, it is imperative to
ensure the rational and cost-effective delivery of health services, and one of the mechanisms
which can affect the cost effectiveness of healthcare is a healthcare payment system. When
creating an adequate payment system, one should be familiar with its basic implications for
healthcare expenditure, one of which is the fact that in the shift from paying for services, to a
prospective payment system, to a capitation system, the uncertainty is transferred to the
healthcare providers. In other words, the very same are more motivated to limit costs and
use health resources more rationally.
1. Baumol, W. (1988). Price Controls for Medical Services and the Medical Needs of the
Nation’s Elderly. Chicago, IL: American Medical Association.
2. Eastaugh, S.R. (1992). Health Economics: Efficiency, Quality, and Equity, Westport, CT:
3. Feldstein, M. (1970). The Rising Price of Physicians’ Services. Review of Economics and
Statistics, 52(2), 121–33.
4. Hankins, R. W., Baker, J. J. (2004). Management Accounting for Health Care
organizations: Tools and Techniques for Decision Support. Sudbury, Massachusetts: Jones
and Bartlett Publishers.
5. Jensen, M. C. and Meckling, W. H. (1976). Theory of the Firm: Managerial Behavior,
Agency Costs and Ownership Structure. Journal of Financial Economics, 3(4), 305-360.
Retrieved 11.4.2012. from http://ssrn.com/abstract=94043.
6. Mills, A. et al. (ed.) (1990). Health System Decentralization. Concepts, Issues and Country
Experience, Geneva: World Health Organization.
7. Preker, A. S., Liu, X., Velenyi, E. V. and Baris, E. (2007). Public Ends, Private Means:
Strategic Purchasing of Health Services. Washington: IBRD/WB.
8. Reinhardt, U. E. (1985). Compensation of physicians: Approaches used in foreign
countries. Quality Review Bulletin, 11(12), 366–377.
9. Rígoli, F. (1999). Incentives in health services: An overview of theories and paradoxes.
Retrieved 15.4.2012. from http://www.opas.org.br/rh/admin/documentos/ACF2F.pdf.
10. Sloan, F. (1976). Physician Fee Inﬂation: Evidence from the Late 1960s. In R. Rosett
(ed.), The Role of Health Insurance in the Health Service Sector. New York: Watson
11. Strizrep, T. Dijagnostičko-terapijske skupine (DTS): Priručnik za zdravstvene radnike.
Retrieved 10.4.2012. from http://www.medri.uniri.hr/informatika/assets/DTS_prirucnik.pdf.