Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The impact of public expenditures on health care on total health expenditures: an exploratory analysis of selected OECD countries
(USC Thesis Other)
The impact of public expenditures on health care on total health expenditures: an exploratory analysis of selected OECD countries
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE IMPACT OF PUBLIC EXPENDITURES ON HEALTH CARE ON TOTAL
HEALTH EXPENDITURES:
AN EXPLORATORY ANALYSIS OF SELECTED OECD COUNTRIES
by
Bart Verbelen
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PUBLIC ADMINISTRATION)
December 2007
Copyright 2007 Bart Verbelen
TABLE OF CONTENTS
List of Tables ……………………………………………………………………...v
List of Figures ……………………………………………………………………..vi
Abstract. ........................................................................................................................................ vii
Introduction .............................................................................................................1
Chapter 1 : Health Care Systems: Definitions, Typologies and Individual
Health Care Systems Described .............................................................................5
1.1. Defining a Health Care System. .....................................................................5
1.2. From definition to typology. ..........................................................................7
1.3. Comparing Health Care System Types. .........................................................9
1.4. Health Care Systems at Glance. ...................................................................11
1.4.1. National Health Service Systems ..........................................................12
1.4.2. Non National Health Service Systems ..................................................46
1.4.3. Entrepreneurial Health Care Systems....................................................64
1.5. How Are Health Care Systems Different and Alike?...................................67
1.5.1. The public and private roles in the health care system..........................67
1.5.2. The providers in the health care system. ...............................................69
1.5.3. The individual in the health care system. ..............................................71
Chapter 2 : Trends in Selected OECD Countries 1970-2000. ...........................73
2.1. Trends in Health Total Expenditures............................................................73
2.1.1. Total Health Expenditures per Capita 1970 – 2000. .............................73
2.1.2. Trends in Total Health Expenditures Relative to GDP. ........................79
2.2. Trends in Source of Health Expenditures.....................................................82
2.2.1. Public Share of Total Health Expenditures. ..........................................82
2.2.1. Out of Pocket Share of Total Health Expenditures ...............................86
2.3. Trends in Health Care System Components.................................................88
2.3.1. Inpatient Health Expenditures...............................................................88
2.3.2. Pharmaceutical Expenditures. ...............................................................90
2.4. Trends in Availability of Technological Applications .................................93
ii
2.5. Common Trends or Not?..............................................................................96
2.5.1. Total Health Expenditures: Long Run, Short Run and Government. ...96
2.5.2. Public Share of Total Health Expenditures: All the Same
in the End?.......................................................................................................97
2.5.3. Out of Pocket Payments and Inpatient Share: Relative
Contradicts Absolute Payments.......................................................................97
2.5.4. Pharmaceutical Expenditures and Technology: Are they to blame?.....98
2.6. Puzzling trends and more questions than answers. ......................................99
Chapter 3 : Multivariate Analysis of Total Health Expenditures and its
Determinants........................................................................................................100
3.1. Literature Review.......................................................................................100
3.1.1. Health Expenditures and a Nation’s Income.......................................100
3.1.2. Public Share of Total Health Expenditures. ........................................102
3.1.3. Other Determinants of Total Health Expenditures..............................104
3.2. The Model. .................................................................................................109
3.2.1. Conceptual Framework. ......................................................................110
3.2.2. The Empirical Model...........................................................................111
3.2.3. The Dependent Variable: Total Health Expenditures. ........................112
3.2.4. The Independent Variables..................................................................113
3.3. Data and Methods.......................................................................................127
3.3.1. Data. ....................................................................................................127
3.3.2. Methods...............................................................................................128
3.4. Results. .......................................................................................................134
3.4.1. The Unrestricted Model.......................................................................134
3.4.2. Restricted Model 1. .............................................................................136
3.4.3. Restricted Model 2. .............................................................................139
3.5. Discussion. .................................................................................................142
3.5.1. Issues Comparative Data and Model...................................................142
3.5.2. Discussion of Results. .........................................................................144
Conclusion............................................................................................................155
1. A Look Back to the Future............................................................................155
2. How is the United States Alike. ....................................................................157
3. How is the United States Different and, does it matter? ...............................159
4. The Sisyphus Labor of Changing the United States’ Health Care System. ..161
5. The Ideal Health Care System Reform..........................................................162
6. Final Remarks and Future Research..............................................................164
iii
Bibliography.........................................................................................................165
Appendix ............................................................................................................191
iv
v
LIST OF TABLES
Table 1.1. Differences between Individual and Collective Insurance Schemes......10
Table 2.1. Total Health Expenditures (in $ PPP). ...................................................75
Table 2.2. Growth Rates of GDP and Health Expenditures ($ PPP)......................80
Table 2.3. Public Share of Total Health Expenditures by Country
(1970 – 2000)...................................................................................................83
Table 2.4. Per Capita Expenditures ($ PPP) on Pharmaceuticals
(10 year interval)..............................................................................................92
Table 2.5. MRIs and CT Scanners per 1 Million Population. ................................94
Table 3.1. Statistical Test Results for all Models. .................................................131
Table 3.2. LSDV Regression Estimates of the Unrestricted Model. (Cross-
Sectional and Time-Series Estimates not Included) ......................................135
Table 3.3. LSDV Regression Estimates of Restricted Model 1. (Cross-Sectional
and Time-Series Estimates not Included)......................................................138
Table 3.4. LSDV Regression Estimates of Restricted Model 2. (Cross-Sectional
and Time-Series Estimates not Included) ……………………………………....141
vi
LIST OF FIGURES
Figure 1.1. Typology of Health Care Systems. .........................................................9
Figure 2.1. Total Health Expenditures per Capita 1970 – 2000 ($ PPP).................74
Figure 2.2. Mean Average Growth Rate of Total Health Expenditures ($ PPP)
by Health System 1970 – 2000........................................................................76
Figure 2.3. Mean Average Growth Rate of Total Health Expenditures ($ PPP)
by Health System per Decade..........................................................................77
Figure 2.4. Mean Average Growth Rate of Total Health Expenditures ($ PPP)
before and after the Introduction of a NHS. ....................................................78
Figure 2.5. Total Health Expenditure as a Percent of GDP (1970-2000)...............79
Figure 2.6. Total Health Expenditures as a Percent of GDP before and after the
introduction of the NHS...................................................................................81
Figure 2.7. Share of Public Health Expenditures by Type of Health Care System
(1970 – 2000)...................................................................................................82
Figure 2.8. Mean Average Growth Rate of Public Financing ($ PPP) by
Health System Type 1970 – 2000....................................................................84
Figure 2.9. Mean Average Growth Rate of Public Financing ($ PPP) of
Health Care Systems by Decade......................................................................85
Figure 2.10. Mean Average Growth Rate of Public Financing before and after
the introduction of NHS after the Introduction of a NHS System...................85
Figure 2.11. Share of Out of Pocket Expenditure 1970 - 2000. (% Total Health
Expenditures)...................................................................................................86
Figure 2.12. Share of Out of Pocket Expenditure in 1970, 1980, 1990 and 2000
by Health System ($ PPP). ..............................................................................87
Figure 2.13. Share of Inpatient Care by Health Care System 1970 - 2000. ...........88
vii
Figure 2.14. Per Capita Inpatient Care Expenditures 1970 – 2000 ($ PPP)...........89
Figure 2.15. Share of Pharmaceutical Expenditures by Health Care
System 1970 - 2000. ........................................................................................90
Figure 2.16. Per capita Pharmaceutical Spending ($ PPP) 1970 – 2000................91
Figure 2.17. Mean Average Growth Rate of Pharmaceutical Expenditures. .........93
Figure 2.18. Kidney Transplants per 100,000 population .......................................95
Figure 3.1. Conceptual Framework: Determinants of Total Health Expenditures110
Figure 3.2. Musgrove's Two Dimensions of Health Care .....................................115
vii
Abstract
This study has three parts. First, health care systems are analyzed along their key
organizational, financing and delivery characteristics. Next, trends in total health
expenditures, source of financing, expenditure structure and technology are
described. The combined results show that the United States’ health care system
very similar to other health care systems with respect to organizational and
delivery characters but substantially differs from the other health care systems in
terms of financing mechanism. In addition, the United States has much higher total
health expenditures and a much smaller public share of total health expenditures
than any other country included in this study.
Finally, based on the findings from the descriptive analysis and building on
previous research, a multivariate analysis using fixed effects focuses on the
relationship between total health expenditures and public share of health care
system in nineteen OECD countries over a thirty-year period. This analysis differs
from previous research in two ways. First, this analysis includes a sensitivity
analysis to test the robustness of the models and, different models are tested
separately as well as in one model. While GDP remains the most important
explanatory variable, public share is inversely related to total health expenditures.
Public integrated health care systems relying predominantly on public provision
viii
have an expenditure-increasing effect, supporting the public choice theory. Third,
consistent negative estimates are obtained for share of health expenditures spent on
pharmaceuticals, supporting the substitution theory. Out of pocket payments have
an expenditure decreasing influence. Global budgets can contain expenditures
while DRGs do not. System that predominantly use fee for service remuneration,
have higher total expenditures compared to other payment methods. Finally, an
ageing population and increased urbanization are positively related to total health
expenditures.
1
Introduction
During the last two decades, governance approaches that emphasized deregulation,
privatization and decentralization have become a common trend in public policy in
general and in health policy in specific. Because of increasing health expenditures,
for many years, industrialized countries have been searching for strategies and
techniques to slow the growth of health care spending from incentives for greater
productivity, efficiency and effectiveness to stimulating free-market competitive
dynamics.
Many international organizations such as the Worldbank (Wang, et al., 1999), the
World Health Organization (Evans et al., 2000; Murray & Frenk, 1999) and the
Organization for Economic Cooperation and Development (OECD, 2002) have
promoted these policies and have reported on a regular basis how countries were
performing.
A common conclusion was that industrialized countries have continuously rising
health expenditures. At the same time, health outcomes such as life expectancy
have improved steadily. However, there are differences among industrialized
countries in terms of expenditures and, higher health expenditures do not
necessarily seem to be related with better health outcomes.
2
The goal of this study is to review health care financing options and the ability of
various approaches to control expenditures. Through the development of a
conceptual model, the goal is to have a better understanding of the structure of
financing sources and to evaluate their differential effects on total health
expenditures at an aggregate level. With respect to total health expenditures, the
emphasis is on whether public funding has as potential to control total health.
Based on a definition of a health care system, a two dimensional typology is
developed to categorize different types of health care systems. Health care systems
and their philosophical foundations are discussed.
Next, the main features of the health care systems of the countries included in the
analysis are described along their key components: the organization of the health
care system, the financing features of the system and the delivery of health
services. Health care systems have evolved over the last three decades and
continue to undergo changes. This part focuses on the characteristics of the health
care systems in the countries included in the further analysis as they were around
1997 – 2000. The chapter concludes with summarizing how health care systems are
alike and how they differ.
3
The second part of the study reviews trends that took place over the last three
decades. Trends in types of health care systems are described with respect to total
health expenditures in absolute as well as relative terms, public expenditures on
health and out of pocket payments. In addition, trends in health care subsystems
such as demand and supply side characteristics as well as expenditure structure are
described.
Since the influence of government involvement in the financing of health care
systems is the major research objective, the three existing types of health care
systems are compared against one another with respect to public financing. Since
some countries made fundamental changes in their health care system, before and
after trends that accompanied the introduction of a National Health Service are
discussed.
Based on the trends described in the second part, the third part starts with
summarizing the results of previous multivariate statistical analysis. This is
followed by constructing and extending a conceptual framework that is based on
previous models tested by Leu (1984) and Gerdtham (1991).
This multivariate analysis differs from previous research in that it attempts to
underpin the empirical results by well-established economic and new institutional
economic theories with respect to the role of government and market by a priori
4
stating hypothesis with respect to the direction of the estimates. A second
difference with previous research is the use of a different regression strategy by
separating different health system components that are tested in separate models as
well as altogether in one model. Third, this study differs by introducing new
variables on the supply side and the expenditure structure. In order to test the
robustness, all models are subjected to a sensitivity analysis by using several (sub)
datasets that contain the same countries but cover different eras. The chapter ends
with summarizing and discussing the results from the analysis.
The study concludes with summarizing the findings from the three parts of the
analysis and reflects on the findings from a United States’ perspective.
5
Chapter 1 : Health Care Systems: Definitions, Typologies
and Individual Health Care Systems Described
After defining a health care system, the definition is used to construct a typology of
health care systems. In the second part, the different types of health care systems
are discussed and countries being part of the analysis are categorized accordingly.
Next, the countries’ health care systems are described along their organizational,
financing and delivery characteristics. This chapter concludes with reviewing
overall similarities and differences of the different health care systems.
1.1. Defining a health care system.
Starting from the premise (Roemer: 1998) that every country has some kind of
health care system that is the result of its history, economic development and
ideology, health care systems have been defined in various ways. Several
approaches are used.
A first approach broadly defines health care systems with respect to illness. While
focusing on treating illness with economic, institutional and human resources,
Field (1993: 10) for example, defines a health care system as a “convention to
6
delineate or to differentiate the totality of formal efforts, commitments, personnel,
institutions, economic resources, research efforts (both basic and applied) that a
nation-state or society earmarks or devotes to illness, premature mortality,
incapacitation, rehabilitation, and other health related problems”.
A second approach is similar to Field with respect to the resources, but rather than
focusing on illness or incapacity of some kind, Roemer (1991:31) focuses on the
delivery of services and defines health care systems as “the combination of
resources, organization, financing, and management that culminated in the delivery
of health services to a population”. Similar to Roemer, but more detailed Lassey et
al. (1997: 3) define a health care system as “the combination of health care
institutions, supporting human services, financing mechanisms, information
systems, organizational structures that link institutions and resources, and
management structures that collectively culminate in the delivery of health services
to patients”. Finally, the WHO (2000) defines health care systems as “A formal
structure for a defined population, whose finance, management, scope and content
is defined by law and regulations. It provides for services to be delivered to people
to contribute to their health … delivered in defined settings such as homes,
educational institutions, workplaces, public places, communities, hospitals and
clinics.”
7
A third approach defines a health care system through its components or
subsystems. Green and Matthias (1997) define health care system through its four
subsystems of policy development, financing, delivery and allocation and quality
of services while Blank and Burau (2004) define a health care system through its
three subsystems: funding, provision and governance. Finally, Matcha (2004: 4)
defines a health care system as “any combination of those components identified
by a society that facilitates the provision of health care for its members”.
All the definitions above have in common that to they make a distinction between
elements determining the delivery of health care and the financing of care.
Therefore, health care systems are here, defined as the institutional arrangements
regarding the delivery and financing of health care – both of which can be public or
private in nature - to a population.
1.2. From definition to typology.
Different typologies have been used to study health care systems comparatively.
These similarities are used to design a typology of health care systems. Maxwell
(1974) for example made a two dimensional typology based upon the degree of
government control of the health care system and the degree of public financing of
the health care system.
8
Roemer (1977, 1991) categorizes the different health care systems according to the
degree of government involvement (entrepreneurial and permissive countries,
welfare oriented countries, universal and comprehensive countries and socialist and
centrally planned countries) and the economic wealthy-ness (affluent countries,
resource rich countries, developing countries and poor countries). Similarly, Elling
(1980; 1989; 1994) too creates a typology in which a distinction is made between
core capitalist countries, social welfare countries, industrialized socialist-oriented
countries, capitalist dependency countries in periphery and semi-periphery and
socialist-oriented countries that are quasi independent of the world system.
The typology used here is based on how a health care system is defined.
Previously a health care system was defined as the institutional arrangements
regarding the financing and delivery of health care to a population. Based on the
Maxwell typology and with respect to the delivery as well as the financing, the
institutions involved vary along a public-private continuum (figure 1.1.). Two
broad categories can be distinguished based on this typology: collective health
insurance schemes and individual health insurance schemes.
National Health Services schemes provide health care through publicly owned
institutional arrangements and finance these health services through resources
generated by government. Most of the times, but not necessarily exclusively,
financial resources for health care are generated through general taxation.
Figure 1.1. Typology of Health Care Systems.
National Health
Service
(NHS)
Not Existing in
OECD
Non - National
Health Service
(Bismarck or NHI)
Entrepreneurial
Model
(Entrepreneurial)
Public Private
Public
Private
FINANCING
D
E
L
I
V
E
R
Y
Non-NHS Systems can be subdivided into two categories: Social Insurance or
Bismarck Systems on the one hand and National Health Insurance systems on the
other. Both have in common that health services are provided privately. The main
difference lays in the way they publicly finance health care systems.
1.3. Comparing Health Care System Types.
At the outset, the individual and collective health care systems start from a
different assumption with respect to the nature of health care as a good. The first
one considers health care as a regular good that can be provided efficiently through
a market mechanism with limited government intervention. The latter rather
considers health care as a non-marketable good that is likely to suffer from market
9
10
failures when distributed through a market. Therefore, this view attributes a larger
role for government in health care.
Whether one considers health as a regular good or a not, has opposite
consequences. (Table 1.1) A first consequence related to the different view of the
nature of the good is that health care in collective health care systems is considered
a basic right for all residents or citizens while individual health care systems see it
as an individual choice to purchase health insurance. Consequently, the first view
virtually covers the total population (universal coverage) since health insurance is
mandatory while the latter potentially covers no individual at all since the
individual has the individual choice to purchase health insurance or not.
Table 1.1. Differences between Individual and Collective Insurance Schemes.
Individual Insurance Collective Insurance
Entitlement Individual Choice Basic Right
Mechanism Personal Interest Solidarity
Membership Voluntary Mandatory
Covered Population Potentially None Potentially All
Covered Benefits Willingness to Pay Need Based Basic Coverage
Rate Setting Individual/group National pool
(Modified from Dror, 2000)
11
A second consequence is related to the financing of the health care system. In
individual health insurance schemes, rates are set at the individual or group level
according to risks, and the extend of the benefits depends on each individual’s
preferences and willingness to pay. Collective insurance schemes are based on
solidarity among individuals and risks are pooled at the national level including
every individual. Financial resources are mainly generated through general
taxation, social insurance or a combination of both.
1.4. Health Care Systems at Glance.
In this section the different health care systems of the countries included in this
study will be described conform to the “HiT” (Health in Transition) Methodology
of the European Observatory on Health Care Systems and Policies, the WHO
Regional Office for Europe. The collection of comparable information on country
health care systems by the Observatory is based on their Production Template and
Questionnaire (1999). The focus of the country-by-country description here is
threefold: General Organizational Characteristics, Financing and Delivery of health
care.
12
1.4.1. National Health Service Systems.
1.4.1.1. Denmark.
General Organizational Characteristics.
For the last thirty years, Denmark’s statutory health insurance has covered its total
population. Opting out is not a possibility but additional private voluntary
insurance supplementing the statutory health care is available (Mossialios &
Thompson 2004).
In Denmark, all three levels of government (central, county and local) are involved
in running the health care system. The national government’s responsibility is
rather limited in that it creates a coordinating framework in which counties and
municipalities operate. In annual negotiations with the Association of County
Councils and the National Association of Local Authorities (non binding) budget
targets are agreed upon and are further filled in at the lower levels of government
(Christiansen, 2002).
What services should be provided and how they organized is determined at the
county level. Counties own and mange the hospital system while municipalities
are responsible for the provision of certain services (Van Kemenade, 1997).
13
Financing of Health Care.
Health care in Denmark is mainly financed through income taxes levied by the
government (Pedersen et al., 2005). County and local governments also receive
additional revenues from property taxes to finance health care (Dixon & Mossialos,
2002). There are no social insurance contributions from the employers, but there
are out of pocket expenses for each individual at the point of use of health services
except for general practitioner and referred hospital visits (Van Kemenade, 1997).
Delivery of Health Care.
General practitioners have private practices and being the first point of entrance to
the health care system they serve as gatekeepers. They are remunerated through a
mix of capitation and fee-for-service. The number of general practitioners as well
as the number of registered patients per general practitioner is fixed through
negotiations between the National Health Service and the physicians’ professional
organization. The choice of Physician is not necessarily limited to the municipality
(Christianson et al., 1999). Specialists either work as salaried employees of public
hospitals or practice medicine in a private practice for the National Health Service
on a negotiated fee-for-service basis (European Parliament, 1998).
14
Most hospitals are public hospitals owned by the county. They provide free
medical care based on need and, are financed through prospective global budgets
based on historical costs and activities. In order to make hospitals more cost
conscious, counties established soft contracts with hospitals that outset the
objectives of both county and hospital (Christianson et al. 1999).
Pharmaceutical products are distributed to individuals through a limited number of
private pharmacies and through hospital pharmacies for the hospital sector. There
is a price ceiling set at the average price of pharmaceutical products in the
European Union (WHO, 2001).
1.4.1.2. Finland.
General Organizational Characteristics.
Finland’s statutory health insurance scheme virtually covers its whole population.
Private insurance is available and is purchased to supplement the coverage of the
statutory insurance (Van Kemenade, 1997).
In Finland, all three levels of government (central, province and local) are involved
in the health care system. However, central and local governments play the most
15
important role. While the central government determines the general policies and
monitors their implementation, municipalities decide on the specific planning and
organization of health services they have to provide (WHO, 2001). Besides a
monitoring role, provinces play also the role of advocates of the central
government’s objectives. Besides the provinces, hospital districts are also
involved at the intermediate level of the health care system. Hospital districts are a
federation of municipalities who own the hospital (European Union, 1998).
Financing of Health Care.
The Finnish health care system is primarily co-financed through national general
taxation and local income taxes. National contributions are transferred as
prospective subsidies to local governments who can tailor them according to the
specific local needs. In addition, there are mandatory employer and employee flat
rate contributions to the National State Sickness Insurance (WHO, 2001). At the
municipality’s discretion, maximum out-of-pocket payments as co-payments can
be charged for certain municipal health services. For short-term hospital stays,
there is a fixed per diem fee (Hurst & Siciliani, 2003).
16
Delivery of Health Care.
General Practitioners working in health centers are gatekeepers and are paid a
monthly salary or a combination of salary, capitation, fee-for-service and local
allowances if they have the status of personal doctor (Häkkinen, 1999).
Individuals are limited in their choice of provider, since they have to register with a
physician in the locality where they reside (van Kemenade, 1997). Physicians are
allowed to work outside the health center in a private practice on a fee for service
basis for which patients are partially reimbursed (Häkkinen
, 2005). Similarly,
specialists are salaried hospital employees who also can practice medicine in a
private practice on a fee-for-service base (WHO, 2001).
The payment of hospitals, which are owned by several municipalities, is primarily
through services purchased by the municipality. Most of the time, the hospital
sends a monthly invoice to the municipality. Prices for services are determined
without guidelines and vary from hospital to hospital. The most common way to
charge municipalities for the services provided is bed days (Van Kemenade, 1997).
There are however fixed maximum prices. Only a few hospitals use Diagnosis
Related Groups as their basis for determining the price of their services (Häkkinen
,
2005).
Pharmaceuticals are sold by privately owned pharmacies to individuals. The retail
price of pharmaceutics is based on a “reasonable” wholesale price, which is a
17
maximum price private and hospital pharmacies have to pay. The retail price is
calculated based on this wholesale price, a state set profit margin and a value-
added-tax. Pharmaceutics with sales permission are reimbursable (European
Parliament, 1998).
1.4.1.3. Greece.
General Organizational Characteristics.
Since 1983, Greece has a compulsory National Health Services system that
coexists with the remainder of its previous social health insurance system (Niakis,
2001). The Greek system covers virtually its whole population (Sissouras et al.,
1999). Private insurance is available and relatively widely purchased (Mossialos
2002).
The Greek health care system is organized at three governmental levels. The
central government’s Ministry of Health advised by the Central Health Council
plays a predominant role and is responsible for planning and organizing the health
care system through the regional centers as well as the coordination of the private
health sector (WHO, 1996). As such, the central government is involved in
legislating almost all aspects of the health care system (Mossialos et al., 2005). At
the regional level, the thirteen Regional Health Councils are responsible for the
18
implementation of the national policies. They also give advice to the central
government about the needs of their population (Van Kemenade, 1997). At the
local level, districts are managed by a central government representative and are
responsible for the provision of health care (WHO, 1996). Finally, public social
insurance funds whose membership is mandatory and based on occupation are
highly regulated and, offer insurance packages including supplemental insurance as
well as health services (Niakis, 2001).
Financing of Health Services.
General taxation and social insurance are the major sources of financing of the
Greek health care system (European Parliament, 1998). Social insurance funds
generate their resource through contributions from employees and employers who
contribute the largest part (Van Kemenade, 1997). Some insurance funds receive
state subsidies in addition to their regular funding sources (Sissouras, 1999). Out
of pocket expenditures are also an important source of financing. Although cost
sharing differs from insurance fund to insurance fund, it generally applies to
services such as private inpatient and outpatient care, public outpatient care,
diagnostic services and pharmaceuticals (Van Kemenade, 1997). Retired and
chronically ill individuals are exempt from co-payments (Sissouras et al., 1999).
Private insurance is available and mainly supplements the insurance fund coverage
with an emphasis on inpatient care (Mossialos, 2002).
19
Delivery of Health Services.
Physicians working for the National Health Services, irrespective of whether they
are general practitioners in a primary care center or specialists in a hospital, are
salaried employees of the state (Niakis, 2001). General practitioners are not
gatekeepers and their salary is determined based on their number of registered
patients (European Parliament, 1998). Private primary practitioners contract with
insurance funds on a fee for service basis (Mossialos et al., 2005).
There are three types of hospitals: Public NHS hospitals, public non-NHS hospitals
owned by the insurance funds and private hospitals (Van Kemenade, 1997). Public
NHS hospitals are receiving their resources from the state and the sickness funds
on a retrospective per diem basis (WHO, 1996). Public non-NHS hospitals and
contracted private hospitals receive a per diem supplemented with fees for the
provision of diagnostic services from the insurance funds. Private hospitals
generate revenues through insurance fund reimbursements, out of pocket payments
and voluntary private insurance reimbursements (European Parliament, 1998).
Pharmaceutical prices are regulated in two ways. First, there is a positive list of
reimbursable drugs based on whether the drugs are available in other EU countries
(Van Kemenade, 1997). Secondly, as is the case in other countries prices are
20
regulated based on the costs of the ingredients, profit margins and other costs such
as promotion and distribution (WHO, 1996).
1.4.1.4. Iceland.
General Organizational Characteristics.
Iceland’s statutory health insurance covers all residents. Although not outlawed,
private health insurance does almost not exist in Iceland (OECD, 2001). The major
actor in Iceland’s health care system is located at the central level where the
Ministry of Health and Social Security is in direct control of all the key functions
of the health care system such as the health facilities approval and the collection
and publication of health statistics (WHO, 2003). The Ministry also is responsible
for determining the policy framework (Van Kemenade, 1997). The State Social
Security Institute administers the health insurance under the supervision of the by
parliament elected Social Security Board (WHO, 2003).
Financing of Health Care.
Health care is mainly financed through personal and corporate income taxes as
well as value-added-taxes and comes directly and solely from the central
21
government’s budget or the State Social Security Institute. The remainder of the
health care cost is complemented by household out-of-pocket payments (WHO,
2003). There are out of pocket expenses for pharmaceuticals, general practitioners
and specialist visits but not for preventive care and inpatient care. Private
voluntary insurance is available (OECD, 2001).
Delivery of Health Care.
A majority of the general practitioners receive are salaried employees of the health
center or hospital. In addition to the salary, they receive a fee for service from the
State Social Security Institute. The salary is calculated the same way civil servants’
salaries are calculated (WHO, 2003). General practitioners are the first point of
entry and serve as gatekeepers. Only a few general practitioners in the capital are
paid on a capitation bases supplemented with a fee for service (Van Kemenade,
1997). Individuals have a free choice of provider (WHO, 2003). Most specialists
practice medicine in a solo practice and, are paid on a fee for service basis. The
fees are negotiated between the health authorities and the medical association.
Most specialists also work as salaried hospital employees on a part time basis. All
hospitals are public and financed through taxes. Hospitals have fixed budgets
(WHO, 2003).
22
With respect to pharmaceuticals, Iceland has a system of reference pricing with
capped detail and wholesale margins. With respect to reimbursement,
pharmaceuticals are categorized according to how essential they are. Essential
drugs are reimbursed fully, while non-essential drugs are not reimbursed at all. In
addition, some prescription drugs are only reimbursed for a maximum period of
time (WHO, 2003).
1.4.1.5. Ireland.
General Organizational Characteristics.
All of Ireland’s residents are eligible for public coverage (Hughes, 1999). There is
however a distinction between two categories of beneficiaries of the national health
insurance. “Category I” residents have an income under a certain ceiling and are
entitled to free medical care from public providers. Residents with an income
higher than the ceiling (Category II) are also eligible for public health services but
have to pay co-payments for ambulatory and hospital care as well as a general
practitioner’s fee. Most of Category II insured purchase highly regulated private
voluntary health insurance that also cover medical care provided by private
providers (Wiley, 2005).
23
At the Central level, the Department of Health is responsible for determining the
general health policy and planning of health services (European Parliament, 1998).
The eight regional health boards manage the public health services (Hughes, 1999).
Health Boards have an annual budget based on demographic characteristics of the
region and are tightly monitored on a monthly basis (European Parliament, 1998).
The central government also regulates the Voluntary Health Insurance Board.
Financing of Health Services.
There are two main financers in the Irish health care system: the central
government and voluntary private health insurance. The central government
generates its resources for its health services mainly through general taxation. In
addition, a health contribution based on the gross income is levied for Category II
insured (Van Kemenade, 1997). The Voluntary Health Insurance Board, a non-
profit state health insurance company, operates private health care and generates its
financial resources through premiums for Category II patients who also pay out of
pocket payments (Wiley, 2005).
Delivery of Health Services.
General Practitioners mainly operate in solo practices and are self-employed.
There is free choice of physician but Category I patients must register with a
24
personal physician who has a contract with the regional health board. Every
general practitioner receives a capitation payment for each registered Category I
patient as well as a lump sum for practice costs. Category II patients pay a fee-for-
service that is determined through negotiations between the Irish Medical
Association and the Ministry of Health. All general practitioners serve as
gatekeepers (Van Kemenade, 1997). Specialists working in a hospital are salaried.
They are allowed to provide medical services in private practice on a fee-for-
service basis that is reimbursable up to a maximum by the Voluntary Health
Insurance (Hurst & Siciliani, 2003).
Public hospitals are funded through prospective global budgets and, for some
hospitals DRGs are used to determine the budg2ets (European Parliament, 1998;
Hughes, 1999). Public hospitals receive their resources from the Health Boards and
predominantly serve Category I patients at no charge. Every public hospital also
has some “private” beds for the treatment of Category II patients who pay on a fee-
for-service basis. Besides public hospitals, there are also public non-profit
hospitals owned by religious orders and private for profit hospitals who
predominantly serve category II patients who have voluntary private health
insurance. While the non-profit hospitals receive subsidies from the Department of
Health, the for-profit hospitals rely on negotiated direct payments from the
voluntary health insurance for their services (Wiley, 2005).
25
The price of prescription pharmaceutical products is set by the Ministry of Health
after negotiations with the pharmaceutical industry and is based on the price in
different European Union member states (Hughes, 1999). Prescriptions are
monitored for Category I patients who receive medicines that are on a positive list
free of charge. Category II patients pay a co-payment (European Parliament,
1998).
1.4.1.6. Italy.
General Organizational Characteristics.
Since 1978, Italy has a statutory Nation Health Service system that virtually covers
its whole population. Additional private insurance is available on a voluntary basis
but is not widely purchased (Hurst & Siciliani, 2003).
The Italian Health Care System is organized at three levels of government (France,
et al., 2005). At the central level, the Ministry of Health determines the national
health policy framework and generates and allocates financial resources to the
regions (Angletti, 2002). The regional authorities are responsible for the
implementation of the national health plan through developing regional health
plans that meet the needs of their geographical population and is conform to the
26
national health plan (Fattore, 199). In addition, they have financing and
monitoring responsibilities of the local health authorities in the region (European
parliament, 1998). Finally, they are responsible for the delivery of health services
and assuring access to those services for their population (Gaydos & Zanola,
2002). At the local level, the local health authorities are responsible for providing
a minimum level of primary health care (Spence, 1996).
Financing of Health Services.
The Italian National Health Services is financed through the National Health Fund
that generates its financial resources equally through general taxation and income
based health contributions (Gaydos & Zanola, 2002). Tax revenues come from
general taxation revenues from the central government (Angeletti, 2002) as well as
from regional taxes (Hurst & Siciliani, 2003). Income based contributions are
regressive (WHO, 2001) with the employer paying the largest part (Van
Kemenade, 1997). Out of pocket payments are another source of financing and
take the form of co-payments for services such as diagnostic procedures,
pharmaceuticals and specialist visits as well as direct payments to private providers
(WHO, 2001). Lastly, private insurance is available on a voluntary basis to
supplement the statutory health coverage (Van Kemenade, 1997).
27
Delivery of Health Services.
General practitioners are self-employed (European parliament, 1998) and are paid
on a capitation basis up to a maximum of registered patients (Gaydos & Zanola,
2002) who are free to choose their physician (European parliament, 1998). They
also serve as gatekeepers to specialist visits and inpatient care (Spence, 1996).
General practitioners working in public hospitals or local health units are paid a
salary (France et al., 2005). When working in public hospitals as a salaried
employee, specialists have the option to work either on a full time or part time
basis (European parliament, 1998). Only specialist working part time in a public
hospital can work in a private hospital (Hurst & Siciliani, 2003). Specialists
practicing medicine in a private practice are paid on a fee for service bases that is
determined by the Ministry of Health (France et al., 2005).
Most inpatient care is provided by public hospitals that receive their resources
through prospective global budgets (Angletti, 2002) based on case mix with
adjustments for historical costs and cross-boundary flows (WHO, 2001). Most
private hospitals are private non-profit organizations and have annual contracts
with local health authorities if they are accredited (Angletti, 2002). Their payment
is similar tot the public hospitals in that it is based on case mix. The main
difference with public hospitals is that private hospitals they can charge patients
additionally for certain services (Van Kemenade, 1997).
28
Pharmacies can be either private businesses who contract with local health
authorities, hospital pharmacies or owned by the local health authorities (Gaydos &
Zanola, 2002). Italy has a positive list of reimbursable pharmaceutical products
(WHO, 2001) based on efficacy, cost, risk-benefit balance and acceptability of
treatment for the patient (Fattore, 1999). Drug prices can be freely determined but
cannot exceed the average European price (WHO, 2001).
1.4.1.7. New Zealand.
General Organizational Characteristics.
New Zealand’s statutory health insurance covers its whole population. Additional
voluntary insurance is available to complement as well as supplement the basic
package covered by the national scheme (Bloom, 2000).
There are two main players in New Zealand’s National Health Service. At the
central level, there is the Ministry of Health that is in charge of the financial flows
in the health care system. The ministry also advises the Minister of Health with
whom it has an annual performance agreement regarding strategic health planning
(Boyd & Sheridan, 2001). Furthermore, the ministry also administers regulation
29
and legislation and monitors its annual performance contracts with the District
Health Boards.
At the regional level, the District Health Boards who are statutory corporations
have to arrange health services for their geographic population. They either can
provide the services themselves or contract third parties. Districts also have to
develop 5-10 year strategic plans in cooperation with the local communities as well
as annual plans and funding agreements with the providers. In order to avoid
duplication of certain functions, “shared services” were established for needs
assessment, contracting and monitoring of health services (WHO, 2001).
Financing of Health Services.
There are four sources for financing health care. The most important source of
financing is general taxation, which is based on income and taxes on goods and
services. A second source of financing is the Social Health Contribution through
the Accident Compensation Corporation. It is a no-fault insurance that covers
medical cost for accident related injuries. Both, the general taxation and the social
health contribution are mandatory. Private health insurance complements and/or
supplements the statutory health insurance coverage. Finally, there are out-of-
pocket payments for medical services. These out-of pocket payments are
subsidized for low-incomes (Kriebe, 2000).
30
Delivery of Health Services.
General practitioners are self-employed and a majority of them practices medicine
in a group practice. They receive their income from several sources. The major
part of their income comes from government retrospective subsidies for patient
visits on a fee-for-service basis. A second source of income comes from
capitation-based contracts for registered lower income patients. Finally, general
practitioners also derive income from patient fees and payments from the Accident
Compensation scheme (WHO, 2001). Individuals have a free choice of health care
provider. General practitioners serve as gatekeepers and patients need a referral to
see a specialist. Specialists working in a public hospital are paid a salary. They
are allowed to practice their specialty in a private practice in addition to their
hospital employment. Specialists in private hospitals are paid on a fee for service
basis.
Most hospitals are publicly owned and receive fixed annual prospective budgets
from the District Health Board which is calculated according to Diagnosis Related
Groups. Adjustments are made for rural areas. (WHO, 2001) Hospital treatment is
free of charge (Bloom, 2001).
Most pharmacies are privately owned local businesses. Pharmaceutical products
must be licensed for distributions. Medsafe, a government agency, licenses
31
medicines based on safety. Pharmac, also a government agency, is in charge of the
price setting and determining the subsidy level as well as the list of reimbursable
pharmaceuticals (WHO, 2002). Patients pay a co-payment for their drugs but
residents with low incomes or serious illness that need frequent treatment can be
exempted. (WHO, 2002)
1.4.1.8. Norway.
General Organizational Characteristics.
All Norwegian inhabitants are cover by the statutory national health insurance,
which is mandatory. Voluntary health insurance is available but not widespread
purchased (WHO, 2001).
The Norwegian health care system is a three-tier system. The central government‘s
Ministry of Health and Social Affairs sets out the legal and policy framework
within which the counties and municipalities have to operate. The National Board
of Health who supervises health services and gives advice with respect to quality
improvement, legal consistency, data collection and information dissemination
(Hagen et al., 1998) assists the Ministry.
32
The 19 counties are grouped into five health regions in order benefit from the
economies of scale for the specialized care they have to provide. They also have to
submit health plans that have to be approved by the ministry. Municipalities are
responsible for primary care and the right of patients to have access to satisfactory
health care in their local community (WHO, 2001).
Financing of Health Services.
The Norwegian health care system is predominantly financed through taxes from
the different levels of government. General taxes from the central government are
transferred as global budgets to health regions and municipalities (WHO, 2001).
The central government also generates income through contributions to the
National Insurance Scheme from employers and employees (OECD 2001).
Local governments also generate revenues for health care through levying income
taxes (van Kemenade, 1997). Treatment in public hospitals is free as opposed to
treatment in private hospitals where patient has to pay the full bill. There is a co-
payment up to an annual maximum for physician visits (WHO, 2001).
Private voluntary insurance to complement the existing statutory health insurance
is available but not purchased by many (WHO, 2001).
33
Delivery of Health Care.
Primary care is provided by general practitioners in group practices. Patients have
a free choice of health provider. Most general practitioners who are also
gatekeepers are self-employed and have a contract with the local government to
provide medical services on a fee-for-service basis and capitation. Some general
practitioners are salaried employees of the municipality (Iverson & Lura, 2000).
Finally, there are general physicians who privately provide medical services. They
can charge a fee-for-service at that they determine themselves. Specialists working
in a public hospital are paid a salary.
Most hospitals in Norway are public hospitals and provide free inpatient services to
Norwegian citizens and residents. For outpatient services, a patient pays a fee-for-
service charge that is partially reimbursable by the National Health scheme. Public
hospitals receive prospective budgets based on case mix (Mikkola et al., 2002) and
annual expected activities from the intermediate government. Additional
earmarked grants are given to teaching and research hospitals and hospitals with
high levels of specialization.
In Norway, there is a difference between drugstores and pharmacies. They differ
in that only the latter have pharmacists and provide prescription drugs, which the
34
first cannot. Pharmaceutical prices are controlled through maximum prices set by
the government (WHO, 2001).
1.4.1.9. Portugal.
General Organizational Characteristics.
Portugal has an amalgam of health care systems. Besides the National Health
Service system that was established in 1979 and provides universal coverage,
Portugal also has special insurance schemes for specific professions (Perreira et al.,
1999). Voluntary private insurance is available to supplement the statutory
coverage (Mossialos, 2004).
The Portuguese National Health Service system is organized in three tiers. The
central government’s Ministry of Health is responsible for developing a health
policy framework as well as its implementation and evaluation (Duaraes, 2003). In
addition, the Ministry maintains its network of publicly owned health care
providers and supervises the private health sector (European Parliament, 2003).
The Ministry also coordinates health related issues with other Ministries and
guides, supervises and coordinates the regional health administrations (WHO,
2004). Regional health administrations have to submit regional health plans that
35
implement the national health plan for approval to the Ministry (Duaraes, 2002).
They also manage and supervise the public hospitals in their geographical area
(WHO, 2004). Regional Administrations are also responsible for integrating
public and private health services (Van Kemenade, 1997) through contracting with
private for profit and non-profit hospitals (WHO, 2004). Local governments are
responsible for the management of primary health care centers (Oliveira & Pinto,
2005).
Health subsystems are remainders of the social insurance system before 1979 and
cover specific occupational groups (mostly public employees) of the population.
Subsystems either provide health services themselves or contract with public
and/or private providers
(WHO, 2004).
Financing of Health Services.
Portugal’s National Health Services system is financed through general taxation
(Oliveira & Pinto, 2005). Health subsystems finance the coverage of their
members through income-based contributions from employees and employers with
the latter paying the largest fraction (European Parliament, 1998). Out of pocket
payment is a third source of financing and applies to physician visits,
pharmaceuticals and diagnostic services. Retired, disabled and chronically ill
individuals are exempt from co-payments (Van Kemenade, 1997). Finally,
36
voluntary private health insurance is available to supplement the statutory coverage
provided by the National Health Service and the health subsystems (Mossialos,
2004).
Delivery of Health Services.
General practitioners and specialist who work for the National Health Service are
paid a salary. They are allowed combine their employment for the NHS with a
private practice for which they are paid on a fee for service basis (Van Kemenade,
1997). Physicians who solely work from their private medical office negotiate
their fees with the NHS, health subsystems and private insurance companies
(WHO, 2004). General practitioners are gatekeepers and their referral is
mandatory for specialist and inpatient care (Duares, 2002). Although individuals
have a free choice of physician, they must register with a primary care physician
(Pereira et. al., 1999).
Most of the impatient care is provided by public hospitals who receive prospective
global budgets based on historical costs and case mix. Private for profit and non-
profit hospitals negotiate contracts with the NHS for the delivery of services
(WHO, 2004).
37
Portugal has a positive as well as a negative list of reimbursable drugs (European
Parliament, 1998). Although pharmaceutical prices are not highly regulated
(OECD, 2004), there is an agreement between the government and the
pharmaceutical industry to voluntary cap the budget which if exceeded the
pharmaceutical industry has to partially pay back the excess (Pereira et al., 1999).
1.4.1.10. Spain.
General Organizational Characteristics.
Spain has switched from a social health insurance scheme to a National Health
Service scheme in 1983 (WHO, 2000). Spain’s current statutory National Health
Service (INSALUD) scheme is mandatory and virtually covers its entire
population, except for the civil servants who have their own social insurance
scheme (Rico & Lazenby, 2002).
The Spanish health care system is mainly organized at two levels of government
(OECD, 2001). The Central government is responsible for setting general health
policies and guaranteeing the right of health care for every citizen through general
coordination and financing (European Parliament, 1998). In addition, the central
government still administers health services in ten of the seventeen regions. In
38
seven regions, regional health authorities (regional health department) manage and
organize health services in their geographic area themselves (Granados & Gallo,
2001). All regions receive the financial resources needed for providing health care
from the national government on a per capita basis. Regions can increase their
health budget through regional taxes (Van Kemenade, 1997).
Financing of Health Services.
The main source of financing is general taxation. Social insurance contributions
are a second source of financing but only apply to civ++il servants (WHO, 2000).
There are also out of pocket expenses. While at the outset, health services are free
regardless of whether they are provided on an inpatient or outpatient basis, co-
payments have been introduced for certain services such as pharmaceuticals, and
inpatient and outpatient private services (Van Kemenade, 1997). Individuals who
are retires are exempted for pharmaceutical co-payments (Rico & Lazenby, 2002).
Finally, voluntary private insurance to supplement the services covered by the
statutory health insurance scheme (Mossialos, 2004).
Delivery of Health Services.
General practitioners are the first contact and serve as gatekeepers to specialists or
hospital care (Hurst, & Siciliani, L, 2003). They are employed by primary care
39
centers and are paid through a mix of salary and capitation (Lopez-Casanovas,
1999). Specialists are salaried employees of hospitals and increasingly receive a
combination of salary and capitation (WHO, 2000). When working in public
hospitals, specialists are not allowed see private patients in that hospital. They are
however allowed to practice medicine in private hospitals. Specialists working
exclusively for public hospitals receive a higher salary than those working in
public as well as private hospitals (Hurst & Siciliani, 2003).
Most hospitals are public hospitals owned by the national or regional government
(Van Kemenade, 1997). All public hospitals receive prospective budgets based on
case mix that are negotiated annually (Lopez-Casanovas, 1999). Private hospitals
of which a majority is for profit (Rico & Laxenby, 2002) can contract with the
NHS to provide health services to meet needs or can provide services for private
insurance to generate resources (WHO, 2000).
Pharmaceutical prices are determined through negotiations between the
pharmaceutical industry and the Spanish government (European Parliament, 1998).
In addition, annual growth is limited to a ceiling. There is a negative list for
reimbursement of pharmaceutics based on their therapeutic value (Lopez-
Casanovas, 1999).
40
1.4.1.11. Sweden.
General Organizational Characteristics.
The Swedish health care system is comprehensive in terms of population coverage.
Private insurance is available to supplement the coverage from the statutory health
insurance. (Saltman, 1998).
At the National level the Ministry of Health and Social Affairs is responsible for
determining the overall policy objectives and fundamental principles of the health
care system (Wall, 1996). The National Board of Health and Welfare assists the
Ministry with respect to the implementation of policy and legislation by the
counties. At the national level, both municipalities and counties are represented by
respectively the Swedish Association of Local Authorities and Federation of
county Councils. Both organizations not only look after their members’ interest
but are also an information channel from lower level of government to central
government (WHO, 2001).
At the regional level, counties are responsible for organizing and financing equal
access to medical services based on the population’s needs with the national
framework. This includes the regulation of the private health care providers.
41
Counties are also responsible for managing primary as well as secondary care
facilities (Van Kemenade, 1997).
At the local level, municipalities are responsible for institutional facilities for the
elderly and the disabled (European Parliament, 1998).
Financing of Health Services.
Swedish health care is financed through four mechanisms (WHO, 2001). Counties
are the most important source of financing health care. In order to provide health
services, they generate financial resource through income taxes (Hurst & Siciliani,
2003). The central government provides block grants generated from income taxes
and indirect taxes to the counties (WHO, 2002). Next, there are mandatory
contributions to the Nation Health Service. These contributions are a mandatory
fixed percentage from the employees’ income paid by the employers (Matcha,
2003). Patients also pay an out of pocket fee for every physician visit. Voluntary
private insurance supplementing the statutory insurance is available but not widely
purchased. If purchased, it often is purchased by the employer (WHO, 2001).
Delivery of Health Services.
Most general practitioners are salaried employees of the county hospitals or
primary care centers (WHO, 2002). Physicians with a private practice can make
42
agreements with the National Health Services County Councils and are paid on a
capitation basis (Matcha, 2003). General practitioners have a gatekeeper role in
that referral is required for inpatient hospital care but not for outpatient health care
(Hurst & Siciliani, 2003). Specialists are employed by the hospital and are paid a
salary. Their time spent in private practice for which they charge on a fee-for-
service charge is limited by their contracts (Annel, 2005).
Most hospitals are public hospitals and receive and negotiated annual budget
(European Parliament, 1998). The way the budget is calculated differs from
county to county. Sometimes prospective budgets are calculated on a historical
cost, sometimes based on DRGs. Some counties created a split between provider
(hospital) and purchaser (county) by creating and internal market (Van Otter,
1999).
The National Agency for Medicine the number of pharmacies. The retail price is
determined by a formula that includes the reasonableness of the wholesale price,
which is determined by the National Board for Drug Reimbursement, and the price
of the drug in other countries (Anell & Svanvar, 1999). Prescription drugs
included on a positive list are reimbursable through the National Health Service but
individuals pay a fixed co-payment per prescription (WHO, 2001).
43
1.4.1.12. United Kingdom.
General Organizational Characteristics.
The United Kingdom has a NHS system and provides statutory health insurance to
all its residents (Van Kemenade, 1997). Private insurance is available and
complements the statutory health coverage (Mossialos, 2004).
The two most important organizations in the UK’s health care system are the
Department of Health and the National Health Service that resorts under the
Department of Health (Oliver, 2005). The Department of Health sets the general
policy framework, sets and plans national health goals and allocates resources to
the NHS (WHO, 1999). The Department of Health is led by the Secretary of State
for Health (Fattore, 1999).
The NHS oversees the provision and delivery of health care in the United Kingdom
(Gaydos & Fries). It also sets health priorities as well as planning guidance (WHO,
1999) and is organized in three tiers (Matcha, 2003). The NHS Executive Regional
Offices receive their financial resources from the NHS and are accountable to the
NHS.
44
Health Care Financing.
The British health care system has four financing sources. The most important
source are the revenues from general taxation (hurst & Siciliani, 2003) at the
national level (Value Added Taxes and income taxes) and local council taxes
(WHO, 2002). Secondly, there are national insurance contributions at a fixed
percent of the income paid by the employee and employer, with the employee
paying a slightly higher percent (Bloor & Maynard). Next, there are also out of
pocket costs for certain services, especially prescription drugs though certain
individuals such as elderly, children, pregnant women and chronically ill are
exempted (Graig, 1999). Finally, private health is available on a voluntary base to
supplement the NHS coverage (Mossialos, 2004).
Delivery of Health Services.
General practitioners are independently operating medical professionals working in
a group setting (primary care group) (Oliver, 2005) and act as gatekeepers for
specialty and inpatient care (WHO, 1999). They are remunerated through different
payment mechanisms. In addition to their budget covering the care of their
registered patients, general practitioners are also paid fees for certain services and
negotiated practice allowances (Matcha, 2003). Specialists can be employed on
either a full or a part time basis by the NHS to practice medicine in NHS trusts on a
45
salary basis. If full time employed, specialists are still allowed to obtain 10 percent
of their income from private practice. This restriction does not apply to part time
employed specialists who can generate an unlimited part of their income from
private practice (WHO, 2002). Specialists are paid on a fee for service basis for
private consultations (Hurst & Siciliano, 2003).
There are private as well as public hospitals (NHS Trusts) in the UK. The latter
became as the result of the purchaser/provider split organizations autonomous from
the NHS (WHO, 2002). They generate their financial resources through
contracting with the Primary Care Groups (Fattore, 1999). Contracts are made for
three years but annual bargaining is possible (Bloom & Maynard). NHS trusts
have besides the majority of public beds also private beds available (WHO, 1999).
Most private hospitals are for profit hospitals and are owned by private insurance
companies and generate their revenues through procedures covered by private
insurance (WHO, 2002).
Pharmaceutical prices are regulated in that profit margins are fixed by the
government although prices are determined by the industry (Duffy, 2002) and,
compliance assures contracting opportunities with the NHS Hurst & Siciliani,
(2003).
46
1.4.2. Non National Health Service Systems.
1.4.2.1. Austria.
General Organizational Characteristics.
The Austrian compulsory health insurance is part of a statutory social insurance
scheme and covers the medical care of its virtually total population (Van
Kemenade, 1997).
The Austrian health care system is constitutionally organized at three levels of
government (Theurl, 1999). At the central/federal level, the Ministry of Health is
responsible for the health care system in terms of outlining the general policy
framework and generating the funds to finance health care. The regional authorities
have each their own Department of Health and are responsible for the
implementation of the policies set by the federal government. They are also
responsible for the provision of hospital and long term care. The Federal
government monitors regional governments’ the hospital investment plans. Local
governments have limited responsibilities (Stepan & Sommersguter-Reichmann,
2005).
47
In addition to the three levels of government, the 27 self-governing social
insurance funds cover the labor force and their dependents. Their tasks are to
administer health insurance and negotiate contracts with representatives of health
care providers regarding the services to be provided and the reimbursement rates.
They are united in the Austrian Federation of Social Insurance Institutions to look
after their interests (WHO, 2001). Membership to the insurance funds is based on
occupation resulting in a restriction of insurer choice (Van Kemenade, 1997).
Financing of Health Services.
The major financing sources of the Austrian health care system are equal
contributions from employee and employer to the social insurance scheme.
Contributions are limited to a maximum percent of the income and taxation
(European Parliament, 1998). Additional financing is generated through out-of
pocket payments and voluntary private health insurance (WHO, 2001). Out-of-
pocket payments differ among occupational insurance funds and apply mostly on
hospital care in the form of co-payments. For prescription drugs, a fixed amount is
charged per prescription (Van Kemenade, 1997). However, lower incomes are
exempted (Stepan & Sommersguter-Reichmann, 2005). Voluntary private health
insurance is available to supplement and complement the statutory health insurance
(Mossialos, 2004).
48
Delivery of Health Services.
Most physicians, whether they are general practitioners or specialist, are self-
employed and their referral is needed to obtain outpatient care (WHO, 2001).
Physicians who have a contract with the insurance funds are paid on fee for
services basis that is negotiated at the national level between the insurance fund
association and the physician association. This master contract is used as the basis
for individual contracts between insurance funds and individual physicians
(European Parliament, 1998). Some sickness funds use a combination of capitation
and fee for service (Theurl, 1999). While the fees for contracted physicians are
fully reimbursed through a voucher system, fees for physicians without a contract
are not totally reimbursed (Van Kemenade, 1997). Physicians working in a
hospital are salaried but can have an additional practice (WHO, 2001).
Most hospital care is provided by public hospitals that are allowed to us a fraction
of their beds for the treatment of private patients (Theurl, 1999). The private
hospital sector is dominated by non-profit hospitals ((Stepan & Sommersguter-
Reichmann, 2005). Hospitals are financed through a combination of prospective
and retrospective payments (WHO, 1999). The prospective global budgets are
determined on a case mix basis with regional adjustments (Theurl, 1999).
49
Pharmaceutical prices are regulated by the central government (European
Parliament, 1998). Maximum prices are set based on estimated costs, international
comparisons (Van Kemenade, 1997) and profit margins for wholesalers and
pharmacies (Theurl, 1999). There is also a positive list of reimbursable drugs set
by the sickness funds. Prescribe drugs on that list are automatically reimbursed,
while those not on the list have to be approved for reimbursement (WHO, 2001).
1.4.2.2. Germany.
General Organizational Characteristics.
Germany has universal and comprehensive statutory health care system that is
eligible for the total population. A typical feature is that individuals above a
certain income can opt out and purchase private insurance (European Parliament,
1998).
The health care system is organized in a three tier governmental setting
complemented with important responsibilities for the quasi public corporations or
sickness funds. The Federal government’s responsibility is defined in the
Constitution. As such, the central government must provide a legislative
framework for the delivery of health services except for hospitals (WHO, 2004). It
includes regulating the sickness funds, determining the benefits, creating guidelines
50
for negotiations between Sickness funds and providers and determining the
financing mechanism (Graig, 1999). The regional government (Länder) is
responsible for supervising the sickness funds and managing the state hospitals and
planning of capital investments (Swami, 2002). The Lander also are responsible
for public health but many have devolved this responsibility to the local
governments (Wörz
& Busse, 2005).
In addition to the government, there are also corporatist organizations (sickness
funds) that are involved in the health care system. These sickness funds operate
autonomously from government but membership is mandatory for employees with
an income under a certain level (Brenner & Rublee, 2002). Employers have a free
choice of sickness fund although they typically have members based on profession
and/or location. In general, a distinction can be made between two broad
categories of funds. On the one hand, there are the primary funds. These are local
funds that cover individuals that do not belong to a particular professional group.
They also include funds that are established by individual employers. On the other
hand, there are the substitute funds whose members are white-collar employees
(Swami, 2002).
51
Financing of Health Services.
Financing takes place predominantly through social insurance and sickness funds
are at the core of this system (WHO, 2004). Sickness funds receive equal
contributions that are a flat percent of the income up to a maximum from
employers and employees below a certain income level (Graig, 1999).
Contribution rates are set annually based on projections (Brenner & Rublee, 2002).
Furthermore, the health care system is also financed by general taxes. Out of
pocket in the form of hospital per diems and co-payments, for certain services are a
third source of financing (Van Kemenade, 1997). Finally, voluntary insurance is
available as a substitute for the statutory for employees above a certain income
(Swami, 2002). For employees with an income under the ceiling, complementary
as well as supplementary insurance is available (Mossialos, 2004).
Delivery of Health Services.
While Germany has no gatekeeper system, referral to hospital care is required.
Most physicians have an independent practice, the remainder works as salaried
hospital employees (Brenner & Rublee, 2002). Ambulatory care is provided by
office-based physicians independent of whether they are general practitioners or
specialists (WHO, 2000). Physicians are paid on a fee for service basis by their
professional associations (Swami, 2002). Fee schedules are negotiated between the
52
regional governments and the professional associations resulting in a global budget
for the professional associations (WHO, 2004).
Two types of hospitals can be distinguished in Germany. With an equal amount of
public and non-profit hospitals, for profit hospitals only provide a minor part of
hospital services (Busse, 2001). All hospitals that are enlisted on Länder plans
receive their budgets from sickness funds through negotiations (European
Parliament, 1998). These target budgets are calculated on case-fees (DRG based),
procedure fees (intervention based) and a per diem (WHO, 2004). If hospitals
exceed the target budget, they have to pay back the difference. If hospitals do not
exceed the target budget, they can keep 40% of the difference (Busse & Howorth,
1999). Private hospitals contracted by the sickness funds are treated similarly
when they treat statutory insured patients. If they do not have a contract, they only
treat privately insured patients (WHO, 2002).
Pharmacies are privately owned businesses or hospital pharmacies (WHO, 2002).
Germany has uniformly set by the federal association of sickness funds (WHO,
2004) as well as a system of reference pricing to determine the upper limit of the
reimbursable part of pharmaceutics (European Parliament, 1998). In addition,
Germany has spending caps on pharmaceutical expenditures (Busse & Howorth,
1999) and physicians are held accountable for overprescribing and are fined when
exceeding the budget (Swami, 2002).
53
1.4.2.3. Japan.
General Organizational Characteristics.
Japan’s health insurance system is based on employment and virtually covers its
total population (Match, 2003). Membership of an insurance fund is mandatory
and opting out is not allowed (Graig, 1999).
The role of the government is limited to monitoring the implementation of the laws
regarding its universality and coverage package as well as setting the fees
(Yaijima, & Takayanagi, 2002). The government also manages the insurance fund
for public employees, teachers and their dependents (Kyosai) and the fund for self-
employed, and employees of small end middle-sized businesses (Seikan). The
local governments manage the insurance fund for the elderly (Roken) (Yoshikawa
& Bhattacharya, 2002). Large private companies manage the fund (Kenpo) of their
employees and dependents (Yoshikawa, 2001).
Financing of Health Services.
The major financing source is social insurance through compulsory fixed rate
payroll contributions in which employer and employee participate equally (Matcha,
2003). Additional financing comes from income taxes levied at the local and
54
national level (Yajima, & Takayanagi, 2002). A final source of financing are out-
of-pocket expenditures by individual patients. Voluntary private insurance is
available but not widely purchased (OECD, 2001).
Delivery of Health Services.
Physicians regardless of whether they are general practitioners or specialists
practicing medicine in a private practice are paid on a fee-for-services basis, while
physicians working in a hospital are salaried (Ikegami & Campbell, 2004). An
official national schedule based on a point system is used to determine the fees for
services covered under the statutory health insurance (Shoji & Hironari, 1999).
General practitioners are not gatekeepers and no referral is needed for hospital care
(Graig, 1999).
A large majority of hospitals are private non-profit hospitals owned by physicians.
Hospitals are not paid through budgets but through a national fee schedule for
inpatient services(Yoshikawa & Bhattacharya, 2002). There are only a few of
public and teaching hospitals that are subsidized by public funds and provide more
advanced care (Yajima, & Takayanagi, 2002).
Pharmaceutical fee schedules are revised biannually and are calculated based on
the market price and a profit margin for health providers (Yoshikawa, 2001).
55
1.4.2.4. The Netherlands.
General Organizational Characteristics.
The Dutch health care system is a mixed system of private and statutory health
insurance that virtually covers its total population (Hurst & Siciliani, 2003).
The national Ministry of Public Health, Welfare and Sports oversees the health
care system with respect to the planning of health services and the implementation
of the health policy. In addition, the ministry is responsible for the statutory social
insurance for which it determines the income related premiums (European
Parliament, 1998). The Minister of Health must approve capital investments, the
establishment of new hospitals, and the negotiated physician fees and hospital rates
(Abel-Smith, B. & Mossialis, E., 1994).
There are three schemes of health insurance in the Netherlands: two (public)
statutory health insurances and one private. The first (public) statutory health
insurance covers the whole population regardless of income for exceptional
medical expenses including long-term care, care for the disabled and home care
(OECD, 2001). A second (public) statutory health insurance covers elementary
health care and is mandatory for all individuals with an income below a certain
level including self-employed (Okma, 2001). This type of health insurance is
56
provided through thirty sickness funds that are regionally organized and they
cannot deny coverage to any eligible individual. Sickness funds cannot provide
medical services themselves but must contract with providers and compete with
one another for members beyond their geographical area. They also offer
supplemental coverage for their members (Graig, 1999). Employees and self-
employed with an income above the ceiling, may opt out of the statutory health
insurance scheme and purchase private health insurance. Private insurers must
offer a basic package that is similar to the coverage offered by the sickness funds to
all eligible individuals (van het Loo, et al., 1999). Private insurance companies
also offer supplemental health insurance, which they can refuse to sell to high-risk
individuals (WHO, 2004).
Financing of Health Services.
The Dutch financing of health care is in accordance with the different schemes,
each having their own main financing source (WHO, 2004. Health care is largely
financed through social insurance contributions. Exceptional medical expenses are
financed through a flat percentage of the taxable income up to a ceiling paid by the
employer (Van Kemenade, 1997). Normal medical care for individuals under the
income ceiling is financed by two sources. On the one hand, there is a flat rate
contribution collected by the sickness funds and paid for by the employee. On the
other hand, there is percentage of the income that is paid by the employer and
57
employee with the latter paying the largest share (Ottewill, R., 1996). Individuals
with an income above the ceiling can purchase private insurance. There are also
out of pocket expenses for exceptional medical expenses up to a monthly
maximum, for certain services and for prescription drugs (WHO, 2004). Taxes
play a minimal role in financing the Dutch health care system although government
provides additional financial resources through grants (Schut & van de Ven, 2005).
Delivery of Health Services.
Primary care is provided by self-employed general practitioners who are
gatekeepers for specialist and hospital care (Ottewill, 1996). Individuals insured
by the sickness fund must register with a by the sickness fund contracted general
practitioners who is paid on a capitation basis (Okma, 2001). Patients covered by
private insurance do not have to register and their physician is paid on a fee for
service basis (European Parliament, 1998). A majority of specialists is paid on a
fee for service basis in hospitals (Van Kemenade, 1997) for both patients covered
by private insurance and sickness funds but with different schemes (Graig, 1999).
However, there are target expenditure caps at the national and hospital level for
specialist care and, hospital management and specialist units negotiate on the
volume to be provided (WHO, 2004). While working for a hospital, specialists are
allowed to have a private practice (Hurst & Siciliane, 2003).
58
Only a small fraction of inpatient care is provided by public hospitals and, since
profit hospitals are outlawed (Graig, 1999), the vast majority of hospitals are non-
profit organizations (Schut & van de Ven, 2005). Hospitals receive a target budget
that is based on fixed and variable costs. Fixed costs are related to location or
population covered by the hospital costs as well as infrastructure costs based on
beds and specialty units. Variable costs are related to capacity (number of beds)
and volume related costs (admissions) (WHO, 2004).
There is positive as well as a negative list of reimbursable pharmaceuticals that
apply for the privately insured and those insured by the sickness funds (Van
Kemenade, 1997). In addition, there are maximum prices for pharmaceutical
products outside the hospital through a reference price system that is based on the
European average for drugs treating the same illness (WHO, 2004).
1.4.2.5. Australia.
General Organizational Characteristics.
Australia’s statutory health insurance (Medicare) covers of medical treatments for
all residents regardless their nationality (Wall, 1996).
The health care system is organized at three governmental levels. The
Commonwealth (central government) is responsible for funding and regulating the
59
health care system as well as determining the general policy framework (WHO,
2001). At the intermediate level, six State and two Territory governments are
responsible for the provision and administration of health services and the strategic
planning of their health services. Every five year, Australian Health Care
Agreements are negotiated between the Commonwealth and the State and Territory
governments. In these agreements, prospective block grants are determined for the
public hospitals. The local level of government is responsible for public health
issues (Wall, 1996).
Financing Health Services.
Financing the health care system is a shared responsibility between the
Commonwealth and the State (Podger, 1999). The Australian health care system is
predominantly financed through a mix of general taxation and mandatory tax based
health insurance levied on higher incomes (WHO, 2001). In addition, there are out
of pocket costs mainly for not covered pharmaceutics but also for fees charged by
physicians to the patient directly (WHO, 2002). Finally, voluntary private
insurance is available to supplement and complement the standard Medicare
insurance,
60
Delivery of Health Services.
General practitioners who are gatekeepers are self-employed and work on a fee for
service basis (WHO, 2002). They also have the choice to bulk bill Medicare
directly at 85% of the fee schedule, making health care free for patients (WHO,
2001). Patients have a free choice of physician without mandatory registration
(hurst & Siciliani, 2003). Specialists are employed by hospitals. They can either
be salaried hospital employees or private visiting specialists who are either paid on
a fee for service basis or an agreed upon service contract (WHO, 2001).
Most Australian hospitals are independent government enterprises. Hospitals
receive global prospective budgets from the government (WHO, 2002). These
budgets are calculated based on case mix, hospital costs and demographic
characteristics of the area the hospital services (Hurst & Siciliani, 2003).
Pharmaceutical prices are determined through negotiations between the central
government and the wholesalers (WHO, 2003). The basis for determining the
price of drugs is mainly the cost based. There is a positive list (Pharmaceutical
Benefit Scheme) of drugs that are subsidized by government although there are co-
payments (Wall, 1999).
61
1.4.2.6. Canada.
General Organizational Characteristics.
The Canadian health care system (Medicare) provides statutory health insurance to
all its residents (Graig, 1999). Voluntary private health insurance is available to
supplement Medicare coverage (Match, 2003).
The health care system of Canada is organized at the federal and the provincial
level. The federal government (Department of Health) partially finances the health
care system through transferring financial resources to the provinces (Hurst &
Siciliani, 2003). In addition, it determines and oversees the compliance with the
principles on which the Canadian health care system is built (Deber & Baranek,
1998). These principles are universality, accessibility, comprehensiveness,
portability and public administration (Patel & Rushefsky, 2001). The federal
government also provides health care directly to certain population groups such as
the military, mounted police, native Canadians and federal inmates (Noseworthy,
2001). At the provincial level, the health departments manage the health services
through planning, financing and evaluating the health services that are part of the
basic coverage package mandated by the federal government (Match, 2003).
Provincial governments also fund public hospitals and negotiate salaries and fees
with health professional associations (WHO, 1996).
62
Financing of Health Services.
Taxation is by far the largest financing source of the Canadian health care system.
Taxes are levied on corporate and individual incomes by the federal as well as the
provincial government (Goldsmith, 2002). Some provinces use additional financial
resources such as sales taxes and lottery revenues that are nominally targeted for
health care while other provinces impose health care premiums (OECD, 2001).
There are out of pocket expenditures for individuals seeking health care services
that are not covered by the provincial health plan and for which they do not have
supplementary voluntary private health insurance and pharmaceuticals (WHO,
1996). A final source of financing health care is private insurance, which by
federal law cannot offer parallel health insurance but is restricted to offering
supplemental insurance (Graig, 1999).
Health Delivery Characteristics.
Primary care is provided by general practitioners and family practitioners in private
practices on a fee-for-service basis paid directly by the province (WHO, 1996).
Fees are determined through negotiations between the provincial government and
the provincial medical associations (Match, 2003). However, some provinces have
hard budgetary caps for physician expenditures resulting in fee reductions for the
physicians if the cap is exceeded. Canadians have a free choice of physician who
63
serves as gatekeepers for inpatient care (Goldsmith, 2002). Specialists are treated
similar to primary care physicians in that they practice medicine in a private office
and are paid on a negotiated fee for service basis (Hurst & Siciliani, 2003).
A large majority of Canadian hospitals are non-profits (WHO, 1996). There is no
uniform mechanism of financing hospitals in Canada. Depending on the province,
the hospital financing methods are: the population based method that takes into
account the population profile of the province to determine the hospital budgets,
the global budget based on historical costs and case mix, the line by line method
determining the financial resources by item line, program or department and, the
ministerial discretion method that is based on hospitals’ requests (Hurst &
Siciliani, 2003).
Prices of patented prescription drugs are regulated in Canada (Goldsmith, 2002).
In addition, different provinces have different policies with respect to
pharmaceuticals varying from out of pocket costs, the use of formularies and
managed drug plans (Graig, 1999).
64
1.4.3. Entrepreneurial Health Care Systems.
The United States.
General Organizational Characteristics.
The United States’ health care system is a pluralistic system predominantly relying
on voluntary private insurance with statutory insurance for a limited part of the
population. Medicaid and Medicare provide statutory coverage for the poor and
elderly (Iglehart, 1995). Medicaid is a state and federal program. Medicare is a
federal program that consists of two parts. Part A is mandatory and covers certain
inpatient care and home care while Part B is optional and partly covers outpatient
care and medical equipment including pharmaceuticals. Public programs cover a
relatively small portion of the population.
In addition to the public insurance programs, private insurance is the predominant
insurance coverage. It is most of the times provided by the employer as fringe
benefit with employers paying part of the insurance (Kuttner, 1999). Besides the in
importance declining indemnity insurance, insurance is provided through managed
care organizations.
65
These managed care organizations negotiate contracts with providers or provide
health services themselves. Three types of managed care organizations can be
distinguished. First, there are the Health Maintenance Organizations plans that
provide health services for their enrollees who pay a monthly enrollment fee. A
second type are the Preferred Provider Organizations who negotiate contracts with
health providers that will offer health services to their enrollees. Both, HMOs and
PPOs have in common that the patients’ choice of health care provider is limited.
Finally, there are the point of service plans in which enrollees have the choice of
provider but have to pay less when a contracted provider is chosen (Lassey et al.,
1997).
Employers can also choose self-insurance in which they organize their own
insurance coverage (Upshaw & Mathews, 2002). Both state and federal
government have legislative power to regulate the private insurers (OECD, 2001).
Since health insurance is voluntary, the United States has a significant and growing
number of uninsured. The proportion of uninsured grew from 14.2% in 1995 to
16.1 percent in 1997 (Kuttner, 1999).
Financing of Health Services.
Statutory coverage is financed by the federal government for the Medicare
enrollees and by the state and federal governments for individuals eligible for
66
Medicaid. Medicare is financed through payroll taxes and retiree supplemental
premiums, while Medicaid is financed by the state supplemented by federal
matching grants (Iglehart, 1999).
Private insurance is most of the time paid for by the employer and partly by the
employees. Additional financing comes from out of pocket payments and
philanthropy (Pohl, 2002).
Delivery of Health Services.
With the exception of physicians working in public hospitals, general practitioners
and specialists operate in a private practice setting. Most of the time general
practitioners have a gatekeeping function. The way specialists and general
practitioners are remunerated depends on the type of health insurance their patients
have. Indemnity insurance predominantly pays on a fee for service bases and
managed care organizations either pay a salary or pay the physicians through
capitation. With respect to Medicare, specialists are paid according to a
prospectively set reimbursement rate (resource-based relative value scale) (Graig,
1999).
In the United States, a minority of hospitals is owned by the federal, state or local
governments. While many hospitals are for profit organization, there is also a
67
substantial number of not for profit community clinics. With the latter more
focusing on short-term care (Sekhri, 2002). Hospitals are also financed through
different mechanisms: fee for services, per diem and fee per case. Medicare pays
hospitals prospectively through Diagnosis Related Groups (Lassey et al., 1997).
Pharmaceutics are provided through the market and the government does not set
maximum prices or reference prices. However, at the level of the health plan,
prices are negotiated and there are drug formularies and pharmaceutical benefit
management systems (Chandler, 1996).
1.5. How Are Health Care Systems Different and Alike?
1.5.1 The public and private roles in the health care system.
In all types of health care systems, government plays an important role with respect
to legislation and setting the broader framework for the health care system in which
it operates. Compared to the United States where the role of the government tends
more toward creating a working market, governments in countries with a NHS and
countries without an NHS play more active role in terms of planning, creating a
policy framework and determining prices in health care sub sectors.
68
In general, most health care systems are organized at three levels of government
with the central level mapping out the general policy framework and planning.
Intermediate governments are responsible for the implementation of the central
governments’ policies and supervise the local governments if they have a role in
the health system. Most local governments, if they have a role in the health care
system, are responsible for the provision of certain services.
With respect to financing, most NHS countries as well as the United States
(Medicaid and Medicare) rely on several levels of government to finance their
health care system. While Denmark relies almost solely on local governments,
Ireland, Iceland and New Zealand only generate financial resources through the
central government. Most countries that do not have a NHS rely on the central
government to finance their health care system through social insurance
contributions. The two exceptions here are Canada and Australia where
respectively provinces and states participate in health system financing.
Sources of financing also differ among the different health care systems. All
countries generate resources through taxation. In the United States, taxes are used
to finance public insurance schemes. All countries with the exception of Denmark,
Iceland, Australia and Canada rely on taxes and health related contributions to
finance their health care system . Countries with a Social Insurance system
predominantly fund their health care system through contributions from employers
69
and employees who most of the times contribute equally. All countries that
switched to a National Health Services still have financing remainders from their
previous social insurance system.
While all countries have private involvement in financing the health care system,
there are differences in the degree of private financing. All countries have out of
pocket payments contributing to the financing of the health care system.
While the United States relies predominantly on private health insurance to finance
health care, all other countries have limited private financing through private
insurance. Private insurance is available in all countries and is offered to
complement statutory insurance to cover out of pocket expenses or to supplement
statutory health insurance to finance elective medical services.
1.5.2 The delivery in the health care system.
While physicians in most NHS countries and in the United States are more likely to
be remunerated on a salary and/or capitation basis, physician payments in non-
NHS countries are predominantly paid on a fee for service basis. Fees are the
result of a negotiation process between government and professional physician
associations and/or social insurance organizations. Physicians in NHS countries
are sometimes treated the same as public servants as in the case in Denmark,
70
Finland, Iceland Portugal, Spain and Sweden. However, in Ireland, New Zealand,
Greece, Italy and the United Kingdom they are considered private practitioners, as
is the case in the United States and countries without a NHS. In most countries,
general practitioners are gatekeepers to specialist or inpatient care. Specialists that
are considered public employees are allowed in most countries with a NHS system
to practice medicine part time in a private practice.
In NHS countries with exception of Ireland and New Zealand a large majority of
inpatient care is provided by public hospitals. In non-NHS countries with the
exception of Japan and the Netherlands, public hospitals still provide a substantial
amount of inpatient care. The remainder of most of the hospital care in NHS and
non NHS is countries is provided by not for profit hospitals with a minor share of
hospital care provided by investor owned hospitals if they are not outlawed as is
the case in Japan and the Netherlands. In the United States most hospitals are
private for profit organizations.
With respect to the financing of hospitals, all countries with the exception of Japan
have some kind prospective payment system based on Diagnosis Related Groups.
In all countries, except for the United States and Japan, DRGs are combined with
global budgets.
71
All countries with the exception of the United States regulate pharmaceutical
prices. Most of the time maximum prices are set through reference pricing taking
into account the prices of drugs in other countries or calculating the prices based on
the costs of ingredients and profit margins for wholesalers and detail salers. All
countries use positive and/or negative lists of reimbursable drugs. While in most
countries, governments decide which pharmaceuticals are on these lists, in the
United States this is determined by the health plans.
1.5.3 The individual in the health care system.
In the entrepreneurial health care system of the United States, individuals have at
least a choice among insurance plans if the employer offers health insurance. Once
a health plan is chosen, the choice of provider may be limited if the health plan is a
managed care plan.
In countries with a NHS system, statutory health insurance provides an insurance
package that is the same for every individual and there is no choice between
alternative plans. However, additional voluntary private insurance is available and
for this type of insurance individuals have a choice between different insurers.
With respect to the choice of provider, all but one country restrict the individual
72
choice of provider. Only Finland limits the choice of general practitioner to the
geographical boundary in which one resides.
In non-NHS countries that have a Bismarck health care system, individuals have a
choice of insurer with the exception of Austria where membership of an insurance
fund is mandatory linked to occupation.
73
Chapter 2 : Trends in selected OECD Countries 1970-2000
The previous chapter has demonstrated that health care systems can be
distinguished along two dimensions: the publicness and privateness of financing
on the one hand and provision on the other. This chapter reviews trends that took
place in the different health care systems from 1970 through 2000.
2.1. Trends in Health Total Expenditures.
2.1.1. Total Health Expenditures per Capita 1970 – 2000.
For countries that did not change their health care system, total health expenditures
in dollars PPP, have increased steadily over the last 30 years (Figure 2.1). Annual
per capita total health expenditures increased from 232.9 dollars in 1970 to 3,008.8
dollars in 2000. The United States with an entrepreneurial health system had the
highest per capita health expenditures increasing from $ 347 dollars in 1970 to $
4,540 in 2000. Countries with a NHS system had the lowest per capita health
expenditures. Their average per capita health expenditure grew from $160 in 1970
to $2,107 in 2000. Countries without an NHS closely followed the countries with
an NHS system with total per capita health expenditures rising from $191.7 in
1970 to 2,379.2 in 2000.
Figure 2.1. Total Health Expenditures per Capita 1970 – 2000 ($ PPP).
0
1,000
2,000
3,000
4,000
5,000
1970 1975 1980 1985 1990 1995 2000
Year
$ PPP
NHS Countries
Non NHS Countries
United States
Source: OECD Health Data 2003
Not only are their differences between the different types of health care systems,
within each type there is variation among the countries (Table 2.1). Within the
group of countries with an NHS, Scandinavian countries with the exception of
Finland have the highest per capita expenditures exceeding $ 2,000 while the other
countries in that group stayed well below $ 2,000 per capita in 2000. Since 1980,
Scandinavian countries with the exception of Finland have been the highest
spenders among the NHS countries. Compared to the United States most countries
with a NHS system spend less than half the amount on health.
74
75
In 2000, among the countries with no NHS based health care system Japan had the
lowest per capita annual health expenditures and Germany the highest. All other
countries had per capita expenditure within a range of $ 250. With Canada and
Germany alternating as highest per capita spenders in countries without an NHS,
Japan had the lowest health expenditures since 1970. In 2000, with the exception
of Germany and Canada, all countries have spent about half of what the United
States spent on health care.
Table 2.1. Total Health Expenditures per Capita(in $ PPP).
1970 1975 1980 1985 1990 1995 2000
NHS System
Denmark - 489 819 1177 1453 1880 2398
Finland 161 303 509 847 1295 1414 1699
Iceland 137 288 576 947 1376 1823 2562
Ireland 99 233 452 590 719 1223 1793
New Zealand 177 354 458 587 937 1238 1611
Norway 132 297 632 915 1363 1865 2787
Sweden 270 465 850 1171 1492 1680 2195
United Kingdom 144 271 445 671 977 1330 1813
Average 160.0 337.5 592.6 863.1 1201.5 1556.6 2107.3
Non NHS System
Australia - 435 658 985 1300 1778 2350
Austria 159 358 662 813 1204 1831 2233
Canada 255 428 709 1192 1674 2115 2580
Germany 223 475 824 1242 1600 2263 2780
Japan 130 262 523 818 1082 1581 1984
Netherlands - 396 668 896 1333 1787 2348
Average 191.8 392.3 674 991 1365.5 1892.5 2379.2
Entrepreneurial System
United States 347 586 1055 1759 2738 3654 4540
Grand Average 232.9 438.6 773.9 1204.4 1768.3 2367.7 3008.8
Source: OECD Health Data 2003
While per capita total health expenditures have increased for all countries and all
types of health care systems over a 30-year period, the mean average growth rate
(MAGR) (Figure 2.2.) over that same period is similar for the different health care
systems (8.68 % for Non NHS countries, 8.78 % for NHS countries and 8.95 % for
United States’ entrepreneurial system).
Figure 2.2. Mean Average Growth Rate of Total Health Expenditures by Health
System 1970 – 2000.
8.78
8.68
8.95
0
1
2
3
4
5
6
7
8
9
10
NHS Countries Non NHS Countries United States
Source: OECD Health Data 2003
As opposed to the long term MAGR, 10 year MAGR sheds a different light on
total per capita health expenditures (Figure 2.3). The seventies have the highest
MAGR followed respectively by the eighties and nineties. There are however
differences between the different types of health care systems for each era. During
the seventies, the MAGR is the lowest for the United States at 11. 8 %. Countries
with a NHS system and those without a NHS had a similar MAGR of respectively
76
13 % and 13.3 %. During the eighties, compared to the seventies, the United
States’ MAGR dropped 1.8 % to 10 %. Despite this drop, the United States had the
highest MAGR compared to the other two health care systems. Both, countries
with a NHS system and countries without a NHS almost cut their MAGR in half in
the eighties as compared to the seventies. Their MAGR dropped respectively
from 13.3 % to 7.4 % and from 13 % to 7.3 %. In the nineties as compared to the
eighties, the United States was able to reduce its MAGR by half from 10% to 5.2
% making it the health system with lowest MAGR for that decade.
Figure 2.3. Mean Average Growth Rate of Total Health Expenditures by Health
System per Decade.
13.3
7.4
5.9
13.0
7.3
5.8
11.8
10.0
5.2
0
2
4
6
8
10
12
14
1970-1980 MAGR (%) 1980-1990 MAGR (%) 1990-2000 MAGR (%)
Growth
Rate (%)
NHS Countries
Non NHS Countries
United States
Source: OECD Health Data 2003
While still reducing their MAGR compared to the nineties, countries with a NHS
system and countries without a NHS based health care system have a higher
MAGR than the entrepreneurial health care system of the United States in the
77
nineties. Their MAGR for the nineties was respectively 5.9 % and 5.8% compared
to 5.2% for the United States.
Four countries have changed their health care system from a non NHS system to a
NHS system (Figure 2.4). These countries are Greece (in 1983), Italy (in 1979),
Portugal (in 1978) and Spain (in 1983). While no data is available for Greece and
Italy before the switch to a NHS system, Portugal and Spain have experienced a
drop in MAGR almost by half. Portugal’s MAGR dropped from almost 20 %
before the introduction of a NHS to 10.78 % afterwards. Spain’s MAGR dropped
from 14.24 % before to 7.69 % after the switch to a NHS system. Greece and Italy
respectively had a MAGR of 8.2 % and 5.48 % after their NHS was established.
Figure 2.4. Mean Average Growth Rate of Total Health Expenditures before and
after the Introduction of a NHS.
19.86
14.24
8.20
5.48
10.78
7.69
0
5
10
15
20
25
Greece Italy Portugal Spain
Growth
Rate(%)
Before NHS
After NHS
Source: OECD Health Data 2003
78
2.1.2.Trends in Total Health Expenditures Relative to GDP.
Total health expenditures in terms of percent GDP (Figure 2.5), increased over the
last three decades on average from 6 % in 1970 to nearly 10 % in 2000. Total
health expenditures as a percent of GDP in countries with a NHS system grew
from 5.2 % in 1970 to 7.7 % in 2000. For that same period, countries with no NHS
system experienced a growth from 5.8 % to 8.8 %.
Figure 2.5. Total Health Expenditure as a Percent of GDP (1970-2000).
0
2
4
6
8
10
12
14
1970 1975 1980 1985 1990 1995 2000
Year
NHS Countries
Non NHS Countries
United States
% GDP
Source: OECD, Health Data 2003
While spending only about one percent more than the other health care systems in
1970, the United States’ health expenditures in terms of percent GDP skyrocketed
to almost 13.1% in 2000. This is almost double the percentage of the other
systems.
79
80
When growth rates of GDP and total health expenditures are compared, GDP
growth rates outpace the growth rates of total health expenditures for all health care
systems (Table 2.2.). However, there are decade wise differences. For countries
with a NHS system, the average difference between GDP growth and total health
expenditure growth in the seventies was 2.95 %, compared to 2.47 % for the
United States and 2.86 % for countries with no NHS system.
Table 2.2. Growth Rates of GDP and Health Expenditures.
MAGR
1970-1980
MAGR
1980-1990
MAGR
1990-2000
GDP
Health
Expenditures
GDP
Health
Expenditures
GDP
Health
Expenditures
NHS System
Denmark 9.36 11.12 6.64 5.90 5.33 5.14
Finland 10.71 12.18 7.51 9.78 4.39 2.75
Iceland 12.94 15.43 6.40 9.09 5.03 6.42
Ireland 10.80 16.40 8.09 4.75 9.16 9.57
New Zealand 8.45 9.95 5.68 7.43 4.12 5.56
Norway 11.71 17.01 6.82 7.98 7.46 7.42
Sweden 9.10 12.13 6.61 5.79 3.68 3.93
United Kingdom 9.47 11.95 7.49 8.19 4.39 6.38
Average NHS 10.32 13.27 6.91 7.36 5.45 5.90
Non NHS
System
Australia 8.75 12.06 5.92 7.04 4.67 6.10
Austria 11.14 15.30 6.92 6.16 5.17 6.37
Canada 10.53 10.76 6.42 8.97 4.24 4.42
Germany 10.11 13.95 7.08 6.87 3.38 5.68
Japan 10.91 14.96 8.45 7.54 3.57 6.26
Netherlands 9.60 11.23 6.48 7.15 5.06 5.83
Average Non
NHS
10.17 13.04 6.88 7.29 4.35 5.78
Entrepreneurial
System
United States 9.30 11.77 6.59 10 4.15 5.19
Grand Average 9.93 12.70 6.79 8.22 4.65 5.62
Source: OECD, Health Data 2003
During the seventies, the average difference between GDP growth and total health
expenditure growth for NHS countries was 2.95 percent points. In the nineties, the
difference dropped to .45 percent points. A similar trend took place for countries
not having a NHS system whose difference dropped from 2.87 to 0.41 and the
United States dropping from 2.77 to 0.97. All health care systems have in
common that the difference got smaller over a period of thirty year. With the
exception of countries without a NHS, both the United States and the NHS
countries experienced a continuous reduction of the difference between GDP
growth and health expenditure growth.
Countries that changed their health care system from a social insurance system to a
NHS system saw their total health expenditures in terms of share of GDP increase.
Figure 2.6. Total Health Expenditures as a Percent of GDP before and after the
introduction of the NHS.
6.60
4.14
4.95
8.52
7.87
6.81
6.73
0
1
2
3
4
5
6
7
8
9
10
Greece Italy Portugal Spain
Percent
Before NHS
After NHS
Source: OECD Health Data 2003
81
Among those countries (Figure 2.6), Portugal had the highest increase of 2.67
followed by Greece (1.92) and Spain (1.78).
2.2.Trends in Source of Health Expenditures.
2.2.1.Public Share of Total Health Expenditures.
The public share of total health care expenditures is the highest in countries with a
NHS system (Figure 2.7). From 1970 until the early eighties, the average public
share increased from 83.2 % to 85.6 %. This trend was followed by a decrease to
an average public share of 80.5 % in 2000.
Figure 2.7. Share of Public Health Expenditures by Type of Health Care System
(1970 – 2000).
0
10
20
30
40
50
60
70
80
90
100
1970 1975 1980 1985 1990 1995 2000
Year
NHS Countries
Non NHS Countries
United States
Percent
Source: OECD Health Data 2003
82
83
Within each type of health care system (Table 2.3), there are some country wise
variations. Scandinavian countries (with the exception of Finland) and the United
Kingdom had the highest public share in the group of countries with a NHS
system. Over the last 20 year, most countries experienced a decline in the public
share of total health expenditures.
Table 2.3. Public Share of Total Health Expenditures by Country (1970 – 2000).
1970 1975 1980 1985 1990 1995 2000
NHS System
Denmark - 85.4 87.8 85.6 82.7 82.5 82.5
Finland 73.8 78.6 79 78.6 80.9 75.6 75.1
Iceland 81.7 87.2 88.2 87 86.6 84.5 84.4
Ireland 81.7 79 81.6 75.7 71.9 71.6 73.3
New Zealand 80.3 73.7 88 87 82.4 77.2 78
Norway 91.6 96.2 85.1 85.8 82.8 84.2 85
Sweden 86 90.2 92.5 90.4 89.9 86.7 85
United Kingdom 87 91.1 89.4 85.8 83.6 83.9 80.9
Average 83.2 85.2 86.5 84.5 82.6 80.8 80.5
Non NHS System
Australia - 73.1 63 71.4 62.5 66.7 68.7
Austria 63 69.6 68.8 76.1 73.5 70.9 69.6
Canada 69.9 76.2 75.6 75.5 74.5 71.4 70.4
Germany 72.8 79 78.7 77.4 76.2 80.5 78.8
Japan 69.8 72 71.3 70.7 77.6 83 81.3
Netherlands - 67.9 69.4 70.8 67.1 71
Average 68.9 73.0 71.1 73.7 71.9 73.9 73.8
Entrepreneurial System
United States 36.4 41 41.5 39.9 39.6 45.3 44.4
Grand Average 62.8 66.4 66.4 66.0 64.7 66.7 66.2
Source: OECD, Health Data 2003
Mean average growth rates of per capita $ PPP are similar for the different types of
health care systems (Figure 2.8). From 1970 to 2000, the United States had the
MAGR of 9.6% followed by countries without NHS Systems and countries with a
NHS system with respectively a MAGR of 9.11% and 8.87%.
Figure 2.8. Mean Average Growth Rate of Public Financing by Health System
Type (1970 – 2000).
8.87
9.11
9.66
0
1
2
3
4
5
6
7
8
9
10
NHS Countries Non NHS Countries United States
Source: OECD, Health Data 2003
A closer look at growth rates by decade reveals that there are differences between
the different health system types (Figure 2.9). During the seventies, variation of
the growth rate of public share of total health expenditures is relatively small with
a growth rate of 14% for NHS systems, 14.5% for non-NHS systems and 13.3% for
the United States. During the eighties with the exception of the United States,
countries with a NHS system and countries without a NHS system saw their
growth rate of public share reduced by almost half to respectively 7.1% and 7. 5%.
The United States’ public share growth rate fell to 9.5%. This decreasing trend
continued during the nineties.
84
Figure 2.9. Mean Average Growth Rate of Public Financing of Health Care
Systems by Decade.
14.0
7.1
5.7
14.5
7.5
5.5
13.3
9.5
6.3
0
2
4
6
85
Source: OECD Health Data 2003
Of the countries that switched to a NHS system, Greece and Spain had an increase
of public share of total health expenditure after the NHS system was established
(Figure 2.10.).
Figure 2.10. Mean Average Growth Rate of Public Financing before and after the
introduction of NHS after the Introduction of a NHS System.
49
63
75
54.0
75.0
60.8
76.9
0
10
20
30
40
50
60
70
80
90
100
Greece Italy Portugal Spain
Before NHS
After NHS
Source: OECD Health Data 2003
8
10
12
14
16
1970-1980 MAGR (%) 1980-1990 MAGR (%) 1990-2000 MAGR (%)
Year
NHS Countries
Non NHS Countries
Growth
Rate
United States
Their public share increased respectively from 49% to 54% and from 75% to
76.9%. Portugal saw its growth rate of public expenditures on total health
expenditures decrease from 63% to 60.8%. Italy and Spain had a higher public
share after the introduction of a NHS system compares to Greece and Portugal.
2.2.1. Out of Pocket Share of Total Health Expenditures.
Out of pocket medical expenses as a percentage of total health expenditures was
reduced for all types of health care systems (Figure 2.11). The United States had
the highest decrease with the share of out of pocket expenses dropping from 35.3%
in 1970 to 15.2% in 2000.
Figure 2.11. Percent of Out of Pocket Expenditure (1970 – 2000).
0
10
20
30
40
50
60
70
80
90
100
1970 1975 1980 1985 1990 1995 2000
Year
Percent
NHS Countries
Non NHS Countries
United States
Source: OECD Health Data 2003
86
Countries with a NHS had a decrease of the proportional share of out of pocket
payments from 20% in 1970 to 15.9% in 2000.
The only countries that had an increase over the time span from 1970 through 2000
were the countries without a NHS. Their out of pocket share went up from 13.9%
to 14.8%.
While the share of total health expenditures on health in terms out of pocket
payments remained rather stable for most countries except for the United States,
absolute out of pocket expenses vary substantially by decade and by health care
system type (Figure 2.12).
Figure 2.12. Per Capita Out of Pocket Expenditure in 1970, 1980, 1990 and 2000
by Health System.
34.7 31
122
68.2
96
256
167.6
212
550
324.2
349
690
0
100
200
300
400
500
600
700
800
Average NHS Average Non NHS United States
1970
1980
1990
2000
Source: OECD Health Data 2003
87
2.3.Trends in Health Care System Components.
2.3.1. Inpatient Health Expenditures.
All types of health care systems saw their share of total health expenditures spent
on inpatient health care increase until the early eighties (Figure 2.13). However,
compared to the other to the two other types of health care systems, countries with
a NHS system had the highest proportion of their total health expenditures spent on
inpatient care.
Figure 2.13. Percent of Inpatient Care by Health Care System (1970 – 2000).
0
10
20
30
40
50
60
70
80
90
100
1970 1975 1980 1985 1990 1995 2000
Year
Percent
NHS Countries
Non NHS Countries
United States
Source: OECD Health Data 2003
Initially, the inpatient share of 54% in 1970 increased to 59.9% in the mid eighties
after which it dropped to 47.1% in 2000.
88
In terms of per capita expenses, expenditures for inpatient care have increased for
all types of health care systems from 1970 through 2000 (Figure 2.14.). The
United States’ per capita spending has been the highest throughout that era. The
amount of per capita expenditure on inpatient care increased from $142 in 1970 to
$1,253 in 2000. A similar trend took place in the two other types of health care
systems. Inpatient expenditures per capita were alike in systems with and without
a NHS. In 1970 approximately $ 80 per capita was spent on inpatient care. This
was almost half what the United States was spending. In 1980, this amount went
up to respectively $ 363.4 and $ 304.2. In 1990, the gap widened to $ 713.2 for
NHS countries and $ 598.6 for non NHS countries after which it decreased to
respectively $989.8 and $919.5 in 2000.
Figure 2.14. Per Capita Inpatient Care Expenditures (1970 – 2000).
0
200
400
600
800
1000
1200
1400
1970 1975 1980 1985 1990 1995 2000
Year
$ PPP
NHS Countries
Non NHS Countries
UnitedStates
Source: OECD Health Data 2003
89
2.3.2. Pharmaceutical Expenditures.
Pharmaceutical expenditures as a percent of total health expenditure show a
decreasing trend from 1970 to the mid-eighties after which they started to increase
again (Figure 2.15.). The United States’ share of pharmaceutical expenditures is
the lower than the other two types of health care systems.
Figure 2.15. Percent of Pharmaceutical Expenditures by Health Care System
(1970 – 2000).
0
2
4
6
8
10
12
14
16
1970 1975 1980 1985 1990 1995 2000
Year
NHS Countries
Non NHS Countries
United States
Percent
Source: OECD Health Data 2003
In the United States, while experiencing a steady decrease of pharmaceutical share
of total health expenditures from 12.4 % in 1970 to 8.6% in 1994, a steep increase
took place from 1995 to 2000. During that era, the pharmaceutical share increased
with 3 percent from 8.9% to 11.9% of total health expenditures.
90
Per capita spending on per capita pharmaceutical expenditures ($ PPP) increased
steadily from 1970 to 2000 for all types of health care systems (Figure 2.16).
Figure 2.16. Per Capita Pharmaceutical Expenditures (1970 – 2000).
0
100
200
300
400
500
600
1970 1975 1980 1985 1990 1995 2000
Year
$ PPP NHS Countries
Non NHS Countries
United States
Source: OECD Health Data 2003
In 1970, compared to the United States with $ 159 per capita pharmaceutical
expenditures, countries without a NHS spent $ 119.4 per capita, while NHS
countries spent the lowest per capita on pharmaceuticals with $ 18.5 per capita.
Ten years later in 1980, per capita spending on health expenditures has doubled.
This doubling trend has continued in for the eighties and the nineties (Table 2.4).
Table 2.4 also shows that there are differences among countries within each type of
health care system.
91
92
Table 2.4. Per Capita Expenditures ($ PPP) on Pharmaceuticals (10 year interval).
1970 1980 1990 2000
NHS System
Denmark - 50 109 208
Finland 20 54 122 264
Iceland 22 92 216 -
Ireland 22 49 88 191
New Zealand - 54 129 -
Norway 10 55 98 -
Sweden 18 55 120 305
United Kingdom 21 57 132 -
Average 18.8 58.3 126.8 242.0
Non NHS System
Australia - 53 117 292
Austria - - - -
Canada 29 60 192 404
Germany 36 110 228 379
Japan - 111 232 315
Netherlands - 53 128 237
Average 32.5 74.3 179.0 358.3
Entrepreneurial System
United States 43 96 251 541
Grand Average 31.4 76.2 185.6 380.4
Source: OECD 2003
With respect to average growth rate, figure 2.17 shows that the growth rate was the
highest for NHS countries in the seventies at 12% while in the eighties and nineties
pharmaceutical growth rates remained stable at about 8%. The United States
pharmaceutical growth rates were lower than NHS and non NHS countries in the
seventies but higher in eighties. In non NHS countries growth rates decreased
moderately but continuously.
Figure 2.17. Mean Average Growth Rate of Pharmaceutical Expenditures by
Decade.
0
2
4
6
8
10
12
14
1970-1980 MAGR (%) 1980-1990 MAGR (%) 1990-2000 MAGR (%)
Growth
Rate (%)
NHS Countries
Non NHS Countries
United States
Source: OECD 2003
2.4.Trends in Availability of Technological Applications.
The availability of technological applications has steadily increased in all types of
health care systems (Table 2.5). With respect to CT Scanners, non NHS countries
have the highest number of CT Scanners per million inhabitants from the nineties
on, taking over the lead from the United states who had the highest availability in
the throughout the eighties. For the last 20 years, NHS countries had the lowest
availability of CT Scanners. However, there is quiet some variation of availability
of CT Scanners within each type of health care system. Compared to Denmark,
Ireland, New Zealand and the United Kingdom, Finland, Iceland, Norway and
Sweden had twice as many CT Scanners per million inhabitants in 1990.
93
94
Table 2.5. MRIs and CT Scanners per 1 Million Population.
1980 1985 1990 1995 2000
Scan MRI Scan MRI Scan MRI Scan MRI Scan MRI
NHS System
Denmark 0.2 - 1.6 - 4.3 2.5 7.3 - 11.4 6.6
Finland 1.5 - 5.3 0.4 9.8 1.8 11.7 4.3 13.5 9.9
Iceland 0.0 - 8.3 - 11.8 3.9 18.7 7.5 21.3 10.7
Ireland - - - - 4.3 - - - - -
New Zealand - - - - 3.6 - - - 8.8 -
Norway - - - - 11.6 0.7 - - - -
Sweden 1.9 - 4.9 0.2 10.5 1.5 - 6.8 - -
United Kingdom - - 4.3 1.0 - 3.5 - -
Average NHS 0.9 - 5.0 0.3 7.5 1.9 12.6 5.5 13.75 9.1
Non NHS System
Australia 0.7 - 8.2 - 13.8 0.6 20.8 2.9 - 4.7
Austria - - - - 11.6 - - - 25.8 10.8
Canada - - - 0.2 7.1 0.7 8.0 1.3 - 2.5
Germany - - 6.6 0.5 10.1 1.9 15.6 4.8 - -
Japan - - - - 55.2 6.1 - - - -
Netherlands - - - - 7.3 0.9 3.9 - -
Average Non
NHS 0.7 7.4 0.4 17.5 2.0 14.8 3.2 25.8 6.0
Entrepreneurial
System
United States - - 12.7 1.1 14.6 3.7 14.1 13.1 8.1
Switch Countries
Greece 0.6 - 1.5 - 6.5 0.4 - - - -
Italy - - - - 6.0 1.7 - 3.4 20.6 -
Portugal - - - - - 0.8 8.3 - 12.1 -
Spain - - - - - - - 2.7 - 4.9
Average Switch 0.6 1.5 6.3 1.0 8.3 3.1 16.35 4.9
Grand Average 0.73 6.7 0.4 11.5 2.2 12.4 4.8 17.3 7.0
Source: OECD 2003
Within the group of countries not having a NHS health care system, Japan has
about three times as many CT scanners per million inhabitants than that group’s
average. With respect to CT Scanners the United States remained stable with
respect to the availability of CT Scanners.
Similar trends took place for NHS and non NHS health systems with respect to
MRIs. However, as opposed to CT Scanners, the United States also experienced
an increase in availability of MRIs.
Rather than looking at the availability of technologically advanced diagnostic
devices, Figure 2.18 shows the actual use of high tech procedures in terms of
kidney transplants. For all types of health care systems, a steady increase took
place from 1970 through 1985. However, only in the United States this trend
continued throughout the eighties and nineties while the number of kidney
transplants remained almost stable for the two other types of health care systems.
Figure 2.18. Kidney Transplants per 100,000 population.
95
0
1
2
3
4
5
6
1970 1975 1980 1985 1990 1995 2000
Year
NHS Countries
Non NHS Countries
United States
Source: OECD 2003
96
2.5.Common Trends or Not?
2.5.1. Total Health Expenditures: Long Run, Short Run and Government.
Total Health Expenditures have increased over the long term in absolute (PPP) as
well as relative terms (% GDP) for all types of health care systems but to a larger
extend for the entrepreneurial system of the United States. However, there is not
that much difference in the 30-year mean average growth rate of absolute total
expenditures among all systems.
On the short run there are however differences between the systems. Mean
average growth rates of per capita health expenditures decreased every decade for
all health care systems. For NHS and non NHS countries, the largest decrease took
place in the eighties while for the use the major decrease took place in the nineties.
With respect to relative expenditures, all systems show identical patterns when
GDP growth is compared health expenditure growth by decade. In addition,
percentages by decade are almost identical for the three systems.
The introduction of a NHS system seems to be related with increased per capita
health expenditures and share of GDP spent on health. After the introduction of a
NHS system, per capita health expenditures increased in Greece, Italy, Portugal
97
and Spain. However, growth rates were reduced after the introduction of a NHS
system.
2.5.2. Public Share of Total Health Expenditures: All the Same in the End?
Although there are substantial differences between the systems, the public share of
total health expenditures remained rather stable over the long run. The United
States’ public share of total health expenditures is much lower than in any other
country. However, in the United States the mean average growth rate of public
financing of health care is higher than in the other systems in the long run as well
as in the short run. On the short run, all systems reduced the growth rate of their
public share by almost one third. However, the United States lagged behind in this
trend during the seventies. With respect to the introduction of a NHS system
trends are not so clear. If there is an increase, this increase in public share of total
health is limited.
2.5.3. Out of Pocket Payments and Inpatient Share: Relative Contradict
Absolute expenditures.
While the share of out of pocket costs remained relative constant for NHS and non-
NHS countries, the United States experienced a drastic decline over a 30-year time
98
span. This trend is not confirmed by absolute out of pocket expenditures. On the
contrary, per capita out of pocket expenditures have increased dramatically over
that same period.
With respect to the inpatient share of total health expenditures, a similar trend can
be observed. While relative inpatient health expenditures went down from 1970
through 2000, per capita inpatient health expenditures skyrocketed over that same
era.
2.5.4. Pharmaceutical Expenditures and Technology: Are they to blame?
Trends in pharmaceutical expenditures are somewhat contradictory but not that
different from trends in other health care system characteristics in terms of absolute
and relative pharmaceutical spending. Pharmaceutical expenditures increased in
terms of per capita dollars for all health care systems, while to a certain extent the
opposite is true for percent of total health expenditures spent on health care. As a
share of total expenditures, they show a decreasing trend until the mid-eighties.
With respect to technology, NHS and non NHS countries have increased
availability of high-tech diagnostic devices such as MRIs and CT Scans. While in
the United States the availability of MRIs has increased over the last thirty years,
99
the availability of CT Scans has remained rather constant. With respect to the
utilization of technology as proxied by the number of kidney transplants, the
United States experienced a higher increase than the other countries.
2.6. Puzzling trends and more questions than answers.
Trends in health care systems described in this section suggest several issues.
First, what is the actual impact of public financing on total health expenditures
given the major differences between the United States’ health care system and the
other health care systems. Second, are there system effect differences since NHS
countries have lower total health expenditures? Third, what is the impact of
demand and supply policies and the structure of the expenditures. The following
chapter will elaborate on these issues through a multivariate analysis.
100
Chapter 3 : Multivariate Analysis of Total Health
Expenditures and its Determinants
3.1. Literature Review.
3.1.1. Health Expenditures and a Nation’s Income.
Simple regression analysis - cross sectional as well as pooled time series - has
shown that GDP per capita explains a substantial percent of variation of total per
capita health expenditures independent of whether exchange rates or purchasing
power parities (PPPs) are used. (Newhouse, 1977; OECD, 1987; Parkin, 1987;
Parkin, 1989; Hitiris, 1992; O’Connel, 1996).
Similar results with respect to GDP have been obtained in multivariate cross
section and time series analysis. Kleiman’s (1974) seminal cross sectional
multivariate analysis established that GDP in terms of exchange rates was
positively related to total health expenditures in all models tested. This was
confirmed by later cross sectional analysis using exchange rates (OECD, 1985;
Leu, 1986, Getzen, 1992; Gerdtham & Jönsson, 1991) and PPPs (OECD, 1985;
Pfaff, 1990; Gerdtham & Jönsson, 1991; Gerdtham et. al., 1992, Getzen 1992;
Murillo, Piatecki & Saez, 1993).
101
Multivariate time series analysis has yielded less stable results. Pfaff (1990) using
data for three countries confirmed the positive relationship between GDP and total
per capita health expenditures for Canada and the United States but not for
Germany where the relationship became negative but not significant. Kanavos and
Yfantopoulos (1999) found that GDP had no significant explanatory power in 7 of
the 14 countries of the European Union independent of measurement in terms of
exchange rates, PPPs or national currencies.
For more than a decade, pooled cross section time series multivariate analysis is
the predominant way to analyze the determinants of health expenditures.
Gerdtham (1992) using five pooling methods on a static equilibrium model as well
as a dynamic error correction model for 22 countries covering 16 years (1972 –
1987) found that for both types of models GDP per capita PPP and per capita
health expenditures were positively related. In a follow up studies for 22 countries
from 1971 through 1991(Gerdtham et al. 1994; 1998, Gerdtham, 2003) the effects
of GDP on total health expenditures were confirmed. Roberts (1999) using data of
a selection of 20 OECD countries from 1966 to 1993 also obtained a significant
positive estimate for per capita GDP.
In other words, studies investigating the determinants of total health expenditures
have found a significant and positive relationship between per capita GDP and total
per capita health expenditures, independent of the measurement of GDP, the
102
functional form of the empirical model, the type of data (cross-sectional or pooled)
and whether using simple or multivariate analysis.
3.1.2. Public Share of Total Health Expenditures.
Public expenditures as a share of total health expenditures have been introduced in
several multivariate analyses. In the OECD (1985) cross sectional models
covering 20 countries in three separate years (1970, 1976 and 1980) public
financing is never significant but is consistently inversely related to total per capita
health expenditures regardless of functional form or whether expenditures are
measured in terms of exchange rates or percent of GDP spent on health.
In the same OECD study, public financing becomes significant and has a negative
sign when the three samples are pooled. Similar results the OECD study were
obtained by Gerdtham et al. (1992) with 1987 data for 19 countries and. When
using pooled data for 22 countries from 1972 through 1992 (Gerdtham, 1992), the
sign was always negative but significance depended on the statistical method used.
O’Connell (1996) also found a significant and inverse relationship between public
share of total per capita health expenditures for a different sample of 21 countries
from 1975 through 1990. In Roberts’ (1999) analysis using a sample of 20
countries from 1960 through 1993, significance of the public share estimates
103
always positive but significance depended upon the statistical method used. In the
same study, a cross section regression using long-term averages for the same
counties used in the pooled time-series, yielded an inverse and insignificant
relationship between public expenditures and total per capita health expenditures.
Finally, in Gerdtham et al. (1993) negative but insignificant estimates were
obtained for public share in a sample of 24 countries from 1970 to 1991. Later, a
follow up study by Gerdtham et al. (1998) using four different samples only
resulted in negative and significant public share estimates for sample containing
data on 17 countries from 1970 to 1991. For the same sample from 1981 to 1991
and a sample of 22 countries from 1981 to 1991 estimates were positive but not
significant.
In other words, unlike per capita GDP, the public share of total health expenditures
has produced less consistent results. Both direction and significance depend on the
sample. Multivariate analysis using either cross sectional or pooled data,
sometimes yield positive and sometimes negative estimates indicating that in the
seventies public share had a expenditure lowering effect but not thereafter. In
addition, different statistical methods also produce different estimates in terms of
sign and significance.
104
3.1.3. Other Determinants of Total Health Expenditures.
3.1.3.1. Health System Effect.
Leu (1984) found that countries with a public integrated health care system
(National Health Service) have significantly lower health expenditures than those
who had not. Gerdtham et al. (1992) could not replicate these results. When
included in the model with public share of total health expenditures the parameter
estimate for public integrated health care system became insignificant and, when
entered in the model with public hospitals and public financing share, the sign of
public integrated health care system switches directions. While significantly
affecting total health expenditures inversely in the follow up study (Gerdtham et al.
1994), this result became insignificant in later studies (Gerdtham et al. 1998,
2003).
3.1.3.2. Public Provision.
Inpatient care as a share of total health expenditures has been entered in the
Gerdtham et al. study (1992) as a proxy for supplier induced demand substituting
Leu’s (1984) ratio of public and non-profit hospital beds to total hospital beds. In
105
all three models tested in the Gerdtham study, the share of inpatient care was
positive and significantly related to total health expenditures.
Later studies confirmed the cost increasing impact of share of inpatient care on
total health expenditures. In addition, an inverse relationship was found between
inpatient and ambulatory care suggesting substitution between the two types of
medical care (Gerdtham et. al. 1994, 1998, 2003; Gerdtham & Jönsson, 2000).
3.1.3.3. Open Ended Financing of the Health Care System.
Gerdtham et al. (1992) also included two measures of open ended-ness of the
financing: fee-for-service and global budgets. As opposed to Leu (1984) who from
a public choice perspective argues that public financing will increase health
expenditures, Cuyler (1989) argued that the financing mechanism is more
important than the public or private nature of the provider. His argument is based
on the lack of incentives to control health expenditures.
Gerdtham et. al. (1992) operationalize this argument through the means of two
dummy variables: fee for service and global budgets. The first applies to
physicians and has an increasing influence on health expenditures when used as
predominant overall physician payment method. (Gerdtham, 1992) However,
106
when using the dummy as predominant inpatient remuneration the variable
becomes insignificant and even negative for a subset of European Union member
states. (Gerdtham, et al. 1994, 1998, 2003).
Restricted or closed-ended financial resources for hospitals are translated into a
global budget dummy variable. In the initial study (Gerdtham et al. 1992), global
budgets for inpatient care did not have a significant impact on total health
expenditures. More recent studies that make a distinction between budget ceilings
for ambulatory and inpatient care yielded similar results for both. (Gerdtham et al.
1994, 1998, 2003).
3.1.3.4. Technology.
Technology has not been systematically introduced in the comparative health
expenditure literature as a determinant of total health expenditures. Kanavos and
Yfantopoulos (1999) include growth rate of pharmaceutical expenditures as a
proxy variable for technology in a country-by-country time series analysis. This
proxy variable turns out to be a significant driver of total health expenditures in all
countries of the European Union except for Italy and the United Kingdom.
107
3.1.3.5. Control Variables: Ageing and Urbanization.
Ageing.
International comparisons have used different variables to measure the influence of
an ageing population on total health expenditures and no matter what variable used
the estimates of the effects of an ageing population is mixed.
The most common indicator is share of population age 65 and older. One cross-
sectional analysis of the OECD (1985) found that the percent population age 65
and older had no significant impact on total per capita health expenditures and that
that there was an inverse relationship except for the years 1970 and 1976 where the
sign was positive. When using change in per capita health expenditure over
periods of 5 years, age was inversely related with the dependent variable. Getzen
(1992) found a negative but insignificant relationship between ageing and total
health expenditures.
Using pooled time series, positive and significant estimates for the percentage of
population 65 and older were only obtained by Hitiris & Posnett (1992) and
O’Connell (1996) while in the OECD study (1985) the only significant parameter
estimates were negative. Similar to the OECD results, Barros (1998) for four 10-
108
year time periods and Roberts (1999) for a 33 year time span never obtained
positive or significant estimates. Finally, Okunade (2004) attained no significant
relationship between health care expenditure growth in $ PPP and ageing on the
short run but did a find positive and significant relationship on the long run.
A second way to measure the demographic structure of a society is age dependency
ratio, which is a ratio of not-working population (ratio of the population under age
18 and over 65) over the working population (population between 18 and 65).
Gerdtham’s analysis (Gerdtham, 1992) shows that age dependency is significant
and has a positive sign when entered in the model.
A third measure capturing age related demographic changes is the percent of
population age 75 or older together with the percent of population 4 years old and
younger. Both variables have been introduced in the model together since over a
life time health care expenditures take U-shaped shape being higher both in early
life and before death. Gerdtham et al. (1994, 1998) results with respect to effect of
ageing on total health expenditures was that most of the times the variables were
inversely related and always negative. With respect to the percent of population
under the age of 4, the estimates are positive and significant.
109
Urbanization
Urbanization has been introduced in the earlier comparative health expenditures
studies. Leu’s analysis (1984) generated positive estimates for the impact of
urbanization on total health expenditures but urbanization was not always
significant. In Gerdtham’s (1992) replication of Leu’s model, urbanization had a
negative sign and was never significant. When introduced in his own model,
urbanization was significant and inversely related to total health expenditures. A
similar result was obtained by Karatzas (2000) for the United States where
urbanization had a negative sign but was not significant.
3.2. The Model.
In this section, an alternative framework is offered to analyze the differential
impacts of independent variables on total health expenditures. Based on this
framework, an empirical model is specified and, the relationships between
dependent and independent variables are a priori determined based on theory and
previous research.
3.2.1. Conceptual Framework.
Previous research has mainly focused on different determinants of total health
expenditures such as income, financing source, openness of financing and
demographic determinants without offering a larger conceptual framework.
The framework proposed here is in accordance with definition from chapter 2 and
encompasses several categories that make up a system in terms of the delivery and
financing of health care services. (Figure 3.1.)
Figure 3.1. Conceptual Framework: Determinants of Total Health Expenditures.
GDP
PUBLIC
SHARE
Supply Side
• Physician Payment
• Hospital Payment
Demand Side
• Out of Pocket.
Health System Effect
Control variables
• Population Age
• Urbanization
Other Explanatory Variables
Primary
Explanatory
Variable
Key Explanatory
Variable
of Interest
Health Expenditure Structure
• Inpatient Care
• Pharmaceutics
• Technology
TOTAL
EXPEN
Dependent
Variable
110
111
Besides the primary explanatory variable (income in terms of per capita GDP) and
the variable of interest (public share of total health expenditures), several other
determinants are introduced in the empirical model.
3.2.2.The Empirical Model.
Based on the literature and the above presented conceptual framework, the
following empirical model can be proposed:
THEXP = ƒ GDP + PUBSHARE + NHS + INPATIENT + PHARMA + KIDNEY
+ FFS + GLOBAL + DRG + POCKET + POP65 + URBAN + µ.
Dependent variable:
THEXP = Total per capita health expenditures in $ Purchasing Power Parities.
Independent variables:
GDP = Per capita Gross Domestic Product in $ Purchasing Power Parities.
PUBSHARE = Public Share of total health expenditures (%).
INPATIENT = Share of total health expenditures spent on inpatient care (%).
PHARMA = Share of total health expenditures spent on pharmaceuticals (%).
KIDNEY = Number of kidney transplants per 100,000 population.
112
FFS = Dummy variable with unity for countries with fee for service as
predominant physician payment.
GLOBAL = Dummy variable with unity for countries with budget caps for
hospitals starting at the year of implementation if applicable.
DRG = Dummy variable with unity for countries with Diagnostic Related Groups
starting at the year of implementation.
POCKET = Share of total health expenditures paid out of pocket (%).
POP65 = Share of total population age 65 and older (%).
URBAN = percent of population living in an urban area as defined by the country
itself.
µ = error term.
3.2.3.The Dependent Variable: Total Health Expenditures.
The research question addresses the determinants of health expenditures in
different OECD countries. Therefore, total health expenditure is the dependent
variable and it is measured in per capita dollars in purchasing power parities.
113
3.2.4.The Independent Variables.
Income.
At different levels of analysis, (household level, health plan level and national
level) a consistent positive relation has been found between income and health
expenditures (Getzen, 2000).
The only difference between studies with different levels of analysis is the
magnitude of the regression estimate. The smaller the unit of analysis, the lower
the estimate indicating that health care is a necessity and not a luxury good.
Moreover, in aggregate comparative studies elasticities of health care differ
depending on the method used to calculate the estimates. Regardless of whether at
the aggregate level of analysis the income elasticity is larger than unity, a positive
relationship is expected between total health expenditures and income measured in
terms of per capita GDP.
Public share of total health care services.
Leu’s theory (1984) is based on public choice predicting a positive relationship
between public share and per capita total health expenditures. His argument is
114
twofold. First, at the national level total health expenditures are largely determined
through a budget process in which decision makers act as budget maximizers.
Second, due to a lack of competition production of health services does not give an
incentive to minimize costs. However, this argument was countered by Cuyler
(1989) arguing that private organizations act in a similar way.
From a welfare economic point of view, the theory of market failure may offer a
sound basis upon which there is a rational to argue in favor of public financing of
health care (Cuyler, A.J., 1971; Pauly, M., 1986; Musgrove & Jönsson, 1993.;
Øvretveit, J., 1996; Musgrove, P., 1996; Rice, T., 1998; Musgrove, P., 1999)
Based on Arrow’s (1963) seminal work that lead to the conclusion that in the
absence of working market for health insurance government can provide health
insurance Musgrove (1996) looks at health care from a two dimensional view,
offering a theoretical framework that allows distinguishing between different types
of health services along costs and nature of the intervention (Figure 3.2.).
While most health interventions are of a private nature, markets can fail for either
economic or equity reasons. Failure because of economic reasons occurs in two
cases resulting in inefficient market outcomes.
Figure 3.2. Musgrove’s Two Dimensions of Health Care.
115
Source: Modified from Musgrove 1996
Risk Sharing
High
Low
Cost
Intervention
Public Private
Market Externalities
Public Good
First, market failure will be present when health care becomes a (nearly) public
good. This is the case when consumption of the good by one individual does not
reduce the availability of that good to other individuals. Consequently, individuals
can consume that good without individually paying an appropriate price for that
good. In addition, the production of these goods is very costly and/or impossible to
produce. Examples of these goods are water and food safety. For these types of
health interventions, there is an appropriate role for government to finance these
goods.
Second, there are health interventions, although not costly, that have substantial
externalities. Here too, markets will result in a less than optimal production of that
good since those individuals paying for that good cannot prevent others to benefit
116
from their purchase. An example of externalities is the control of communicable
disease. Curing one individual from a communicable disease prevents others from
getting that disease without paying a price. If these goods are financed through a
market mechanism, not enough of this good will be provided due to free riding.
Therefore, public financing but not necessarily public provision is desired.
Most medical interventions are routine and can be easily paid for out of pocket.
However, illness strikes randomly and can be catastrophic in the sense that out of
pocket payments can drive an individual into poverty. In the case of catastrophic
costs, collective financing through the means of risk sharing via contributions is a
possible solution to overcome this problem. Therefore, government may finance
insurance for equity reasons since in a market environment individuals with a high
risk for catastrophic illness are not likely to be insured.
While Musgrove’s argument provides a theoretical foundation for the public
financing of health care, empirical results are ambiguous. From the literature
review a common finding is that in the earlier years public financing has a
expenditure decreasing effect while this effect washes out in later years. The goal
here is to replicate these results from earlier studies.
Health System Effect.
The health system effect is measured through a dummy variable for whether or not
a country has a National Health Service that is characterized by predominant public
financing as well as public provision. As such, this variable also measures the
predominant nature of the provision of health services in addition to the source of
financing measured by public share of total health expenditure.
As far as this variable actually captures the public provision, the public choice
argument in the tradition of Buchanan (1969) and Niskanen (1971, 1983) as Leu
suggests may hold strong. The major reason why government can fail is the
absence of a link between the costs of producing and the revenues generated for the
production of the good. The inability of linking both will encourage inefficiencies
and may result in misallocation, overuse of sources or overproduction. A second
reason why government is likely to fail is embedded in the organizational nature of
a nonmarket entity. It refers to internal management problems that have to be
overcome on a daily basis. The last economical reason why government can fail is
the creation of derived externalities.
117
These unwanted effects come into being through policy measures. (Stevens, 1993;
Wolf, 1993; Horn, 1995, Miller, 1995) Therefore, we expect countries with an
NHS to be more expensive than countries without an NHS. The NHS dummy
118
variable was coded as unity for country and year in which a NHS system was in
place similar to Gerdtham et al. (1994, 1998, 2003).
Health Expenditure Components.
The way health expenditures are structured can also influence the total health
expenditures. Three variables are introduced that represent the different
components of health expenditures: inpatient expenditures, pharmaceutical
expenditures and technology.
Inpatient care is more expensive than ambulatory care. Over the years, wherever
possible shifts have taken place from care provided in a hospital setting to care in
outpatient health facilities (Feldstein, 1993). If this substitution is an effective
cost-containment measure, then the lower the share of inpatient care spent should
have an inverse relationship with total health expenditures.
Pharmaceutical expenditures are becoming an increasingly important part of total
health expenditures due to increased prices of pharmaceutical products and
increased volume of consumption, together with growth in insurance coverage of
prescription drugs and increased direct-to-consumer advertisement. (Du Bois et al.,
2000; Berndt, 2001; Reinhardt, 2001; Danzon & Pauly, 2002).
119
On the other hand, an alternative view suggests that in the end many
pharmaceuticals are more cost-effective than other medical interventions such as
surgery. In other words, drugs and especially newer more costly drugs cause
pharmacoeconomic benefits that offset other medical costs. (Goldberg, 1993;
Kleinke, J.D., 2000; Neumann et al. 2000; Soumerai, S.B., 2004).
Over the years, empirical research predominantly focusing on elderly and welfare
recipients has found evidence for this substitution theory (Poisal et al., 1999;
Thomas et al., 2001). Studies in the United States show that limiting reimbursable
medication has increased institutionalization among elderly covered by Medicaid
(Soumerai et al., 1991). Similarly, HMO enrollees who were subject to
pharmaceutical capitation had higher total health costs than those not subjected to
capitation (Popovian et al., 1999). In addition, pharmaceutical cost increases for the
treatment of schizophrenia was lower than cost increases of other medical services
for treating the disease. Finally, Lichtenberg (1996) found that hospitalization and
surgical procedures was inversely related to number of prescribed drugs and that
newer drugs were able to reduce-non pharmaceutical expenditures (Lichtenberg,
2000).
In Canada, Tamblyn et al.(2001) found that larger out of pocket payments for
prescription drugs was related to higher emergency room visits, while Crémieux et
120
al. (2005) found a inverse relationship between provincial pharmaceutical
expenditures and health outcomes.
Until now, comparative multivariate analysis of health expenditures has not
integrated a variable representing the substitution theory in its models. Therefore,
pharmaceutical expenditures as a percent of total health expenditures is entered in
the model to test the substitution theory on an aggregate level and, pharmaceutical
share is expected to be inversely related to total health expenditures.
Finally, technology is also considered an important driving force of increasing
health expenditures (Chernew, M.E. et al., 1998; Fuchs, 1996; Fuchs, 1999; Gelijns
& Koenig, L. et al., 2003; Newhouse, 1992; Rosenberg, 1994; Weisbrod, 1991)
A two-step argument underpins this theory. The more technological innovation is
available, the more it is used and the higher the cost of health becomes. (Baker, L.
et al., 2003; Baker, L. et al. 2003) The first part of this reasoning is supported by
area variation studies on the utilization of health services indicating that there was
a relationship between physicians’ medical practice and the number of medical
procedures performed. (Wennberg & Gittelsohn, 1972, McPherson et al., 1982);
Wright et al., 1999; O’Connor et al.,1999; Fisher et al., 2003) The second part of
the argument is that because of insurance and demand inducement, utilization
increases. This leads on its turn to higher expenditures. (Danzon & Pauly, 2001)
121
In international comparative multivariate analysis, technology is rarely entered in
the models. There are different ways to measure technology. In a study analyzing
the effect of income on technology diffusion in OECD countries, Slade and
Anderson (2001) use five variables to measure technology: MRI, CT Scanners,
liver and kidney transplants and hemodialysis patients. In this analysis, technology
is operationalized in the model as kidney transplants per 100,000 population. There
are two reasons for this. One concerns the availability of data. The number of
kidney transplants is the variable with least missing values. The other is that by
using this variable the whole process of how technology affects costs through
availability and utilization is captured. Therefore, it is expected increased
technology will increase expenditures.
Supply Side Determinants.
One way of controlling health expenditures is related to the openness of financing
mechanisms. Three variables are entered in the model to verify whether the
openness of financing has an influence on total per capita health expenditures. One
variable applies to physicians, the other two are hospital related.
With respect to physician, the openness of payment method refers to the underlying
issue of physician-induced demand. This phenomenon has been measured in two
122
ways. On the one hand, physician to population have been used and, alternatively
physician payment method has been used.
When using physician ratios in international comparisons, contradictory results
have been obtained. Depending on the model tested, the OECD studies (1985;
1987) sometimes indicated a positive and other times a negative relationship with
total health expenditures while others consistently found an inverse relationship
(Gerdtham, 1992; Gerdtham et al. 1994; 1998; 2000). This is consistent with
micro studies using physician ratios having mixed results. Some studies found
support for demand inducement (Evans, Parish & Sully, 1973; Fuchs, 1978;
Hemenway & Fallon, 1985; Cromwell & Mitchell, 1986; Tussing & Wojtowycz,
1986) while others have found no supportive results for this theory. (Wilensky &
Rossiter, 1983; McCarthy, 1985; Stano, 1985; Escare, 1992)
Another way of measuring the demand inducement is the use of physician
remuneration. Three different payment methods can be used: salary, capitation and
fee for service. The first two are most of the time used simultaneously and impose
a limit on the resources available to provide health services to patients. Fee for
service on the other hand does not limit the available resources per patient and may
invoke supplier-induced demand. Unlike physician to population ratios, the
introduction of a fee for service dummy variable has a consistent positive sign and
123
is significantly related to total health expenditure, indication it to be a more robust
measurement for demand inducement.
If the supplier induced demand theory is a valid explanation for higher health
expenditures, health systems that predominantly relay a fee for service payment for
physicians will have higher health expenditures. The FFS dummy is coded unity
for country and year in which it is the predominant payment method as is the case
in Gerdtham et al. (1994, 1998, 2000) studies.
With respect to hospitals, health expenditures can be controlled by the means of
global budget and/or Diagnostic Related Groups (DRG). Both differ from
previously used per diem hospital payment by giving a disincentive to increase
hospital stays and the overuse of medical procedures.
Global budgets most of the times are set prospectively following negotiations and,
are based on historical costs of the hospital as an entity. It can contain
expenditures in two ways. At the aggregate level, policy makers restrict hospitals
costs by setting a cap on their resources. At the micro level, hospitals are expected
to act within the budget limits increasing their efficiency by rationalizing services
and reducing costs (Rice, 1992; Wolfe & Moran, 1993; Monrad Aas, 1995; Carrin
& Hanvoravongchai, 2003).
124
Several studies have assessed global budgets for hospitals as a successful policy
tool to control health expenditures. (U.S. GAO, 1991; Abel-Smith, 1992; Wolfe &
Morane, 1993) However, other studies found that global budgets did not contain
costs and that if budgets were increased more medical procedures would be
performed than needed. (Poterba, 1994; Feldman & Lobo, 1997)
Rather than imposing a cost ceiling at the organizational level, DRGs fix hospital
payments prospectively based on a patients’ diagnosis and the average cost of
treatment. The goal is a more efficient use of available resources. By placing the
financial risk on the hospital, it is expected that lengths of stay and unnecessary
medical procedures be reduced (Lave, 1984). If these underlying assumptions are
valid, then it is expected that countries using a DRG based payment method for
hospitals will have lower health expenditures.
However, previous research has doubted the effectiveness of the DRGs. In a study
of the OECD countries, Forgione and D’Annunzio (1999) found that countries
adopting DRGs had higher total expenditures as well as higher inpatient
expenditures after the DRGs were implemented. Studies on the impact of DRGs in
Finland, Sweden and Norway confirm this finding (Hakansson, 1999; Mikkola et
al., 2001) while studies in the United States were less conclusive (Davis & Rhodes,
1988; Gilman, B.H. 2000; Menke et al., 1998).
125
This variable has never been entered in a multivariate comparative analysis and is
coded unity for country and year if a DRG method is used to cover hospital costs.
The coding is based on two articles discussing the implementation of DRGs in
OECD countries (Lüngen & Lauterbach, 2000; Forgione, Vermeer, Suryskar,
Wrieden & Plante, 2004). Given the previous research results, DRGs are not
expected to have a significant influence on total health expenditures.
Demand Side Determinants.
Rather than aiming at the supply side, policy tools can also be directed towards the
demand side. This is the case with out of pocket expenses. Due to the presence of
health insurance, people become less price sensitive and due to moral hazard use
more health services than they actually need (Pauly, 1968). By increasing the out
of pocket costs, the consumer will become more aware of the costs of medical care
and consume less which will result in lower health expenditures (Ellis & McGuire,
1993).
Important empirical evidence of the demand theory of health care comes from the
RAND health insurance experiment. The study showed that individuals with
higher out of pocket costs had on average lower annual health expenditures
(Manning et al., 1984; Newhouse, J. et al., 1993). It is expected that increased out
of pocket cost are inversely related with total health expenditures.
126
Control Variables: Ageing and Urbanization.
Intuitively one may think that aging causes health expenditures to increase in two
ways. First, the proportion of population of individuals age 65 and older grows.
Second, life expectancy is increasing (Fuchs, 1990). In other words, more people
are age 65 or older and this group of society is living longer causing health
expenditures to increase because of increased demand for health services (Cassel,
C.K. et al., 1992; Jacobzone, S., 2000; Anderson et al. 2000; Lubitz, J. et al., 2001;
Alemayehu, B. et al., 2004). This demographic change combined with the uneven
spread of health care cost in a life cycle is expected to have an expenditure lifting
effect.
Urbanization has been entered in the models on a frequent basis. There are several
reasons why the level of urbanization may have an increasing effect of health
expenditure. From a public health perspective, contagious diseases may spread
faster in urban arias because of the high density of cities. This on its turn can
increase the use of medical services. Second, compared to rural areas, urban areas
have more crime and traffic related injuries leading to a higher use of high cost
trauma centers. Finally, availability and access is likely to be higher in urban
areas. Therefore, the level of urbanization is expected to increase health
expenditures (Crémieux, P. et al., 1999).
127
3.3.Data and Methods.
3.3.1. Data.
The data used for the analysis come from the OECD health data set 2003 for all
variables except urbanization, which was retrieved from the Worldbank’s World
Development Indicators data set 2003. The OECD data set contains information on
30 countries over a time span of more than forty years. Besides demographic and
economic references, the OECD data set is mainly made up of data on health
status, health expenditure and finance, and the use and costs of medical services.
The major advantage of pooled time series is that it extends the number of
observations by combining cross sectional data on spatial units and time units. The
result is a larger number of observations (Hicks, 1996). Since the data set is
relatively small, pooling does not limit the study to restricted statistical analysis, as
would be the case in a regular cross sectional or time series regression analysis.
Pooling increases the number of observations, and thus the degrees of freedom.
This will enable us to introduce more variables and refine our model so that cross-
sectional and longitudinally varying causal forces can be analyzed (Hicks, 1996,
Sarys).
128
For this study 19 countries were withheld for the analysis: Australia, Austria,
Canada, Denmark, Finland, Germany, Greece, Iceland, Ireland, Italy, Japan,
Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom
and the United States. Some countries were not included either because their
comparability with the included countries: the Czech Republic, Hungary, Korea,
Luxembourg, Mexico, Poland, the Slovak Republic and Turkey or, because of lack
of data for the variables on which the multivariate analysis is based: Belgium,
France, Switzerland.
Data have been transformed into their logs. Previous research (Parkin et.al., 1987,
Gerdtham, 1992) has indicated that a double log functional form was preferable
above any other functional form. More recently, this finding was confirmed by
Gerdtham et al. (1994, 1998).
3.3.2. Methods.
3.3.2.1. Regression strategy.
Three different models are tested. The first model is the unrestricted model and
contains all explanatory variables from the conceptual framework with the
129
exception of the NHS and FFS dummy variables. It should be noted that in this
model as a result of controlling for as may phenomenon as possible, not all selected
countries are included because large missing values especially for out of pocket
payments and share of inpatient. Therefore, two models with fewer variables are
included.
In order to test the robustness of the models, each model is tested on two different
samples. The first sample contains data for the 19 countries from 1970 through
2000. The other sample is a subset of this sample and covers data for the same 19
countries between 1980 and 2000. In addition, each model is ran for these two
samples with and without the United States since in many ways the United States
can be considered an outlier and may have an influence on the estimates.
3.3.2.2. Statistical methods.
While pooling has advantages, it also has disadvantages when using multivariate
regression analysis. The combination of cross-sectional and time series data easily
violates the OLS assumptions. Besides possible model misspecification, pooled
data tend to be suffering from autocorrelation and heteroscedasticity (Stimson,
1985; Beck et. al., 1993; Beck & Katz, 1995; Hicks, 1996; O’Connel, 1996). In
130
other words, OLS can only be used when there is no relationship between time
points within a cross-section or between cross-sections (Sayrs, 1989).
In order to determine whether the assumptions of the absence of heteroscedasticity
and serial correlation are violated and to determine what method to use, the
different models are subjected to several tests. While both problems do not affect
the unbiasedness of the estimated coefficients, they may increase the variance of
the distribution of these coefficients and underestimate their standard errors.
In order to test whether or not the variance of the error term is constant for all
observations the χ
2
based White (1980) test is used. This test has the advantage of
being is a general test (Gujarari, 1995) in that it does not assume a specific type of
heteroscedasticity (Studenmund, 1997). With respect to the possibility that in time
series the current observation of the error term is a function of the previous one, the
first order autocorrelation Breush-Godfrey test is used (Breush, 1978; Godfrey,
1978). The Durban Watson t test is not used because of its risk to be inconclusive
(Kennedy, 1998).
In addition, two other tests are performed on all models. On the one hand a Jarque-
Bera (1 980) test is done to check for violations of the normally distributed errors.
As such, this test checks the possibility of outliers or influential observations. On
131
the other hand, every model is tested on the presence of multi-collinearity by the
means of variance inflation and tolerance levels since multicollinearity while not
biasing the estimates, can make the estimates unreliable because OLS cannot
separate the differential effects of the separate but correlated independent variables.
The results are summarized in table 3.1.
Table 3.1. Statistical Test Results for all Models.
Without
US
Without
US
Model with
US
Model with
US
Unrestricted Model
1970-
2000
1980 -
2000 1970 - 2000 1980 - 2000
White Test 97.42
a
75.05 105.78
a
73.09
a
Jarque- Bera Test 0.54 4.7
c
1.46 5.3
c
Godfrey's Test 96.07
a
57.49
a
138.23
a
74.81
a
Hausman Test 6.23
a
18.57
a
6.81
a
22.6
a
Restricted 1 Model
White Test 149.54
a
139.05
a
146.73
a
140.98
a
Jarque- Bera Test 4.4 5.57
c
6.06
b
7.06
b
Godfrey's Test 313.49
a
243.15
a
338.89
a
259.04
a
Hausman Test 22.31
a
41.68
a
19.73
a
36.76
a
Restricted 2 Model
White Test 136.11
a
119.34
a
147.32
a
113.43
a
Jarque- Bera Test 0.67 0.5 1.02 3.37
Godfrey's Test 414.5
a
286.55
a
447.09
a
303.21
a
Hausman Test 27.04
a
24.13
a
24.75
a
25.36
a
(
a
≤ 0.01,
b
≤ 0.05,
c
≤ 0.1)
With respect to heteroscedasticity - with the exception of the unrestricted model of
the 1980 -2000 sample -the White test indicates that all other models suffer from
heteroscedasticity irrespective of sample (Table 3.1.). The Breush-Godfrey test
also points toward the presence of serial correlation in all models irrespective of
sample.
132
The Jarque-Bera test does show that some of the samples’ error term are not
normally distributed. This is the case for the 1980 – 2000 samples of the
unrestricted model that include the United States as well as for the first restricted
model that omitted the United States.
Tolerance and Variance Inflation Factor (VIF) are used to detect multicollinearity
in the model. However, there is no agreed upon rule on what the exact value
confirms the presence of multicollinearity (Newton & Rudestam, 1999). The most
used criteria are that tolerance should not be lower than 0.1 and VIF should not
exceed 10 (Studenmund, 1997; Mermier et al., 2000; Hsu et al. 2000).
Multicollinearity has severe implications. The estimates may not only be wrong in
terms of magnitude, they also may have the wrong sign. In addition, the error will
be inflated rendering an insignificant estimate while the variable is actually
significant.
Appendix A shows that none of the models except for the full model exceed 10
(VIF) or no tolerance score is lower than 0.1 except for the full model where FFS
and NHS have an unacceptable score on both variables. The two variables that are
causing multicollinearity are the NHS & FFS dummy variable. Because of this,
both variables are dropped from the full model.
133
The presence of heteroscedasticity (White test) and autocorrelation (Godfrey test)
indicate that there are unmodeled country and time specific effects. In order to
solve this problem, time and country dummies can be introduced. Country
dummies control for country specific differences that are not included in the model
and, time specific dummies control for time related effects not captured by the
model.
The next step is to determine whether these country and time specific effects are
random or fixed. Random effect models are used when the effect are the result of a
random selection of units of analysis from a population about which inferences are
made while fixed effect models are used when inferences are made about the units
of analysis included in the data only (Jones, 2000). Here, the selected countries do
not come from regular random sample but are selected based on the availability of
data and the comparability of countries. Therefore, a fixed effect model is the most
appropriate statistical method. In addition, previous research (Gerdtham, 1992) has
tested different statistical approaches on a previous version of the OECD data and,
the two way fixed effect models yielded the most robust results. In order to
confirm the appropriateness of this technique, a Hausman test (1978) for fixed
effects was performed on all models resulting in significant values meaning that a
fixed effect is an appropriate method to analyze the models (Table 3.1.).
134
3.4. Results.
3.4.1.The Unrestricted Model.
Table 3.2. shows the results of the model that includes variables representing
income, financing characteristics, hospital payment method, expenditure
components and democratic controls. Because of high multicollinearity with
public share of total health expenditures the health system dummy and the
physician payment dummy variable cannot be included in the model. At the outset,
it should be noted that the unrestricted model does not include Austria, Greece,
Ireland, Netherlands, Portugal, Sweden and United Kingdom for both the 1970 –
2000 and the 1980 – 2000 sample.
Independent of what sample is used, per capita GDP is positively related to per
capita total health expenditures as predicted. All estimates are significant and
below unity. However, when the United States is included, the income effect on
total health expenditures is smaller for both samples.
With respect to the source of financing, public share of total health expenditure is
never significant and has always a negative sign except for the 1970 – 2000 sample
that does not contain data on the United States. With respect to the demand side
financing variable out of pocket payments, the estimates show a more consistent
135
pattern. For all the samples, all estimates are negative and significant except for
the 1970 – 2000 sample for which it was not significant. This indicates that out of
pocket payments have an expenditure decreasing effect.
Table 3.2. LSDV Estimates of the Unrestricted Model. (Cross Section and Time
Series Effects not Included)
No US No US US US
1970-2000 1980-2000 1970-2000 1980-2000
Intercept -3.26 -4.44 2.62 -1.97
-1.46 -1.48 1.4 -0.76
GDP 0.76
a
0.31
a
0.68
a
0.32
a
6.17 2.55 5.77 2.99
Pubshare 0.14 -0.004 -0.23 -0.26
0.58 -0.02 -1.13 -1.57
Inpatient 0.19
a
-0.11
c
0.18
a
-0.12
a
2.28 -1.75 2.24 -2.16
Pharma -0.18
a
-0.27
a
-0.17
a
-0.28
a
-3.63 -6.52 -3.5 -7.42
Kidney -0.02 -0.02 -0.006 -0.02
-1.59 -1.42 -0.49 -1.53
Global 0.01 -0.01 -0.01 -0.02
0.77 -1.05 -0.82 -1.43
DRG -0.01 -0.01 -0.01 -0.01
-0.6 -0.91 -0.81 -0.85
Pocket 0.0001 -0.21
a
-0.3
a
-0.29
a
0 -2.85 -5.46 -6.25
Pop65 -0.02 0.24
a
0.12 0.29
a
-0.29 2.87 1.43 4.06
Urban 0.52
c
2.28
a
-0.05 2.12
a
1.64 4.94 -0.21 5.07
R-Square 0.9948 0.9924 0.9949 0.9946
Obs 174139200 159
N 330220360 240
% Missing 47.27 36.82 44.44 33.75
(
a
≤ 0.01,
b
≤ 0.05,
c
≤ 0.1)
136
With respect to variables representing the different expenditure components of a
health care system, the proportion of expenditure spent on inpatient care yields
different results dependent on the sample. In two out of four samples, the
estimates are significant indicating that inpatient care is a driver of total
expenditures. The only samples that have significant negative estimates are the
1980 – 2000 samples. Independent of the sample the share of pharmaceutical
expenditures is consistently significant and inversely related to total health
expenditures regardless what sample is used. The number of kidney transplants is
also inversely related with total per capita health expenditures but is never
significant.
Finally, with respect to the hospital payment methods global budgets are
significantly related with lower total per capita health expenditures except for the
1970 – 2000 sample without the United States where the estimate is negative but
not significant. However, DRGs are never significant although always negative.
3.4.2. Restricted Model 1.
This model includes data for all the 19 selected countries and includes the health
system and physician payment method dummy variable.
137
Table 3.3. confirms the findings with respect to the health expenditure increasing
effect of income in terms of per capita GDP. All estimates are significant and
below unity. As is the case in the unrestricted model, the estimates drop in value
for the samples when the United States is included.
Public share of total health expenditures shows unstable results with respect to its
impact on total health expenditures as is the case in the unrestricted model. Except
for the 1980 – 2000 sample when the United States is not included, all the
estimates are negative but only significant in the 1970 – 2000 sample when the
United States is omitted.
The estimates of the variables representing the components of health expenditures
are stable in sign. Pharmaceutical expenditures as a percent of total health
expenditures is highly significant and inversely related to total health expenditures.
However, in the 1980 – 2000 samples the estimated impact of pharmaceutical share
is higher than in the 1970 – 2000 sample. The kidney transplant variable is also
consistently inversely related to health expenditures but only insignificant in the
1970 – 2000 sample that include the United States.
138
Table 3.3. LSDV Estimates for Restricted Model 1. (Cross Section and Time Series
Effects not Included)
No US No US US US
1970-2000 1980 - 2000 1970 - 2000 1980 - 2000
Intercept -3.18
a
-0.71 -2.68
a
0.41
-4.82 -1.08 -3.8 0.6
GDP 0.94
a
0.54
a
0.88
a
0.48
a
15.27 7.73 13.48 6.88
Pubshare -0.31
a
-0.06 -0.11 0.09
-3.6 -0.67 -1.33 1.13
Pocket
NHS 0.05 0.04 0.03 0.02
1.21 1.12 0.68 0.66
Inpatient
Pharma -0.19
a
-0.33
a
-0.22
a
-0.34
a
-7.13 -10.88 -7.67 -11.3
Kidney -0.02
b
-0.04
a
-0.01 -0.03
c
-2.02 -3 -1.59 -2.39
FFS 0.28
a
0.11
a
0.28
a
0.1
8.03 3.39 7.65 3.12
Global -0.004 -0.01 -0.03
b
-0.02
a
-0.29 -0.87 -2.41 -1.7
DRG -0.01 0.01 -0.002 0.01
-1.37 0.97 -0.2 1.22
Pop65 0.11
b
0.19
a
0.07 0.15
a
2.25 3.65 1.29 2.94
Urban 0.58
a
0.72
a
0.53
a
0.67
a
6.36 8.12 5.44 7.43
R-Square 0.9936
a
0.9899
a
0.993
a
0.9906
a
Obs 396 314 423 335
N 558 378 589 399
% Missing 29.03 16.93 28.18 16.04
(
a
≤ 0.01,
b
≤ 0.05,
c
≤ 0.1)
139
With respect to payment methods of providers, countries with a predominant fee
for service system to remunerate physicians have higher total health expenditures
than countries using other payment methods. With respect to hospitals, the
introduction of global budgets has a expenditure lowering effect. In other words,
countries using global budgets for hospital financing have lower health
expenditures than those not having global budgets. However, the estimates are
only significant when the United States is included in the sample. The introduction
of DRGs does not significantly affect total health expenditures. In addition the
estimates are positive in the 1980 – 2000 sample.
Finally, an ageing population has a significant expenditure increasing effect in
three of the four samples and, higher degrees of urbanization is linked with higher
total health expenditures.
3.4.3. Restricted Model 2.
As was the case in the two previous models, higher per capita income is associated
with higher per capita total health care systems and in the 1970 – 2000 sample the
estimates are higher compared to the 1980 – 2000 sample.
140
The influence of public share of total health expenditures is less consistent, as is
the case in the other models. If the United States is excluded, this variable is
inversely related to total per capita health expenditures in both samples. However,
the estimate is only statistically significant in the 1970 – 2000 sample, indicating
that over that period of time an increase of 10 percent of public financing has had a
2.6 percent decreasing effect on total per capita health expenditures. When the
United States is included the estimates become insignificant and even has an
opposite sign in the 1980 – 2000 sample.
The health system dummy variable has an expenditure increasing effect and is
always significant independent of the sample used. This means that during the
periods 1970 – 2000 and 1980 – 2000, countries with National Health Service had
on average 1% higher health expenditures than those who had not.
With respect to payment methods, the results of table 3.4 indicate that countries
predominantly relying on fee for service remuneration of physicians have higher
health expenditures than those who rely on other payment methods. The estimates
are higher in the sample covering the larger time span regardless of whether the
United States is included or not. Except for the 1970 – 2000 sample, the use of
global budgets has a significant lowering effect on total health expenditures.
DRGs do not have an impact on total health expenditures.
141
Table 3.4. LSDV Estimates of the Restricted Model 2. LSDV Estimates for
Restricted Model 1. (Cross Section and Time Series Effects not Included)
No US No US US US
1970-2000 1980 - 2000 1970 - 2000 1980 - 2000
Intercept -4.44
a
-2.5
a
-3.86
a
-1.59
b
-7.44 -3.74 -6.05 -2.26
GDP 1.03
a
0.72
a
0.95
a
0.67
a
18.16 9.55 16.24 9.03
Pubshare -0.26
a
-0.06 -0.09 0.07
-3.53 -0.66 -1.23 0.82
Pocket
NHS 0.1
a
0.09
a
0.09
a
0.08
b
4.86 2.65 4.14 2.33
Inpatient
Pharma
Kidney
FFS 0.27
a
0.15
a
0.26
a
0.14
a
7.62 3.94 6.98 3.72
Global -0.0004 -0.03
c
-0.02
b
-0.04
-0.06 -1.7 -2.01 -2.43
DRG -0.004 -0.01 0.01 -0.01
-0.34 -1.06 1.13 -0.84
Pop65 0.09
b
0.11
b
0.05 0.09
c
2 2.03 1.07 1.66
Urban 0.48
a
0.56
a
0.48
a
0.53
a
8.73 7.38 8.21 6.93
R-Square 0.9927
a
0.9839
a
0.9921
a
0.9852
a
Obs 520 363 551 384
N 558378589399
% Missing 6.81 3.97 6.45 3.76
(
a
≤ 0.01,
b
≤ 0.05,
c
≤ 0.1)
142
Finally, urbanization and an ageing population have an increasing effect on total
per capita health expenditures. Urbanization is consistently significant
independent of sample while ageing is only significant for the 1980 – 2000 sample.
3.5.Discussion.
In this section, three aspects related to this study are discussed. First, issues with
respect to the nature of comparative research are addressed. Then, the differences
regarding the estimates in different models and samples are reviewed. Finally, the
actual results of the regression analysis are discussed and related to the existing
theoretical framework presented in the section preceding the analysis.
3.5.1. Issues in Comparative Data and Model.
3.5.1.1 Working Comparatively
Working at an aggregate level of countries brings some advantages as well as
disadvantages with it. Choosing a comparative approach also means choosing for
problems that accompany this approach and is limited in several ways. (Heclo,
1990) First, there is limited by the number of subjects being part of the analysis.
143
As Ragin (1987, 12) states “as the number of observations decreases, the
possibility of subjecting arguments to rigorous statistical testing diminishes”.
Consequently, most of the possible pitfalls relate to methodological issues caused
by the relative small sample size. The units of analysis are macro-social units such
as states, countries or governments, and there are only a limited number of these
units while most statistical analyses assume rather large samples. (Janoski &
Hicks, 1996; Ranade,1998; Hill, 1996) A second issue is related to availability of
comparable data. The essence of the problem is that certain variables could be
measured differently in different countries. In addition, it is always hard to
actually know whether and or when a certain policy implemented. Third, there is
the problem of uniqueness which means that in order to make comparisons, certain
specific features of a country may have to be (Heidenheimer et al., 1990).
Although in this study the problems mentioned have been countered through the
selection of countries and the introduction of time and country specific It is
important at the outset to point at the possible issues with respect to the nature of
comparative research.
144
3.5.1.2. Sensitivity of the models to different samples and countries included.
The Jarque-Bera test indicated that the errors were most of the time normally
distributed. This normallity indicates that there is no influential unit of analysis.
However, in most of the models and samples there are differences in the estimates
when the United States is included meaning that the United States most likely is an
influential outlier.
Furthermore, within each sample signs and significance tend to change.
Because of the nature of the unit of analysis and the differences in signs and
significance estimates among models and datasets, the interpretation and the
implications of the results become more difficult to identify.
3.5.2. Discussion of Results.
3.5.2.1. Elasticity and the Role of Government in Health Financing.
Two of the implications of the Newhouse study (1977) was that health care with an
elasticity larger than unity was a luxury good rather than a necessity good and, that
since GDP was explaining more than 90% of variation in total health expenditures
the effect of other variables was probably close to none.
145
With respect to the second finding, that income is most likely the sole explanatory
variable explaining total expenditures, this study confirms the results of most
inquiries since the Newhouse study on the determinants of total health
expenditures.
With respect to the first conclusion, the results from this analysis show that
elasticity does not exceed unity - except for one model - confirming most of the
results of previous research since the Newhouse study.
Many discussions of studies cover this issue leading to a yes-no discussion on
whether health is actually a necessity good. Comparative multivariate analysis
showed sometimes that health care was a luxury good (Murillo, C. et al., 1993;
Kanavous & Yfantopoulos, 1999; Murthy & Okunade, 2000; Auteri & Constantini,
2004) and other times that it was a necessity good (Condon et al., 2000; Kotzian,
2003; Sen, 2005; Dreger & Reimers, 2005). Still others found differences between
short and long term (Okunade & Karakus ,2001), between public and private health
expenditure (Attela & Marini, 2002) and between high and low-income countries.
(Musgrove et al., 2001; Poullier et al., 2002)
The wide spectrum of results with respect to the nature of health care as a good is
caused by two reasons. First, the wide variety of results can be explained the
146
different statistical analysis used to determine the elasticity of health care.
(Kanavos & Mossialos, 1999)
A second reason is related to the level of analysis. (Kavanos & Mossialos, 1999;
Getzen, 2000) In contrast to aggregate studies, household surveys show that health
care is a necessity good. The reason for these rather contradictory results is
twofold. On the one hand, micro studies investigate the behavior of individuals
and make inferences about the average individual while the macro studies
concentrate on the average behavior of a pool of individuals. On the other hand, at
the different levels of analysis, the income effect is experienced in a different way
because of the level where the budget restrictions take place. At the micro level,
the average individuals are much more sensitive to changes in income and
therefore the spending effect on health. At the macro level, spending restrictions
are set through a budget process for which individuals are not directly sensitive.
Because of those reasons, different results may be obtained at different levels of
analysis.
The elasticities in this study are lower than unity indicating that health care is most
likely not a luxury good but when growth rates of health expenditures are
compared with those of income, health expenditures growth rates are higher than
those of income indicating that health care is a luxury good (Table 2.2.). However,
the simple fact that that an elasticity does or does not exceed unity only means that
147
as income increases, the increase in health care spending will be relatively higher
than the increase in income. This does neither mean that health care is a necessity
good nor that it not a necessity good in way we use these concepts in our daily non-
economic language. As such, a legitimate role in the financing of health care
cannot be based on whether a good is considered a luxury good or not in the
economic sense.
However, a role for government intervention in health care financing can be based
on the Musgrove’s theory (1996) of the role of government in financing health care
based the nature of the good and the likeliness of market failure because of
externalities and asymmetric information as well as equity reasons.
The empirical results from this study show do not fully support that an increase of
percent of government share of total health expenditures can lead to a decrease in
total health. This is however only the case for the samples that do not include the
United States and cover the longer time span. For all the other samples, there is a
not significant effect of public share on total health expenditures indicating that
this impact is most likely a seventies phenomenon.
From the analysis, all but three of the estimates are negatively and mostly not
significantly related to the dependent variable is thought provoking. One possible
reason why publicly financed health care may reduce expenditures is lower
148
administrative costs. All countries that have statutory health insurance also have
standardized levels of covered care for almost the whole population. The financial
resources are generated directly through taxes and/or contributions from employers
and/or employees and, are then pooled on one account. From that account,
government either pays providers directly as is the case in single payer systems or,
transfers the money to social insurances in third payer systems that are responsible
for reimbursing covered health services.
Public and private insurance schemes have administrative costs caused by
processing claims. Commercial health insurances are more likely to have higher
administrative costs that are caused by rating systems and premium structures that
have to be developed and sometimes tailored to the purchasers’ preferences. There
are sales and marketing costs to sell the insurance in a competitive market as well.
Finally, there are the costs of complying with government regulation. (Pauly et al.,
1991, 1999; Thorpe, 1992)
Comparisons between countries show that in the United States administrative costs
of health insurance are higher than in any other country (Himmelstein &
Woolhandler, 2003; Reinhardt et al., 2004; Woolhandler et al., 2003; Woolhandler
& Himmelstein, 2002). In addition, a study on voluntary private insurance in the
European Union showed that administrative costs of private insurance schemes are
149
higher than their public counterparts (Colombo & Tapay, 2004; Mossialos &
Thompson, 2002).
If there is a legitimate role for government in financing health care, the next
question then, is whether the private sector has a role to play in a health care
system. The results from this analysis support a role for private entities. The
health system variable that is indicative of public predominance in a health care
system is assumed to capture also the public provision of health services. The
results show that the presence of an NHS is an expenditure increasing institutional
arrangement in the health care system if the assumption mentioned above holds
strong. This supports the public choice theory that states that public providers
show monopolistic behavior causing inefficiencies, which is supported by results
from research in two areas on the efficiency of health care providers.
First, waiting times is one manifestation of the inefficiency of public providers.
Several studies have looked at differences between countries for referral time lags
(Fleming et al., 1992) and waiting times for different surgical procedures. (Blendon
et al. 2002; Carroll et al., 1995; Siciliani & Hurst, 2003; 2004) Results show that
in countries with public provision of medical care referral and waiting times are
longer.
150
In addition, empirical studies on hospitals have found that in NHS health systems
hospitals behave as bureaucracies. This resulted in budget maximizing behavior as
well as misallocation of resources that on its turn resulted in over utilization of
supplies and over investment in equipment leading to excess capacity. (Rodriguez
et al., 2004; Spicer, 1982)
3.5.2.2. The Expenditure components.
The results with respect to the components of health expenditures are puzzling.
Especially the unexpected negative sign of proportion of total health expenditure
spent on inpatient care. The estimates are not robust since significance depend on
whether or not the United States is included and signs change in the 1970 – 2000
dataset. This means that the more country spends on inpatient care as a percent of
total health expenditure will have an expenditure lowering effect, which is against
common sense.
In the expenditure and the unrestricted model, the other two expenditures structure
variables have negative estimates. With respect to pharmaceutical expenditures,
the results support the substitution theory that argues that drugs can have an
expenditure lowering effect because it has the potential to prevent more costly
151
medical care later on. An increase of 10 percent in pharmaceutical expenditure can
reduce health expenditures between 1.7 and 3.2 percent.
Kidney transplants representing the use of technology and testing whether the
availability of technology increases total health expenditures through higher
utilization is inversely related to health expenditures. This means that a higher
availability of technology not necessarily leads to higher expenditures, at least if
the variable is a good proxy for technology. Some studies have shown that certain
technology intensive procedures have more benefits than costs (Cutler & Meara,
1999; Mohr et al., 2001).
Finally, all the variables included in the model to measure the structure of health
expenditures have in common that they most of the time are inversely related to
total health expenditures. This may come as a surprise but in itself is not that
surprising when one considers the nature of these variables. All three variables
have also in common that they are technology intensive. Although there is no
evidence of severe multicollinearity, this may explain the similarities in the
direction of the estimates.
152
3.5.2.3. Supply or Demand Side Policies?
With respect to the supply side, the results show an expenditure decreasing effect
for policies aimed at hospitals but the estimates are not always significant. Global
budgets seem to have a larger impact although with magnitudes of the estimates
being rather small. Prospective payment systems such as DRGs never have a
significant impact.
This means that countries with global budgets on hospitals have lower
expenditures. Limiting access to open ended financial resources could be an
effective way to limit health expenditures. One should be careful with this type of
inferences. Global budget is a dummy variable whether or not a country uses a
global budget method to finance inpatient care. By now means is it an indicator of
how fixed global budgets actually are fixed. Neither does it say how flexible they
are. Some countries use global budget more as targets than actual caps on hospital
spending and that case budget are most likely not to have an impact on total health
expenditures.
In this study, prospective payment methods like DRGs, appear not be very
effective in affecting total health expenditures. Although the sign has most of the
time the expected sign, the variable is never significant. As in the case of global
153
budgets, one should be cautious with interpreting these results. DRGs are not
universal and do differ substantially from country to country. For example, in the
United States, DRGs originally aimed at hospitals treating Medicare patients. On
the aggregate level, Arras et al. (1995) and Arno et al. (1995) showed that it was
unlikely that DRG had an effect on total health expenditures and most of the slow
down of growth in health expenditures was due to the introduction of managed
care.
The open-ended nature of physician payment through a predominant fee for service
remuneration system contributes to higher total expenditures. On average health
care systems where physicians not paid through the means of a salary and/or
capitation spend on average between 1.4 and 2.7 percent more on health care.
With respect to the demand side, an increase in out of pocket payments also seems
to be an important policy tool for controlling health expenditures. When compared
to supply side policy instruments, out of pocket payments seemingly affects total
health expenditures more than global budgets and DRGs.
154
3.5.2.4. How Important are Demographics?
Unlike most comparative multivariate analysis, this analysis has rather robust
estimates in the expected direction with the impact of an ageing population of total
health expenditure. Urbanization is also positively related to total health care
expenditures. Both variables have relative robust estimates, with the exception of
the datasets with relatively large missing values that may explain the switching
signs. However, whenever turning negative both variables are never significant.
155
Conclusion
1. A Look Back to the Future.
If we look back at the last three decades, mortality rates have improved but it came
at a cost. This is the case for all countries but is even more so for the United States
outspending any other country on the globe. It is also undeniable that the United
States had much higher total expenditures to start with. With respect to the future,
the United States’ health predicament becomes ominous for the health of many
people.
In terms of expenditures, projections for the United States estimate that the percent
of GDP spent on health will grow to from 15.3% in 2003 to 18.4% in 2013 and
18.7% in 2014 (Blendon et al. 2004; 2005). This would mean that the United States
would spend $ 3,585.7 billion on health care annually. On average, every
American would spend $11,043 per capita in 2014 on health care.
Projections also show that that the number of uninsured Americans will continue to
grow. The number of uninsured non-elderly is estimate to climb from 45 million in
2003 to 56 million in 2013. Moreover, the cause of it is the increase in the cost of
health insurance (Gilmer & Kronick, 2005).
156
In addition, Americans are unsatisfied with their health care system. In 1989, 30%
of the participants of a survey responded that there was so much wrong with the
American health care system that it should be redesigned completely. Moreover,
almost 90% of the respondents wanted a different system than the current health
care system (Blandon & Taylor, 1989). A follow up study in 1995, showed that
dissatisfaction with the health care system remained the same. Compared to other
countries, US respondents outscored respondents in Canada and Germany on all
items of the survey. More Americans thought that the United States did not spend
enough on health care, that the system was too bureaucratic, that they were treated
unfairly by the system, that the quality of care was poor, that waiting times were
too long and, that they had limited choice of provider (Blendon et al. 1995). In
another follow up study, general dissatisfaction of Americans remained about the
same but managed care enrollees were more dissatisfied (Blendon et al., 1998;
1999).
Other studies have looked at sub-populations. While fewer elderly in the United
States were dissatisfied with the health care system than the global population
(Blendon et al, 2000), sick respondents were more dissatisfied (Blendon et al,
2003). In both cases, the United States had the highest percent of dissatisfied
respondents compared to the other countries. `However, fewer physicians in the
United States were dissatisfied with the health care system than in the other
countries included in the study.
157
In summary, the United States spends more than any other industrialized country
on health care. However, when compared to other countries, the United States
scores low in terms of satisfaction. Since the number of uninsured is expected to
increase, the level of satisfaction with health care system as a whole is not likely to
improve. This raises the question: What is so different about the United States
health care system that makes it so expensive and so disliked?
2.How is the United States Alike?
The United States’ health care system does not differ that much from other health
care systems with respect to organization and delivery of medical care. First,
governments are involved at several organizational levels with respect to the
provision of health care and the design of policy frameworks including health
insurance. Countries only differ in terms of degree of involvement.
Second, policies to contain costs with respect to the different subsystems are alike
in the United States and in other countries. On the demand side, out of pocket
expenses have increased over the years and, general practitioners serve as
gatekeepers to reduce self-referral to specialists and hospital care in all countries.
On the supply side, physicians are less and less paid on a fee for service basis, and
receive more and more of their remuneration through capitation and/or salary.
158
With the exception of the United States and Japan hospitals in all countries have
been subjected to prospective payment systems such as DRGs and global budgets.
The share of inpatient care has been reduced and replaced by less expensive
outpatient procedures. Here too, with the exception of global budgets, the United
States is not that different from any other country.
Third, with respect to providers, physicians operate in most countries as
independent private medical practitioners. If they are government employees, they
most of the time are allowed to have private patients. Some countries have more
public inpatient facilities than others. In most countries, private providers have a
predominant role. This is also the case in the United States. What is distinctive is
the difference in degree of public involvement. In the United States, public
involvement in inpatient care is much smaller and most private providers are for
profit.
Fourth, in all countries - with one exception - a limited choice of health care
provider is a myth. In addition, all countries have voluntary insurance to
complement or supplement statutory health insurance. Only two countries allow
individuals to opt out of the (mandatory) statutory insurance if certain income
requirements are met. In other words, the free choice that is inherent in the
voluntary nature of the United States’ health care system is also present in the other
systems.
159
Fifth, technology and pharmaceutical related expenditures have been increasing
considerably but – at least from this analysis – do not seem to be the driving forces
of higher health expenditures. On the contrary, both seem to have a decreasing
effect on total health expenditures indicating that the benefits outweigh the costs on
the long term. Finally, there are demographic changes. All countries are
experiencing an ageing population that affects their total health expenditures.
3. How is the United States Different and, does it matter?
The major difference that characterizes the United States’ health care system is the
role that government plays with respect to the financing of the health care system.
The minimal public involvement in health care in the United States would not be
that of an issue if it were not for the fact that public involvement might have an
expenditure decreasing influence as the multivariate analysis demonstrated. It is in
this sense that government involvement in health care financing matters. There is a
theoretical foundation for government involvement in health care financing.
However, there is rather inconclusive empirical evidence that with respect to health
care, government action may result in better outcomes in terms of expenditures
than markets.
160
One of the factors that contribute to the high expenditures in the United States is
not related to the private sector but relates to the nature of government involvement
in health care financing. Currently, the United States’ public sector is financing
health care for the poor and the elderly. This puts extra financial pressure on the
American government for two reasons. On the one hand, the US government
covers a population that is more likely to have higher health care bills resulting in
disproportionately higher costs than other insurers. On the other hand, most
regulatory measures issued apply only to a fraction of the population (Medicare
and Medicaid), limiting the possible impact of government action on total health
expenditures.
The voluntary nature of health insurance for the non-poor and the non-elderly also
may increase the total costs. Since individuals have the choice to insure
themselves and, can decide the degree of coverage, there is the presence of under
and uninsured individuals. (Whether it is their own or their employer’s choice)
Uninsured and underinsured may end up increasing health expenditures since they
most likely end up receiving health care at a point in time at which curing or
treating an illness is much higher.
The United States values individual choice and competition more than any other
countries. The nineties were characterized by increased competition aiming at a
more efficient use of resources and the reduction of the costs of health care. In this
161
policy, managed care organizations became key players and throughout the
nineties, the policy seemed to have a cost reducing effect. However by the start of
the millennium expenditures started to increase again.
4. The Sisyphus Labor of Changing the United States’ Health Care
System.
Health care has been on the political agenda election after election. While always
an important domestic issue, once the health care reform takes the format of actual
proposal it becomes a target of attacks from all political sides. This inevitably has
led to the defeat of fundamental health system changes since the birth of Medicaid
and Medicare in 1965. The result is that the United States’ health care system
remains as it was.
One reason may be the fear of what is still making Americans tremble and shiver.
The concept of “Socialized medicine” still is a concern for most Americans, but
should be placed in the context in the ideological context of the cold war. (Starr,
1992) The health care systems of other countries show that government among
other sources can finance a health care system predominantly, without limiting the
provision by private practitioners, the individual choice of provider and, it has a
cost lowering effect.
162
Another reason may be what Reinhardt (2000, 71) describes as “Americans tend to
be unimpressed by cross-national comparisons of health care systems on the axiom
that American health care is so vastly superior to that anywhere else on the globe
as to render any cross-sectional comparison irrelevant for American health policy”.
Together with the belief in personal choice, incremental policy approaches to
reform the health care system have been focused on a competitive approach.
However, health care costs have been growing again and, it may well worth to look
at policies of other countries.
5. The Ideal Health Care System Reform.
An ideal health care system is a system that finds balance between access, cost and
quality. In other words, individuals should have access to health care of an
acceptable level of quality at a reasonable cost.
Most countries have established mandatory statutory health insurance to assure
reasonable access to health care for virtually the entire population. Not only is
insurance mandatory, the medical care that is covered under the statutory health
insurance is also well defined. Additional insurance to supplement the basic
package is available and, can be purchased from private insurance companies.
163
Expenditures in universal health care system are a top down process in which
governments limit the available resources for health care through annual budget
allocation. In addition, supply as well as demand disincentives are used to control
expenditures at the organizational and individual level.
Quality of care is most of the times established by governments through standards
imposed health care providers. Hospitals as well as physician must comply with
these requirements such as accreditation and licensing in order to practice
medicine. In addition, sometimes information dissemination on quality indicators
such as hospital mortality is used to voice the quality of care. At the macro level,
the goal is to make quality care available to a population. At the organizational
level, quality is monitored through establishing quality assessment criteria and
practice guidelines. Over the years, techniques have been developed focusing on
structure process and outcomes (Donabedian, 1980; 2003) to measure the quality
of care.
With respect to the Unites States, the availability of quality of care is on itself not
problematic. The United States has more state of the art facilities and is at the
cutting edge of medical research resulting in advanced treatment than any other
country the world (Shuster et al., 1998). The issue in the United States is that there
is limited access to regular care at a reasonable cost. With a predominant role of
164
the private sector in financing and delivery of health care, government action is
needed not only in the light of equity but also in the light of controlling health
expenditures given the impact of share of public health expenditures on total health
expenditures.
6.Final Remarks.
Issues related to health care have most of the time been squeezed between two ends
of a continuum. This dichotomized way of thinking has often led to polarized
ideological views in terms “either … or” with respect to public and private
involvement in health care. Especially in the United States, this thinking in
extremes is strong with respect to health care and it has been translated in practice:
government finances health care of the poor and the elderly from one account,
while the private sector finances health care of the employed from their own
account.
From a health economics perspective, this has led to intense pro market versus pro
government arguments based on their failure theories. This often resulted in those
favoring the market that government is to blame if something goes wrong and,
those favoring the government to blame the market if something goes wrong.
165
However, in our day and age things are more complex. The “invisible hand” has
obviously visible and unwanted effects and, the “iron fist” has them too.
In health care, most other countries have walked the “third way” in which public
and private sector go hand in hand and complement each other. In NHS and NHI
countries health care is financed through taxes that are generated from private
firms (large and small) and individuals (old and young, poor and rich) and put into
one account, sometimes supplemented with social contributions. In Social
Insurance countries, resources for financing health care are generated through
social contributions from employees and employers, sometimes supplemented with
general taxation and put into one account that sometimes on its turn is distributed
among the social insurances.
The issue of financing health insurance is not only important for the United States
but it is equally important for an expanding European Union. While movement of
people and free trade of goods and services within the European economic area has
improved, health insurance is the service that is almost only a national matter. For
a European Union with a growing population from “new economy countries” it
will be important as well as a challenge to be able to provide accessible quality
care at a reasonable price to all.
166
In summary, when designing a health care system some arrangements are preferred
over others with respect to containing total health expenditures. First, government
can have an important role to play in terms of preventing market failures to take
place through public financing. However, there is no conclusive empirical
evidence that suggests that higher public involvement in health care financing will
result in lower total health expenditures. Second, the public sector is likely to be
an inefficient provider of health services. Third, pharmaceutical share of total
health expenditures may substitute for more expensive medical procedures and,
technology may do so too. Finally, demand side such as physician payment
policies may be more effective to control total health expenditures than hospital
financing policies such a DRGs and global budgets.
165
Bibliography
Abel-Smith, B. (1992) Cost Containment and New Priorities in the European
Community., In Milbank Quarterly., 70: 393 – 422.
Abel-Smith, B. & Mossialis, E. (1994) Cost Containment and Health Care Reform.
A Study of the European Union, Occasional Paper I Health Policy No 2, London,
London School of Economics & Political Science.
Alemayehu, B. & Warner, K.E. (2004) The Lifetime Distribution of Health Care
Costs. In Health Services Research, 39: 627 – 642.
Anell, A. & Svarvar, P. (1999) Health Care Reforms and Cost Control in Sweden.
In Mossdialos, E. & Le Grand, J. (Eds.) Health Care and Cost Containment in the
European Union. Aldershot: Ashgate.
Anell, A. (2005) Swedish healthcare under pressure. In Health Economics. 14:
S237–S254.
Angeletti, M.A., Italy’s Health Care System. (2001) In Thai, K.V., Wimberly, E.T.
& McManus, S.M. (Eds.), Handbook of International Health Care Systems. New
York: Marcel Dekker Inc.
Arno, P. S. Bonuck, K. & Padgug, R. (1995) The Economic Impact of High-Tech
Home Care. In Arras et al. (Eds.) Bringing the hospital home. Baltimore: The
Johns Hopkins University Press.
Arras, J.D. Porterfield, H.W. & Porterfield, L. O. (1995) Bringing the hospital
home: Ethical and social implications of high-tech home care. In Arras et al. (Eds.)
Baltimore: The Johns Hopkins University Press.
166
Barros, P.P. (1998) The Black Box of Health Care Expenditure Growth
Determinants. In Health Economics, 7: 533 – 544.
Beck, N. & Katz, J. N. (1995) What to do (and not to do) with time series cross-
section data., In American Political Science Review, 89: 634-647.
Beck, N., Katz, J. N., Alvarez, R. M., Garret, G. & Lange, P. (1993) Government
partisanship, labor organization, and macroeconomics performance: a corrigendum,
In American Political Science Review, 87: 945-948.
Berndt, E. R. (2001) Pharmaceutical Industry: Why Major Growth in Times of
Cost Containment., In Health Affairs., 20: 100 – 114.
Blank, R.H. & Burau, V. (2004) Comparative Health Policy, New York: Palgrave
Macmillan.
Blendon, R.J. & Taylor, H. (1989) Views on Health Care: Public Opinion in Three
Nations. 8: 149 – 157.
Blendon, R.J., Benson, J., Donelan, K., Leitman, R., Taylor, H., Koeck, C. &
Gitterman, D. (1995) Who Has the Best Health Care System? In Health Affairs.
14: 220 – 230.
Blendon, R.J., Schoen, C., DesRoches, C.M., Osborn, R., & Zapert, K. (2003)
Common Concerns amid Diverse Systems: Health Care Expenditures in Five
Countries. In Health Affairs. 22: 106 – 121.
Blendon, R.J., Schoen, C., Donelan, K., Osborn, R., DesRoches, C.M., Scoles, K.,
Davis, K., Binns, K. & Zapert, K. (2001) Physicians’ Views on Quality of Care: A
Five-Country Comparison. In Health Affairs. 20: 233 – 243.
Bloom, A.L. (2000) Context and lead-up to Health Reform. In Bloom (Ed.) Health
Reform in Australia and New Zealand., Oxford: Oxford University Press.
167
Bloor, K. & Maynard, A. (2002) Universal Coverage and Cost Control. The
United Kingdom National Health Service. In Thai, K.V., Wimberly, E.T. &
McManus, S.M. (Eds.), Handbook of International Health Care Systems. New
York: Marcel Dekker Inc.
Bonoli, G. (1997) Classifying Welfare States: A Two Dimension Approach. In
Journal of Social Policy, 26, 351 – 372.
Boyd, M-A. & Sheridan, N. (2001) New Zealand. In Wieners, W.W. (Ed.) World
Health Systems. Challenges and Perspectives. Chicago: Health Administration
Press.
Brenner, G. & Rublee, D. (2002) Germany. In Fried, B. & Gaydos, L.M., (Eds.)
World Health Systems. Challenges and Perspectives. Chicago, Health
Administration Press.
Breush, T.S. (1978) Testing for Autocorrelation in Dynamic Linear Models., In
Australian Economic Papers., 17: 334 – 355.
Buchanan, J.M. & Tullock, G. (1962) The Calculus of Consent: Logical
Foundations of the Constitutional Democracy. Ann Arbor: University of Michigan
Press.
Buchanan, J.M. (1969) Cost and Choice: An Inquiry in Economic Theory.
Chicago: Markham.
Busse, R. & Howorth, C. (1999) Cost Containment in Germany: Twenty Years
Experience. In Mossdialos, E. & Le Grand, J. (Eds.) Health Care and Cost
Containment in the European Union. Aldershot: Ashgate.
Busse, R. (2001). Germany. In Wieners, W.W. (Ed.) In Global Health Care
Markets, San Fransisco: Jossey-Bass.
168
Carrin, G. & Hanvoravonghai, P. (2003) Provider Payments and Patient Charges as
Policy Tools for Cost-Containment: How Successful are they in High-Income
Countries? In Human Resources for Health., 1: 1-13.
Carroll, R.J. et al. (1995) International Comparison of Waiting Times for Selected
Cardiovascular Procedures, Journal of the American College of Cardiology, 1: 557
– 563.
Cassel, C.K., Rudberg, M.A.. & Olshansky, S.J. (1992) The Price of Success:
Health Care in an Ageing Society. In Health Affairs. 11: 87 – 99.
Chandler, J. (1996) The United States of America. In Wall, A. (Ed.), Health
Systems in Liberal Democracies. London: Routledge.
Colombo, F. & Tapay, N. (2004) Private Health Insurance in OECD Countries:
The Benefits and Costs for Individuals and Health Systems. OECD Working
Papers No. 15. Paris: OECD.
Christiansen, T. (2002) Organization and financing of the Danish health care
system, In Health Policy, 59: 107-118.
Christiansen, T., Enemark, T., Clausen, J. & Poulsen, P.B. (1999) Health Care and
Cost Containment in Denmark. In Mossialos, E. & Le Grand, J. (Eds.) Health Care
and Cost Containment in the European Union. Aldershot: Ashgate.
Cremer, H. & Pestieau, P. (2003) Social Insurance Competition between Bismarck
and Beveridge. In Journal of Urban Economics. 54: 181 – 196.
Crémieux, P-Y., Meilleur, M-C., Ouelette, P., Petit, P., Zelder, M. & Potvin, K.
(2005) Public and Private Pharmaceutical Spending as Determinants of Health
Outcomes in Canada., In Health Economics., 14: 107 – 116.
Cromwell, J. & Mitchell, J. (1986) Physician-Induced Demand for Surgery. In
Journal of Health Economics., 5: 293 – 313.
169
Cutler, D. & Meara, E. (1999) The Technology of Birth: Is it Worth it? NBER
Working Paper 7390. Cambridge: National Bureau of Economic Research.
Cuyler, A.J. (1971) The Nature of the Commodity ‘Health Care’ and its Efficient
Allocation. In Oxford Economic Papers. 23: 189 – 211.
Cuyler, A. (1989) Cost Supplement in Europe. In Health Care Financing Review.
Annual Supplement. 21 – 31.
Danzon, P.M, & Pauly, M.V. (2002) Health Insurance and the Growth in
Pharmaceutical Expenditures., In The Journal of Law and Economics. 65: 587 –
613.
Deber, R. & Baranek, P. (1998) Canada: Markets at the Margin. In Ranade, W.
(Ed.) Markets and Health Care. A Comparative Analysis.
Di Matteo, L. & Di Matteo, R. (1998) Evidence on the determinants of Canadian
Provincial Government Health Expenditures: 1965 – 1991. In Journal of Health
Economics. 17: 209 – 227.
Di Matteo, L. (2005) The Macro Determinants of health Expenditure in the Unites
States and Canada: Assessing the Impact of Income, Age Distribution and Time.
In Health Policy. 71: 23-42.
Dixon, A. & Mossialos, E. (2002) Health Care Systems in Eight countries: Trends
and Challenges., London: The European Observatory on Health Care Systems.
Donabedian, A. (1980) The Definition of Quality and Approaches to its
Assessment. Ann Arbor: Health Administration Press.
Donabedian, A. (2003) An Introduction to Quality Assurance in Health Care. New
York: Oxford University Press.
170
Donelan, K., Blendon, R.J., Schoen, C., Binns, K. Osborn, R., & Davis, K. (2000)
The Elderly in Five Nations: The Importance of Universal Coverage. In Health
Affairs. 19: 226 – 235.
Donelan, K., Blendon, R.J., Schoen, C., Davis, K. & Binns, K. (1999) The Cost of
Health System Change: Public Discontent in Five Nations. 18: 206 – 216.
Dror, D.M. (2000) Reforming Health Insurance. In International Social Security
Review. 53: 75 – 99.
Du Bois, R. W., Chawla, A., Neslusan, C.A., Smith, M.W. & Wade, S. (2000)
Explaining Drug Spending Trends: Does Perception Match Reality. In Health
Affairs. 19: 231 – 239.
Duffy, D. (2001) United Kingdom, In Wieners, W.W. (Ed.) Global Health Care
Markets. San Fransisco: Jossey-Bass.
Duraes, T. (2002) Portugal. In Fried, B.J. & Gaydos, L.M. (Eds.) World Health
Systems. Challenges and Perspectives. Chicago, Health Administration Press.
Elling, R.H. (1989) The comparison of health systems in world-system perspective.
Research in Social Health Care. 8: 207–26.
Elling, R.H. (1994) Theory and method for the cross-national study of health
systems. International Journal of Health Services. 24:285–309.
Elling, R. H. (1980) Cross-national study of health systems: political economies
and health care. New Brunswick: Transaction Books.
Ellis, R.P. & McGuire, T.G. (1993) Supply-Side and Demand-Side Cost Sharing in
Health Care. In Journal of Economic Perspectives. 7: 135 – 151.
171
Escare, J.J. (1992) Explaining the Association between Surgeon Supply and
Utilization. In Inquiry. 29: 403 – 415.
European Parliament (1998) Health Care Systems in the EU. A Comparative
Study. Luxemburg: European Parliament.
Evans, D.B., Tandon, A., Murray, C.J. & Lauer, J.A. (2001) Comparative
Efficiency of National Health Systems: Cross National Econometric Analysis. In
British Medical Journal. 323: 307 – 310.
Evans, R.G., Parish, E.M., Sully, F. (1973) Medical Productivity, Scale Effects,
and Demand Generation. In Canadian Journal of Economics. 6: 376 – 393.
Fattore, G. (1999) Cost Containment and Health Care Reform in the British NHS..,
In Mossialos, E. & Le Grand, J. (Eds.) Health Care and Cost Containment in the
European Union. Aldershot: Ashgate.
Feldstein, P.J. (1993) Health Care Economics. Albany: Delmar Publications.
Field, M. (1993) Success and Crises in National Health Systems: A Comparative
Approach. New York: Roudledge.
Flemming, D. M., et al. (1992) The European Study of Referrals from Primary to
Secondary Care, In Royal College of General Practitioners. London, Occasional
Paper No. 56.
Forgione,D.A., Vermeer, T.E., Surysekar, K., Wrieden, J.A. & Plante, C.A. (2004)
The Impact of DRG-Based Payments on Quality of Care in OECD Countries. In
Journal of Health Care Financing. 31: 41 – 54.
France, G., Taroni, F. & Donatini, A. (2005) The Italian health-care system. In
Health Economics. 14: S187 - S202.
Freeman, D.G. (2003) Is Health Care a Necessity or a Luxury? Pooled Estimates
from US state-level data. In Applied Economics. 35: 495 – 502.
172
Fuchs, V. (1978) The Supply of Surgeons and Demand for Operations. In The
Journal of Human Resources. 12: 35 – 56.
Fuchs, V. (1990) The Health Sector’s Share of the Gross National Product. In
Science. 247: 534 – 538.
Gawade, A.A., Blendon, R.B., Brodie, M., Benson, J.M., Levitt, L. & Hugick, L.
(1998) Does Satisfaction with Health Plans Stem from Having no Choices. In
Health Affairs. 17: 184 – 194.
Gaydos, L. M. & Fries, B. (2002) The United Kingdom. In Fried, B. & Gaydos,
L.M., (Eds.) World Health Systems. Challenges and Perspectives. Chicago: Health
Administration Press.
Gaydos, L.M. & Zanola, R. (2002) Italy. In Fried, B.J. & Gaydos, L.M. (Eds.)
World Health Systems. Challenges and Perspectives. Chicago: Health
Administration Press.
Gerdtham, U. (1992) Pooling International Health Care Data. In Health
Economics. 1: 217 – 231.
Gerdtham, U. & Jönsson, B. (2000) International Comparisons of Health
Expenditure: Theory, data and econometric analysis. In Cuyler, A. & Newhouse, J.
(Eds.) Handbook of Health Economics. Volume I. Amsterdam: Elsevier Science.
Gerdtham, U., Jönsson, J, McFarlan, M. & Oxley, H. (1994) Factors Affecting
Health Spending: A Cross-Country econometric analysis. In OECD: New
Directions in Health Care Policies: Improving Cost Control and Effectiveness.
Annex. Paris: OECD.
Gerdtham, U., Jönsson, J, McFarlan, M. & Oxley, H. (1998) The Determinants of
Health Expenditures in OECD Countries: A pooled Data Analysis. In Zweifel, P.,
(Ed.) Health, the Medical Profession, and Regulation. Dordrecht: Kluwer
Academic Publishers.
173
Gerdtham, U., Jönsson, J, McFarlan, M. & Oxley, H. (2003) Factors Affecting
Health Spending: A Cross-Country econometric analysis. In OECD: Health Care
Systems: Lessons from the Reform Experience. Paris: OECD.
Gerdtham, U., Sögaard, J., Andersson, F. & Jønsson, B. (1992) An Econometric
Analysis of Health Care Expenditure: A Cross-Section Study of the OECD
Countries. In Journal of Health Economics. 11: 63 – 84.
Gertdham, U. & Jönsson, B. (1991) Conversion Factor Instability in International
Comparisons of Health Care Expenditures. In Journal of Health Economics. 10:
227 – 234.
Getzen, T.E. (1992) Population Ageing and the Growth of Health Expenditures. In
Journal of Gerontology. 47: 98 – 104.
Gilmer, T. & Kronick, R. (2005) It's The Premiums, Stupid: Projections Of The
Uninsured Through 2013. In Health Affairs. WE 5: 143 – 151.
Godfrey, L.G. (1978) Testing Against General Autoregressive and Moving
Average Error Models When the Regressors Include Lagged Dependent Variables.
In Econometrica. 46: 1293 – 1302.
Goldberg, R. (1993) Pharmaceutical Price Controls: Saving Money Today or Lives
Tomorrow? Institute for Policy Innovation Policy Report No. 123.
Goldsmith, L. J. (2002) Canada Germany. In Fried, B. & Gaydos, L.M., (Eds.)
World Health Systems. Challenges and Perspectives. Chicago, Health
Administration Press.
Goldsmith, L. J. (2002) Canada. In Fried, B. & Gaydos, L.M., (Eds.) World Health
Systems. Challenges and Perspectives. Chicago: Health Administration Press.
Graig, L.A. (1999) The Health of Nations. An International Perspective on U.S.
Health Care Reform. Washington, D.C., Congressional Quarterly Inc.
174
Granados, A. & Gallo, P. (2001) Spain. In Wieners, W.W. (Ed.) In Global Health
Care Markets, San Fransisco: Jossey-Bass.
Green, A. & Matthias, A. (1997) Non-Governmental Organizations and Health in
Developing Countries. Basingstoke: Macmillan.
Gujararti, D.N. (1995) Basic Econometrics. New York: McGraw-Hill.
Hagen, T., Iversen, T. & Van den Noord. P. (1998) The Norwegian Health Care
System. In OECD Economics Department Working Papers No. 198, Paris: OECD
Publishing.
Hakkinen, U. (1999) Cost Containment in Finnish Health Care. In Mossialos, E. &
Le Grand, J. (Eds.) Health Care and Cost Containment in the European Union.
Aldershot, Ashgate.
Hakkinen, U. (2005) The Impact of Changes in Finland’s Health Care System. In
Health Economics. 14: S101–S118.
Hall, J. (1999) Incremental Change in The Australian Health Care System. In
Health Affairs. 18: 95 – 110.
Hanush, H. (1983) Anatomy of Government Deficiencies. New York, Springer-
Verlag.
Hausman, J.A. (1978) Specification Tests in Econometrics. In Econometrica. 46:
1251 – 1271.
Heffler, S., Smith, S., Keehan, S, Clemens, K. M., Zezza, M. & Truffer, C. (2004)
Health Spending Projections for 2001 – 2013. In Health Affairs, Health Spending
Projections Through 2013 In Health Affairs. WE 4: 79 – 93.
175
Heffler, S., Smith, S., Keehan, Won, G., Clemens, K. M., Keehan, S. & Zezza, M.
(2002) Health Spending Projections for 2001 – 2014. In Health Affairs, Health
Spending Projections Through 2013 In Health Affairs. WE 5: 74 – 85.
Heidenheimer, A.J., Heclo, H. & Adams, C.T. (1990) Comparative Public Policy.
The Politics of Social Choice in America, Europe, and Japan. New York: St.
Martin’s Press.
Hemenway, D. & Fallon, D. (1985) Testing for physician-Induced Demand with
Hypothetical Cases. In Medical Care. 23: 344 – 349.
Hicks, A.M. (1996) Introduction to Pooling. In Janoski, T & Hicks, A.M. (eds.)
The Comparative Political Economy of the Welfare State. Cambridge: Cambridge
University Press.
Hill, M. (1996) Social Policy. A Comparative Perspective. New York: Prentice
Hall.
Hitiris, T. & Posnett, J. (1992) The determinants and effects of health expenditure
in developed countries. In Journal of Health Economics. 11: 173-182.
Horn, M. (1995) The Political Economy of Public Administration. Institutional
Choice in the Public Sector. Cambridge: Cambridge University Press.
Hsiao, C. (1999) Analysis of panel data, Cambridge University Press, Cambridge.
Hughes, J. (1999) Health Expenditure and cost containment in Ireland. In
Mossialos, E. & Le Grand, J. (Eds.) Health Care and Cost Containment in the
European Union. Aldershot: Ashgate.
Hurst, J. & Siciliani, L. (2003) Annex 3 Detailed Country Reviews in Tackling
Excessive Waiting Times for Elective Surgery: A Comparison of Policies in
Twelve OECD Countries. Paris: OECD Health Working Papers.
176
Iglehart, J.K. (1995) Medicaid and Managed Care. In The New England Journal of
Medicine. 332: 1727 – 1731.
Iglehart, J.K. (1999) The American Health Care System. Medicaid. In The New
England Journal of Medicine. 340: 327 – 332.
Iglehart, J.K. (1999) The American Health Care System. Medicare. In The New
England Journal of Medicine., 340: 403 – 408.
Ikegami, N. & Campbell, J.C. (2004) Japan’s Health Care System: Containing
Costs and Attempting Reform. In Health Affairs. 23: 26-36.
Iversen, T. & Kopperud G.S. (2003) The Impact of Accessibility on the Use of
Specialist Health Care in Norway. In Health Care Management Science; 6: 249–
261.
Jacobzone, S. (2000) Coping with Ageing: International Challenges. In Health
Affairs. 19: 213 – 225.
Jarque, C. M. & Bera, A.K. (1980) Efficient Tests for Normality,
Heteroscedasticity, and Serial Independence of Regression Results. In Economics
Letters. 6: 255 – 259.
Jones, A. M. (2000) Health Econometrics. In Culyer, A.J. & J. P. Newhouse, J.P.,
(Eds.) Handbook of Health Economics Elsevier, Amsterdam.
Jönsson, B. & Musgrove, P. (1993) Government Financing of Health Care. In
Schieber, G. (Ed.) Innovations in Health Care Financing: Proceedings of a World
Bank Conference. Washington DC: The World Bank.
Kanavos, P. & Yfantopoulos, J. (1999) Cost Containment and Health Expenditures
in the EU: A Macro Economic Perspective. In Mossialos, E. & Le Grand, J. (Eds.)
Health Care and Cost Containment in the European Union. Ashgate Publishing:
Aldershot.
177
Karatzas, G. (2000) On the Determinants of US Aggregate Health Care
Expenditure. In Applied Economics. 32: 1085 – 1099.
Kennedy, P. (1998) A Guide to Econometrics. Cambridge, MA: The MIT Press.
Kleiman, E. (1974) The Determinants of National Outlay on Health. In The
Economics of Medical Care. Perlman, M. (Ed.) New York: John Wiley & Sons.
Kleinke, J.D. (2000) Just What the HMO Ordered: The Paradox of Increasing Drug
Costs. In Health Affairs. 19, 78 – 91.
Kriebe, T.A. (2000) New Zealand. Journal of Health Politics, Policy and Law. 25:
926 – 930.
Kuttner, R. (1999) The American Health Care System. Employer-Sponsored
Health Coverage. In The New England Journal of Medicine. 340: 163 – 168.
Kuttner, R. (1999) The American Health Care System. Health Insurance
Coverage. In The New England Journal of Medicine. 340: 248 – 252.
Lassey, M., Lassey, W. & Jinks., M. (1997) Health Care Systems around the
World. Chartacteristics, Issues Reforms. Upper Saddle River: Prentice Hall.
Lave, J. (1984) Hospital Reimbursement under Medicare. In Milbank Memorial
Fund Quarterly. 62: 251 – 267.
Letourmy, A. (2000) Les Politiques de Santé en Europe: Une Vue d’Ensemble. In
Sociology du Travail, 42: 13-30.
Leu, R. E. (1986) The Public-Private Mix and International Health Care Costs. In
Cuyler, A.J. & Jonsson, B. (Eds.) Public and Private Health Services. Oxford: Basil
Blackwell.
178
Lichtenberg, F.R. (1996) The Effect of Pharmaceutical Utilization and Innovation
on Hospitalization and Mortality. NBER Working Paper 5418, Cambridge:
National Bureau of Economic Research.
Lopez-Casanovas, G. (1999) Health Care and Cost Containment in Spain. In
Mossialos, E. & Le Grand, J. (Eds.) Health Care and Cost Containment in the
European Union., Aldershot: Ashgate.
Lopez-Casasnovas, G. Costa-Font, J. & Planas, I. (2005) Diversity and regional
inequalities in the Spanish ‘system of health care services’. In Health Economics.
14: S221–S235.
Lubitz, J., Greenberg, L.G., Gorina, Y, Wartzman, L. & Gibson, D. (2001) Three
Decades of Health Care Use by the Elderly, 1965 – 1998. In Health Affairs. 20: 19
– 32.
Lüngen, M. & Lauterbach, K.W. (2000) Nutzung von Diagnosis-Related Groups
(DRG) im Internationalen Vergleich. In Chirurg. 71: 1288 – 1295.
Manning et al. (1987) Health Insurance and the Demand for Medical Care:
Evidence from a Randomized Experiment. In American Economic Review. 77:
251 – 277.
Matcha, D.A. (2003) Health Care Systems of the Developed World. How the
United States’ System Remains an Outlier. Westport: Praeger.
Maxwell R. (1974) Health Care, The Growing Dilemma: Needs Versus Resources
in Western Europe, the U.S., and the U.S.S.R. New York: McKinsey.
McCarthy, T.R. (1985) The Competitive Nature of Primary Care Physician
Services Market. In Journal of Health Economics. 4: 93 – 117.
179
McGuire, A., Parkin, D., Hughes, D. & Gerard, K. (1993) Econometric Analysis of
National Health Expenditures: Can Positive Economics Help to Answer
Normative Questions? In Health Economics. 2: 113 – 126.
Mikkola, H., Keskimäki, I. & Häkkinen, U. (2002) DRG-related prices applied in a
public health care system – can Finland learn from Norway and Sweden. In Health
Policy. 59: 37-51.
Miller, G.J. (1995) Managerial Dilemmas. The Political Economy of Hierarchy.
Cambridge: Cambridge University Press.
Mohr, P.E., Mueller, C., Neumann, P., Franco, S., Milet, M., Silver, L. &
Wilensky, G. (2001) The Impact of Medical Technology on Future Health Care
Costs. Final Report. Bethesda: Project Hope.
Monrad Aas, I.H. (1995) Incentives and Financing Methods. In Health Policy. 34:
205 – 220.
Mossialos E & Thompson S. (2002) Voluntary health insurance in the European
Union. Bruxelles: Directorate General for Employment and Social Affairs of the
European Commission.
Mossialos, E. (2002) Voluntary Health Insurance in the EU. In International
Journal of Health Services. 32: 19 – 88.
Mossialos, E. (2004) Voluntary Health Insurance in the EU. In Mossialos, E.,
Dixon, A., Figueras, J. & Kutzin, J. (Eds.) Funding Health Care: Options for
Europe. Buckingham: Open University Press.
Mossialos, E., Allin, S. & Davaki, K. (2005) Analysing the Greek health system: A
tale of fragmentation and inertia. In Health Economics. 14: S151–S168.
Murillo, C., Piatecki, C. & Saez, M. (1993) Health Expenditure and Income in
Europe. In Health Economics. 2: 127 – 138.
180
Murray, C.J. & Frenk, J.A. (2000) A Framework for Assessing the Performance of
Health Systems. In Bulletin of the World Health Organization. 78: 717 – 731.
Musgrove, P. (1996) Public and Private Roles in Health: Theory and Financing
Patterns. World Bank Discussion Paper No. 339. Washington DC: The World
Bank.
Musgrove, P. (1999) Public Spending on Health Care: How Are Different Criteria
Related? In Health Policy. 47: 207 – 223.
Neumann, P.J., Sandberg, E.A., Bell, C.M., Stone, P.W. & Chapman, R.H. (2000)
Are Pharmaceuticals Cost-Effective? A Review of the Evidence. In Health
Affairs., 19: 92 – 109.
Newhouse et al. (1993) Free for All? Lessons from the RAND Health Insurance
Experiment. Cambridge: Harvard University Press.
Newhouse, J.P. (1977) Medical Care Expenditure: A Cross-National Survey. In
The Journal of Human Resources. 1: 115-125.
Newton, R.R. & Rudestam, K.E. (1999) Your Statistical Consultant. Answers to
your Data Analysis Questions. Thousand Oaks: Sage Publications.
Niakis, D. (2001) The Political Economy of Health Care in Greece. In Thai, K.V.,
Wimberly, E.T. & McManus, S.M. (Eds.), Handbook of International Health Care
Systems. New York: Marcel Dekker Inc.
Niskanen, W.A. (1971) Bureaucracy and Representative Government. Chicago:
Aldine Atherton.
Niskanen, W.A. (1983) Bureaucrats between Self-Interest and Public Interest. In
Hanusch, H. (Ed.) Anatomy of Government Deficiencies, Berlin: Springer-Verlag.
181
Noseworthy, T.W. (2001) Canada. In Wieners, W.W. (Ed.) Global Health Care
Markets. San Fransisco: Jossey-Bass.
O’Connell, J.M. (1996) The Relationship Between Health Expenditure and the Age
Structure of the Population in OECD Countries. Health Economics. 5: 573 – 578.
OECD (1985) Measuring Health Care 1960 – 1983. Paris: OECD.
OECD (1987) Financing and Delivering Health Care. A Comparative Analysis of
OECD Countries. Paris: OECD.
OECD (2001) Private Health Insurance in OECD Countries: Compilation of
National Reports. Insurance and Private Pensions Compendium for Emerging
economies.
OECD (2002) Towards High-Performing Health Systems. Paris: OECD.
OECD (2003) OECD Health Data. A Comparative Analysis of 30 Countries.
Paris: OECD/CREDES.
OECD (2003), Health Data 2003. A Comparative Analysis of 30 Countries. Paris:
OECD.
OECD (2004) The Reform of the Health Care System in Portugal. Economics
Department Working Papers No. 405. Paris: OECD.
Okma, K.G.H. (2001) The Netherlands., In Wieners, W.W. (Ed.) Global Health
Care Markets. San Fransisco: Jossey-Bass.
Okunade, A., Karakus, M. C. & Okeke, C. (2004) Determinants of Health
Expenditure Growth of the OECD Countries: Jackknife Resampling Plan
Estimates. In Health Care Management Science. 7: 173 – 183.
182
Oliveira, M. & Pinto, G.P. (2005) Health Care Reform in Portugal: An Evaluation
of the NHS Experience. In Health Economics. 14: S203–S220.
Oliver, A. (2005) The English National Health Service: 1979-2005. In Health
Economics. 14: S75–S99.
Ottewill, R. (1996) The Netherlands. In Wall, A. (Ed.) Health Care Systems in
Liberal Democracies. New York: Routledge.
Parkin, D.W., McGuire, A.J. & Yuke, B.F. (1987) Aggregate Health Expenditures
nd National Income: Is Health Care a Luxury Good? In Journal of Health
Economics. 6: 109-127.
Parkin, D.W., McGuire, A.J. & Yuke, B.F. (1989) What Do International
Comparisons of Health care Expenditures really show? In Community Medicine,
11: 116 – 123.
Parks Thomas, C., Ritter, G. & Wallack, S.S. (2001) Growth in Prescription Drug
Spending Among Insured Elders. In Health Affairs. 20: 265 – 277.
Patel, K., & Rushefsky, M.E. (2001) Canada, In Thai, K.V., Wimberly, E.T. &
McManus, S.M. (Eds.), Handbook of International Health Care Systems. New
York: Marcel Dekker Inc.
Pauly, M. (1968) The Economics of Moral Hazard. Comment. In American
Economic Review. 85: 531 – 539.
Pauly, M. (1986) Taxation, Health Insurance, and Market Failure in the Medical
Economy. In Journal of Economic Literature. 24: 629 – 675.
Pedersen,
K.M., Christiansen, T. & Bech, M. (2005) The Danish Health Care
System: Evolution - Not Revolution - in a Decentralized System. In Health
Economics. 14: S41 - S57.
183
Perreira, J., Correia de Campos, A., Ramos, F. Simoes, J. & Reis, V. (1999) Health
Care Reform and Cost Containment in Portugal., In Mossialos, E. & Le Grand, J.
(Eds.) Health Care and Cost Containment in the European Union., Aldershot:
Ashgate.
Pfaff, M. (1990) Differences in Health Care Spending across Countries: Statistical
Evidence. In Journal of Health Politics, Policy and Law. 15: 1 – 24.
Podger, A. (1999) Reforming the Australian Health Care System: A Government
Perspective. In Health Affairs. 18: 111 – 113.
Pohl, C.M. (2002) The United States Health Care System. In Thai, K.H.,
Wimberly, E.T. & McManus, S.M. (Eds.), Handbook of International Health Care
Systems. New York: Marcel Dekker Inc.
Popovian, R., Johnson, K.A., Nichol, M.B. & Liu, G. (1999) The Impact of
Pharmaceutical Capitation to Primary Medical Groups on the Health Expenditures
of Medicare HMO Enrollees. In Journal of Managed Care Pharmacy. 5: 414 – 419.
Ranade, W. (1998) Markets and Health Care. A Comparative Analysis. London:
Longman.
Reinhardt, U.E. (2000) Health Care for the Ageing Baby Boom: Lessons from
Abroad. Journal of Economic Perspectives. 14:71 – 83.
Reinhardt, U. E. (2001) Perspectives on the Pharmaceutical Industry. In Health
Affairs. 20: 136 – 149.
Rice, T. (1992) Containing Health Care Costs in the United States., In Medical
Care Review. 49: 20 – 65.
Rice, T. (1998) The Economics of Health Reconsidered. Chicago: Health
Administration Press.
184
Rico, A & Lazenby, T. (2002) Spain. In Fried, B.J. & Gaydos, L.M. (eds.) World
Health Systems. Challenges and Perspectives. Chicago: Health Administration
Press.
Roberts, J. (1999, Sensitivity of Elasticity Estimates for OECD Health Care
Spending: Analysis of a Dynamic Heterogeneous Data Field. In Health
Economics. 8: 459 – 479.
Rodriguez – Alvarez, A. & Knox – Lovell, C.A. (2003) Excess capacity and
expense preference behaviour in National Health Systems: an application to the
Spanish public hospitals. Health Economics. 13: 157 – 169.
Roemer, M.I. (1977) Systems of Health Care. New York: Springer Publications.
Roemer, M.I. (1991) National Health Systems of the World: Vol.1. The Countries.
New York: Oxford University Press.
Saltman, R. (1998) Health Reform in Sweden: the road beyond cost containment.
In Ranade, W. (Ed.) Markets and Health Care, A Comparative Analysis. New
York: Longman.
Sayrs, L. W. (1989) Pooled Tine Series Analysis, in Lewis-Beck, M.S. (Ed.),
Quantitative Applications in the Social Sciences. Newbury Park: Sage
Publications..
Schuster, M.A., McGlynn, E.A. & Brook R.H. (1998) How Good Is the Quality of
Health Care in The United States. In The Milbank Quarterly. 76: 517 – 563.
Schut, F.T. & Van de Ven W.P.M.M. (2005) Rationing and competition in the
Dutch health-care system, In Health Economics. 14: S59–S74.
Sekhri, N. (2001) United States of America. In Wieners, W.W. (ed.) Global Health
Care Markets. A Comprehensive Guide to Regions. Trends, and Opportunities
Shaping the International Health Arena. Chicago: Health Administration Press.
185
Seshami M., & Gray, A. (2004) Ageing and Health Expenditure: The Red Herring
Argument Revisited. In Health Economics, 13, 303 – 314.
Shoji, A. & Hironari, A. (1999) Reform of the Japanese Health Insurance System,
In International Social Security Review. 52: 101-108.
Siciliani, L. & Hurst, J. (2004) Causes de la Disparité des Délais d’Attente en
Chirurgie Non Urgente dans les Pays de L’OCDE. In Revue Economique de
l’OCDE. 38: 107 – 137.
Sissouras, A., Karokis, A. & Mossialos, E. (1999) Health Care and Cost
Containment in Greece. In Mossialos, E. & Le Grand, J. (Eds.) Health Care and
Cost Containment in the European Union. Aldershot: Ashgate.
Soumerai, S.B. (2004) Benefits and Risks of Increasing Restrictions on Access to
Costly Drugs in Medicaid. In Health Affairs. 23: 135 -146.
Soumerai, S.B., Ross-Degnan, D., Avron, J., McLaughlin, T.J. & Choodnovski, I.
(1991) Effects of Medicaid Drug-Payments on Admission to Hospitals and Nursing
Homes., In The New England Journal of Medicine. 325: 1072 – 1077.
Spence, R. (1996) Italy, In Wall, A. (ed.), Health Care Systems in Liberal
Democracies. New York: Routledge.
Spicer, M.W. (1982) he Economics of Bureaucracy and the British National Health
Services. In Milbank Memorial Fund Quarterly. 60: 657 – 672.
Stano, M. (1985) An Analysis of the Evidence on Competition in the Physician
Services Market. In Journal of Health Economics. 4: 197 – 211.
Starr P. (1982) The Social Transformation of American Medicine: The Rise of a
Sovereign Profession and the Making of a Vast Industry. New York: Basic Books.
186
Stepan, A. & Sommersguter-Reichmann, M. (2005) Monitoring political decision-
making and its impact in Austria. In Health Economics. 14: S7–S23.
Stevens, J.B. (1993) The Economics of Collective Choice. Boulder: Westview
Press.
Stimson, J. A. (1985) Regression in Space and Time: A statistical essay. American
Journal of Political Science. 29: 914-974.
Studenmund, A.H. (1997) Using Econometrics. A Practical Guide. Reading, MA:
Addison-Wesley.
Swami, B. (2002) The German Health Care System. In Thai, K.V., Wimberly,
E.T. & McManus, S.M. (Eds), Handbook of International Health Care Systems.
New York: Marcel Dekker Inc.
Tamblyn, R., Laprise, R., Hanley, J.A., Abrahamowicz, M., Scott, S., Mayor, N.,
Hurley, J., Grad, R., Latimer, E. & Perreault, R. (2001) Adverse Events Associated
with Prescription Drug Cost sharing among Poor end Elderly Persons. Journal of
the American Medical Association. 285: 421 – 429.
Theurl, E. (1999) Health Expenditures and Cost Control in Austria. In Mossdialos,
E. & Le Grand, J. (Eds.) Health Care and Cost Containment in the European
Union. Aldershot: Ashgate.
Theurl, E. (1999) Some Aspects of the Reform of Health Care Systems in Austria,
Germany and Switzerland. In Health Care Analysis. 7: 331-354.
Tussing, A.D. & Wojtowycz, M.A. (1986) Physician-Induced Demand by Irish
GPs. In Social Science and Medicine. 23: 851 – 861.
U.S. General Accounting Office (1991) Health Care Spending Control: The
Experience of France, Germany and Japan. Washington DC: General Accounting
Office.
187
Upshaw, V.M. & Mathews Deal, K. (2002) The United States. In Fried, B.J. &
Gaydos, L.M., (Eds.) World Health Systems. Challenges and Perspectives.
Chicago: Health Administration Press.
Van het Loo, M., Kahan, J.P., & Okma, K.G.H. (1999) Developments in Health
Care Cost Containment in the Netherlands., In Mossialos, E. & Le Grand, J. (Eds.)
Health Care and Cost Containment in the European Union. Aldershot: Ashgate.
Van Kemenade, Y.W. (1997) Health Care in Europe. The Finance and
Reimbursement Systems of 18 European Countries. Ede Tijdstroom: Elsevier.
Von Otter, C. (1999) Cost Control in the Swedish Health Sector., In Powell, F.D.
& Wessen, A.F. (Eds.) Health Care Systems in Transition., Thousand Oaks, Sage.
Wall, A. (1996) Health Care Systems in Liberal Democracies., New York:
Routledge.
Wang, J., Jamison, D.T., Bos, E., Preker, A. & Peabody, J. (1999) Measuring
Country Performance on Health. Selected Indicators for 115 Countries.
Washington DC: The World Bank.
White, H. (1980) A Heteroscedasticity-Consistent Covariance Matrix Estimator
and a Direct Test for Heteroscedasticity. In Ecomometrica. 48: 817 – 838.
WHO (1996) Health Care Systems in Transition: Greece. Copenhagen: European
Observatory on Health Care Systems.
WHO (1996) Health Care Systems in Transition: Canada. Copenhagen: European
Observatory on Health Care Systems.
WHO (1999) Health Care Systems in Transition: United Kingdom. Copenhagen:
European Observatory on Health Care Systems.
188
WHO (1999) Health in Transition. Production Template and Questionnaire.
Copenhagen: European Observatory on Health Care Systems.
WHO (2000) Health Care Systems in Transition: Germany. Copenhagen: European
Observatory on Health Care Systems.
WHO (2000) Health Care Systems in Transition: Norway. Copenhagen: European
Observatory on Health Care Systems.
WHO (2000) A Quick Reference Compendium of Selected Key Terms used in The
World Health Report 2000. Geneva: WHO.
WHO (2000) Health Care Systems in Transition: Spain. Copenhagen: European
Observatory on Health Care Systems.
WHO (2001) Health Care Systems in Transition: Australia. Copenhagen: European
Observatory on Health Care Systems.
WHO (2001) Health Care Systems in Transition: Austria. Copenhagen: European
Observatory on Health Care Systems.
WHO (2001) Health Care Systems in Transition: Finland. Copenhagen: European
Observatory on Health Care Systems.
WHO (2001) Health Care Systems in Transition: Finland. London: European
Observatory on Health Care Systems.
WHO (2001) Health Care Systems in Transition: Italy. Copenhagen: European
Observatory on Health Care Systems.
WHO (2001) Health Care Systems in Transition: New Zealand. Copenhagen:
European Observatory on Health Care Systems.
189
WHO (2001) Health Care Systems in Transition: Sweden. Copenhagen: European
Observatory on Health Care Systems.
WHO (2002) Health Care Systems in Eight Countries: Trends and Challenges.
Copenhagen: European Observatory on Health Systems.
WHO (2003) Health Care Systems in Transition: Iceland. Copenhagen: European
Observatory on Health Care Systems.
WHO (2004) Health Care Systems in Transition: Germany. Copenhagen: European
Observatory on Health Care Systems.
WHO (2004) Health Care Systems in Transition: Netherlands. Copenhagen:
European Observatory on Health Care Systems.
WHO (2004) Health Care Systems in Transition: Portugal. Copenhagen: European
Observatory on Health Care Systems.
WHO. (2002) Health Care Systems in Transition: Denmark., London: European
Observatory on Health Care Systems.
Wilensky, G.R. & Rossiter, L.F. (1983) The Relative Importance of Physician-
Induced Demand in the Demand for Medical Care. In Milbank Memorial Fund
Quarterly. 61: 252 – 277.
Wolf, C. (1993) Markets or Governments. Choosing between Imperfect
Alternatives. Cambridge: The MIT Press.
Wolfe, P.R. & Moran, D. W. (1993) Global Budgeting in the OECD Countries. In
Health Care Financing Review. 14: 55 – 76.
Worldbank, (2003) World Development Indicators. Washington D.C.: The World
Bank.
190
Wörz, M. & Busse, R. (2005) Analysing the impact of health-care systemchange in
the EU member states – Germany. In Health Economics. 14: S133–S149.
Yaijima, R. & Takayanagi, R. (2002) The Japanese Health Care System. Citizen
Complaints and Citizen Possibilities. In Thai, K.V., Wimberly, E.T. & McManus,
S.M. (Eds), Handbook of International Health Care Systems. New York: Marcel
Dekker Inc.
Yoshikawa, A. & Bhattacharya, J. (2002) Japan., In Fried, B. & Gaydos, L.M.,
(Eds.) World Health Systems. Challenges and Perspectives. Chicago: Health
Administration Press.
Yoshikawa, A. (2001) Japan. In Wieners, W.W. (Ed.) Global Health Care
Markets, San Fransisco: Jossey-Bass.
191
Appendix: Tolerance and Variance Inflation Factor
US
No
US US
No
US US
No
US
70-
00
70-
00
70-
00
70-
00
75-
95
75-
95
75-
95
75-
95
80-
00
80-
00
80-
00
80-
00
Tol Vif Tol Vif Tol Vif Tol Vif Tol Vif Tol Vif
System
Intercept . 0.00 . 0.00 . 0.00 . 0.00 . 0.00 . 0.00
GDP 0.64 1.55 0.65 1.53 0.66 1.51 0.69 1.46 0.68 1.47 0.70 1.43
Pubshare 0.62 1.61 0.56 1.79 0.59 1.70 0.52 1.92 0.59 1.69 0.53 1.87
NHS 0.70 1.43 0.71 1.40 0.69 1.45 0.70 1.43 0.66 1.51 0.69 1.45
Pop65 0.72 1.40 0.73 1.37 0.79 1.27 0.79 1.26 0.82 1.22 0.83 1.20
Urban 0.70 1.44 0.58 1.74 0.61 1.63 0.50 2.02 0.56 1.78 0.46 2.15
Structure
Intercept . 0.00 . 0.00 . 0.00 . 0.00 . 0.00 . 0.00
GDP 0.48 2.08 0.49 2.02 0.60 1.67 0.61 1.63 0.54 1.86 0.55 1.81
Pubshare 0.62 1.61 0.44 2.26 0.61 1.64 0.37 2.68 0.57 1.74 0.39 2.55
Inpatient 0.36 2.81 0.27 3.73 0.36 2.81 0.24 4.10 0.36 2.74 0.27 3.73
Pharma 0.48 2.10 0.49 2.02 0.47 2.12 0.49 2.02 0.49 2.03 0.52 1.91
Kidney 0.47 2.12 0.46 2.17 0.52 1.91 0.50 2.01 0.50 2.00 0.48 2.08
Pop65 0.67 1.50 0.62 1.60 0.71 1.42 0.64 1.55 0.66 1.51 0.62 1.61
Urban 0.64 1.56 0.63 1.58 0.57 1.77 0.55 1.82 0.45 2.22 0.44 2.27
Supply
Intercept . 0.00 . 0.00 . 0.00 . 0.00 . 0.00 . 0.00
GDP 0.40 2.47 0.42 2.40 0.45 2.22 0.48 2.10 0.46 2.19 0.48 2.10
Publshare 0.64 1.57 0.69 1.46 0.58 1.73 0.63 1.59 0.53 1.89 0.61 1.64
FFS 0.87 1.15 0.92 1.09 0.86 1.16 0.91 1.09 0.84 1.18 0.89 1.12
Global 0.62 1.62 0.64 1.56 0.65 1.54 0.70 1.44 0.61 1.64 0.69 1.44
DRG 0.66 1.52 0.70 1.43 0.71 1.41 0.79 1.27 0.61 1.63 0.66 1.53
Pop65 0.71 1.41 0.72 1.39 0.76 1.31 0.76 1.31 0.78 1.28 0.79 1.26
Urban 0.71 1.40 0.62 1.62 0.67 1.49 0.57 1.76 0.59 1.71 0.50 2.02
Demand
Intercept . 0.00 . 0.00 . 0.00 . 0.00 . 0.00 . 0.00
GDP 0.81 1.23 0.80 1.26 0.79 1.27 0.74 1.35 0.75 1.33 0.73 1.37
Pubshare 0.65 1.55 0.85 1.17 0.54 1.84 0.82 1.23 0.74 1.35 0.84 1.19
Pocket 0.54 1.87 0.66 1.51 0.39 2.57 0.50 1.99 0.66 1.51 0.70 1.44
Pop65 0.77 1.29 0.80 1.25 0.81 1.24 0.80 1.25 0.84 1.19 0.85 1.18
Urban 0.77 1.30 0.73 1.37 0.66 1.51 0.60 1.66 0.77 1.30 0.74 1.36
192
(Tolerance and Variance Inflation Factor continued)
Full 1
Intercept . 0.00 . 0.00 . 0.00 . 0.00 . 0.00 . 0.00
GDP 0.27 3.70 0.23 4.30 0.21 4.81 0.23 4.39 0.47 2.14 0.32 3.16
Pubshare 0.17 6.02 0.13 7.85 0.10 10.27 0.11 9.09 0.19 5.38 0.11 9.41
NHS 0.10 9.87 0.07 13.91 0.10 10.46 0.07 13.50 0.11 9.32 0.08 13.17
Inpat 0.17 5.78 0.17 5.86 0.17 5.95 0.16 6.21 0.18 5.71 0.16 6.09
Pharma 0.45 2.24 0.49 2.05 0.34 2.94 0.35 2.87 0.42 2.39 0.45 2.22
Kidney 0.39 2.58 0.38 2.63 0.30 3.37 0.30 3.29 0.48 2.08 0.50 2.00
FFS 0.24 4.17 0.27 3.75 0.26 3.84 0.28 3.55 0.28 3.54 0.31 3.22
Global 0.33 3.07 0.31 3.18 0.25 3.94 0.29 3.42 0.40 2.51 0.36 2.75
DRG 0.49 2.05 0.53 1.89 0.44 2.27 0.47 2.11 0.41 2.45 0.46 2.19
Pocket 0.26 3.78 0.19 5.31 0.16 6.09 0.14 7.02 0.32 3.17 0.16 6.31
Pop65 0.37 2.71 0.36 2.78 0.34 2.94 0.31 3.24 0.45 2.22 0.44 2.25
Urban 0.31 3.24 0.30 3.34 0.28 3.53 0.27 3.75 0.30 3.34 0.29 3.46
Full 2
Intercept . 0.00 . 0.00 . 0.00 . 0.00 . 0.00 . 0.00
GDP 0.28 3.63 0.25 4.01 0.28 3.63 0.25 4.01 0.49 2.03 0.37 2.72
Pubshare 0.29 3.50 0.41 2.45 0.29 3.50 0.41 2.45 0.31 3.27 0.32 3.12
Inpat 0.35 2.88 0.22 4.54 0.35 2.88 0.22 4.54 0.38 2.66 0.19 5.35
Pharma 0.48 2.10 0.51 1.95 0.48 2.10 0.51 1.95 0.45 2.24 0.47 2.11
Kidney 0.39 2.54 0.40 2.48 0.39 2.54 0.40 2.48 0.50 1.99 0.54 1.87
Global 0.34 2.93 0.33 3.05 0.34 2.93 0.33 3.05 0.41 2.42 0.41 2.42
DRG 0.49 2.02 0.54 1.85 0.49 2.02 0.54 1.85 0.43 2.33 0.47 2.13
Pocket 0.30 3.35 0.32 3.11 0.30 3.35 0.32 3.11 0.36 2.78 0.30 3.28
Pop65 0.40 2.49 0.37 2.70 0.40 2.49 0.37 2.70 0.48 2.10 0.45 2.21
Urban 0.38 2.60 0.35 2.88 0.38 2.60 0.35 2.88 0.38 2.61 0.34 2.98
Abstract (if available)
Abstract
This study has three parts. First, health care systems are analyzed along their key organizational, financing and delivery characteristics. Next, trends in total health expenditures, source of financing, expenditure structure and technology are described. The combined results show that the United States' health care system very similar to other health care systems with respect to organizational and delivery characters but substantially differs from the other health care systems in terms of financing mechanism. In addition, the United States has much higher total health expenditures and a much smaller public share of total health expenditures than any other country included in this study.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Delivering better care for children with special health care needs: analyses of patient-centered medical home and types of insurance
PDF
An analysis of the impact of the total educational support system direct-instruction model on the California standards test performance of English language learners at experimental elementary school
PDF
The impact of adherence to guidelines on the health care expenditures of COPD patients
PDF
Developing an agent-based simulation model to evaluate competition in private health care markets with an assessment of accountable care organizations
PDF
Implementation of peer providers in integrated health care settings
PDF
Three essays on estimating the effects of government programs and policies on health care among disadvantaged population
PDF
Examining the adoption of electronic health records system in patient care and students’ education using the GAP analysis approach
PDF
Value in health in the era of vertical integration
PDF
Impact of pharmacy-based transitional care on healthcare utilization and costs
PDF
Mental health service use by children and youth in the child welfare system: a focus on need and predisposing factors and caregiver type
PDF
Vehicle routing and resource allocation for health care under uncertainty
PDF
A series of longitudinal analyses of patient reported outcomes to further the understanding of care-management of comorbid diabetes and depression in a safety-net healthcare system
PDF
Understanding barriers in consumer health advocacy: an improvement study
PDF
Foster care placement disruption monitoring and reduction: an innovation study
PDF
Health care disparities and the influence of nurse leader cultural competency: an evaluation study
PDF
Eliciting perspectives on palliative care: outpatient visits, advance care planning, and the impact of COVID-19
PDF
Evaluation of the current state of management of children with central nervous system tumors in Paraguay through a comprehensive needs assessment
PDF
Three essays on emerging issues in hemophilia care
PDF
Barriers in care access for the marginalized individual: a provider’s role
PDF
Impacts of system of system management strategies on system of system capability engineering effort
Asset Metadata
Creator
Verbelen, Bart
(author)
Core Title
The impact of public expenditures on health care on total health expenditures: an exploratory analysis of selected OECD countries
School
School of Policy, Planning, and Development
Degree
Doctor of Philosophy
Degree Program
Public Administration
Publication Date
11/21/2007
Defense Date
12/16/2005
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
comparative health systems,health care financing,Health Economics,OAI-PMH Harvest,OECD
Place Name
USA
(countries)
Language
English
Advisor
Melnick, Glenn (
committee chair
), Biller, Robert (
committee member
), Nichol, Michael B. (
committee member
)
Creator Email
verbelen@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m937
Unique identifier
UC1319778
Identifier
etd-Verbelen-20071121 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-486944 (legacy record id),usctheses-m937 (legacy record id)
Legacy Identifier
etd-Verbelen-20071121.pdf
Dmrecord
486944
Document Type
Dissertation
Rights
Verbelen, Bart
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
comparative health systems
health care financing
OECD